Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Hi All I now have a working work-around: My app can (finally) send SMSes via Telepathy again. Previously I was using ContactMessenger object. Now I have found a vector using the ConversationChannel object. It took a lot of hacking around due to my weak understanding of c++, but in the end I hit on a solution that required minimal change to both my code, and to ConversationChannel. I grabbed the code below: https://github.com/nemomobile/nemo-qml-plugin-messages/blob/master/src/conversationchannel.h https://github.com/nemomobile/nemo-qml-plugin-messages/blob/master/src/conversationchannel.cpp then pimped it slightly so it that on a successful sendMessage, it emits the signal finished Previously, using ContactMessenger my code was void TelepathyHelper::sendSMS(const QString contactIdentifier, const QString message) { Tp::AccountPtr acc = Tp::Account::create(TP_QT_ACCOUNT_MANAGER_BUS_NAME, QLatin1String(/org/freedesktop/Telepathy/Account/ring/tel/account0)); messenger = Tp::ContactMessenger::create(acc, contactIdentifier); connect(messenger-sendMessage(message), SIGNAL(finished(Tp::PendingOperation*)), SLOT(onSendMessageFinished(Tp::PendingOperation*))); } Now (temporarily) using ConversationChannel my code is: void TelepathyHelper::sendSMS(const QString contactIdentifier, const QString message) { ConversationChannel *conversationChannel = new ConversationChannel(/org/freedesktop/Telepathy/Account/ring/tel/account0, contactIdentifier); connect(conversationChannel, SIGNAL(finished(Tp::PendingOperation*)), SLOT(onSendMessageFinished(Tp::PendingOperation*))); //this works! SMS is sent yippeee !!! conversationChannel-sendMessage(message); } Cheers Chris Am 2015-08-12 14:26, schrieb christopher.l...@thurweb.ch: Hi Matt . My main focus now is finding a code approach that works with the current installed packages. I think I have found a vector, but need to do some more coding to know for sure. . Thanks Chris ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Answering my own (academic) question: The package install history can be found in /var/log/zypp/history [root@Jolla nemo]# cd /var/log/zypp/ [root@Jolla zypp]# grep -n telepathy-qt5| history 3423:2014-04-11 20:22:58|install|telepathy-qt5|0.9.3+git7-1.4.2|armv7hl||jolla|2591a70ed6d6a2ff08317e0ede1709239e973c4f1d37141a704ed2bef1a16595 4048:2014-06-09 21:50:03|install|telepathy-qt5|0.9.3+git9-1.5.1|armv7hl||jolla|084265e82b72ef3f27d676f451f6893aeae350c7475d8979bde8ed7dd62990c6 4730:2014-07-20 14:07:11|install|telepathy-qt5|0.9.3+git9-1.5.2|armv7hl||jolla|6c02b8c917fb121929d258d5e48e668204889ecc86278ae874f0f3d30c9d4ab8 5329:2014-12-29 11:10:39|install|telepathy-qt5|0.9.4+git2-1.6.2|armv7hl||jolla|c1c583d34911c2c0b43c2c3731e4e0d48febbd1fbcc968ca96467ab806408fa9 6082:2015-03-05 19:53:59|install|telepathy-qt5|0.9.4+git2-1.7.2|armv7hl||jolla|7dcece67f3f83fcb847d86e9548bf34067f19a1f7d80c67e2f0139a77ea6320d 6701:2015-05-04 21:10:50|install|telepathy-qt5|0.9.4+git2-1.6.2|armv7hl||jolla|c1c583d34911c2c0b43c2c3731e4e0d48febbd1fbcc968ca96467ab806408fa9 8154:2015-07-17 22:29:15|install|telepathy-qt5|0.9.4+git2-1.6.5|armv7hl||jolla|04369ec62817d31d8620541bea9b57ed8de55b11327c2baffa8499ac2f212e9b Then looking in github once again: telepathy-qt5 0.9.4+git2-1.6.2 installed to my Jolla on 2014-12-29 is probably when the problem started. Chris Am 2015-08-12 07:12, schrieb christopher.l...@thurweb.ch: Hi Matt thanks for taking the time to have a look as well. In the meantime I have opened a bug report in the mer / nemomobile bugzila https://bugs.nemomobile.org/show_bug.cgi?id=847. Our of pure academic interest I am curious to find out the change history of the telepathy packages installed on the device - to see which update broke the interface between the 2 packages. On a more practical vein, given that the Jolla SMS app also uses Telepathy, and is able to send SMSes, it must be taking a different route through the Telepathy code: so I am now digging through the nemo-qml-plugin-messages to better understand this. They seem to use the textChannel.send() method, whereas I was using the contactMessenger.sendMessage() Chris ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Hi Matt I went through the release history using this link: https://github.com/nemomobile-packages/telepathy-qt/releases Given that the numbers of the releases in Git are very similar to the number scheme of the packages on my device, I assume (always dangerous) a link. My main focus now is finding a code approach that works with the current installed packages. I think I have found a vector, but need to do some more coding to know for sure. One thing that did not click until this morning is that the telepathy code in the Collabora git is C, while the equivalent in nemomobile is C++. I wonder how is the one translated to the other? Thanks Chris Am 2015-08-12 10:48, schrieb Matthew Vogt: Hi Chris. You can browse the commit history of projects in github by clicking on the little 'clock' button at the top of the project page: https://github.com/nemomobile-packages/telepathy-qt/tree/master/telepathy-qt Or, by using the commits URI: https://github.com/nemomobile-packages/telepathy-qt/commits/master/telepathy-qt Unfortunately, it's not so easy to map from commits in the project history to releases... Thanks, Matt ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] OS X SDK install fails
maybe there's a problem with 4.3.30? I am running successfully with 4.3.26 on Yosemite (10.10.5). However both the SDK and OSX have been through several upgrades / re-intalls over the past few years. Chris Am 2015-08-12 14:02, schrieb Michał Radwański: Hello. I can't install SDK on OS X Yosemite. I got VirtualBox 4.3.30. After running last version of SDK installer, I get information about aborted installation: http://i.imgur.com/rcf20Ff.jpg Any ideas? ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Hi Matt Having spent some hours pouring over the nemomobile TelepathyQt / Mission-Control code in Github, it looks to me like TelepathyQt and MissionControl master branches are out of sync (and hence the packages on my Jolla will also be out of sync). https://github.com/nemomobile-packages/telepathy-mission-control/tree/master/telepathy-mission-control/xml still has the interface in draft form (Channel_Dispatcher_Interface_Messages_DRAFT.xml), but in TelepathyQt the equivalent is now Channel_Dispatcher_Interface_Messages1.xml (i.e. undrafted). I have also had a peek in the collabora git for telepathy-mission-control, and find these 2 commits that undraft the interface on the mission-control side. https://git.collabora.com/cgit/freedesktop.org-mirror/telepathy/telepathy-mission-control.git/commit/?id=ba72fef52e9769e3da6b7f4dddbb8ffdc1cecdde https://git.collabora.com/cgit/freedesktop.org-mirror/telepathy/telepathy-mission-control.git/commit/?id=d24d5591801b9b874cc328c7d516a86e50a1bc2e I guess that the Telepathy modules previously on my device both had the interface in DRAFT form which explains why the code previously worked. What is the best way of reporting this? https://bugs.nemomobile.org/ ? Chris Am 2015-08-10 02:15, schrieb Matthew Vogt: Mission control appears to implement the org.freedesktop.Telepathy.ChannelDispatcher.Interface.Messages1 interface now - see: https://github.com/nemomobile-packages/telepathy-qt/commit/6f0177bc024f3f7bce94c0abc283fba1c3b8a0d1 Perhaps your telepathy-qt packages and your telepathy-mission-control packages are not in sync? Thanks, Matt ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Am 2015-08-12 04:52, schrieb Matthew Vogt: Hi Chris. Although I can't speak with any authority on Telepathy, I agree with your assessment. I think that nemomobile issues are handled using the Mer bugzilla these days, at https://bugs.merproject.org/ Thanks, Matt From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of christopher.l...@thurweb.ch [christopher.l...@thurweb.ch] Sent: Wednesday, August 12, 2015 1:09 AM To: devel@lists.sailfishos.org Subject: Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages Hi Matt Having spent some hours pouring over the nemomobile TelepathyQt / Mission-Control code in Github, it looks to me like TelepathyQt and MissionControl master branches are out of sync (and hence the packages on my Jolla will also be out of sync). https://github.com/nemomobile-packages/telepathy-mission-control/tree/master/telepathy-mission-control/xml still has the interface in draft form (Channel_Dispatcher_Interface_Messages_DRAFT.xml), but in TelepathyQt the equivalent is now Channel_Dispatcher_Interface_Messages1.xml (i.e. undrafted). I have also had a peek in the collabora git for telepathy-mission-control, and find these 2 commits that undraft the interface on the mission-control side. Hi Matt thanks for taking the time to have a look as well. In the meantime I have opened a bug report in the mer / nemomobile bugzila https://bugs.nemomobile.org/show_bug.cgi?id=847. Our of pure academic interest I am curious to find out the change history of the telepathy packages installed on the device - to see which update broke the interface between the 2 packages. On a more practical vein, given that the Jolla SMS app also uses Telepathy, and is able to send SMSes, it must be taking a different route through the Telepathy code: so I am now digging through the nemo-qml-plugin-messages to better understand this. They seem to use the textChannel.send() method, whereas I was using the contactMessenger.sendMessage() Chris https://git.collabora.com/cgit/freedesktop.org-mirror/telepathy/telepathy-mission-control.git/commit/?id=ba72fef52e9769e3da6b7f4dddbb8ffdc1cecdde https://git.collabora.com/cgit/freedesktop.org-mirror/telepathy/telepathy-mission-control.git/commit/?id=d24d5591801b9b874cc328c7d516a86e50a1bc2e I guess that the Telepathy modules previously on my device both had the interface in DRAFT form which explains why the code previously worked. What is the best way of reporting this? https://bugs.nemomobile.org/ ? Chris Am 2015-08-10 02:15, schrieb Matthew Vogt: Mission control appears to implement the org.freedesktop.Telepathy.ChannelDispatcher.Interface.Messages1 interface now - see: https://github.com/nemomobile-packages/telepathy-qt/commit/6f0177bc024f3f7bce94c0abc283fba1c3b8a0d1 Perhaps your telepathy-qt packages and your telepathy-mission-control packages are not in sync? Thanks, Matt ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Hi Matt I have now added the debug packages to the device: Am 2015-08-10 02:15, schrieb Matthew Vogt: Mission control appears to implement the org.freedesktop.Telepathy.ChannelDispatcher.Interface.Messages1 interface now - see: https://github.com/nemomobile-packages/telepathy-qt/commit/6f0177bc024f3f7bce94c0abc283fba1c3b8a0d1 Perhaps your telepathy-qt packages and your telepathy-mission-control packages are not in sync? Thanks, Matt Installedcontactsd-1.2.65.3-1.30.2.armv7hl Installedtelepathy-accounts-signon-0.1.27.2-1.12.3.armv7hl Installedtelepathy-farstream-0.4.0-1.1.6.armv7hl Installedtelepathy-gabble-0.18.3+git1-1.7.1.armv7hl Installedtelepathy-glib-0.20.3-1.2.3.armv7hl Installedtelepathy-mission-control-5.15.0+git9-1.6.3.armv7hl Installed telepathy-mission-control-debuginfo-5.15.0+git9-1.6.3.armv7hl Installedtelepathy-qt5-0.9.4+git2-1.6.5.armv7hl Installedtelepathy-qt5-debuginfo-0.9.4+git2-1.6.5.armv7hl Installedtelepathy-qt5-farstream-0.9.4+git2-1.6.5.armv7hl Installedtelepathy-ring-2.4.0-1.16.1.armv7hl Installedtelepathy-ring-debuginfo-2.4.0-1.16.1.armv7hl Installedvoicecall-qt5-plugin-telepathy-0.6.10-1.22.1.armv7hl After that I get the debug output at the bottom of this post. It strikes me that the error is returned very early in the process b.t.w, checking through sent SMSes proves that the code worked back in November 2014. Cheers Chris [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating new DBusProxy [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedStatuses = QSet(0) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedFeatures = QSet(QPair(Tp::Account,3) , QPair(Tp::Account,2) , QPair(Tp::Account,1) , QPair(Tp::Account,0) ) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating a new SimpleTextObserver [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Client registered - busName: org.freedesktop.Telepathy.Client.TpQtSO__1_145_0 objectPath: /org/freedesktop/Telepathy/Client/TpQtSO__1_145_0 interfaces: (org.freedesktop.Telepathy.Client.Observer) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Observer TpQtSO__1_145_0 registered [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Contact id requires normalization. Queueing events until it is normalized [D] TelepathyHelper::onSendMessageFinished:70 - TelepathyHelper::onSendMessageFinished [D] TelepathyHelper::onSendMessageFinished:72 - Error sending message: org.freedesktop.Telepathy.Error.NotImplemented - Channel Dispatcher implementation (e.g. mission-control), does not support interface CD.I.Messages [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ObserveChannels: account: /org/freedesktop/Telepathy/Account/ring/tel/account0 , connection: /org/freedesktop/Telepathy/Connection/ring/tel/ring [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Inserting to factory cache proxy for QPair(org.freedesktop.Telepathy.AccountManager,/org/freedesktop/Telepathy/Account/ring/tel/account0) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating new DBusProxy [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Connecting to ConnectionError() [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Connecting to StatusChanged() [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Connecting to SelfHandleChanged() [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating new HandleContext for /org/freedesktop/Telepathy/Connection/ring/tel/ring [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedStatuses = QSet(2, 0, 1, 4294967295) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedFeatures = QSet(QPair(Tp::Connection,2) , QPair(Tp::Connection,5) , QPair(Tp::Connection,4) , QPair(Tp::Connection,7) , QPair(Tp::Connection,6) , QPair(Tp::Connection,1) , QPair(Tp::Connection,0) ) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Inserting to factory cache proxy for QPair(:1.19,/org/freedesktop/Telepathy/Connection/ring/tel/ring) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating new DBusProxy [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Creating new Channel: /org/freedesktop/Telepathy/Connection/ring/tel/ring/text16 [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Connecting to Channel::Closed() signal [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: Connection to owning connection's lifetime signals [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedStatuses = QSet(0) [D] Tp::Debug::invokeDebugCallback:146 - tp-qt 0.9.4 DEBUG: ReadinessHelper: new supportedFeatures = QSet(QPair(Tp::Channel,1) , QPair(Tp::Channel,0) ) [D] Tp::Debug::invokeDebugCallback:146 -
Re: [SailfishDevel] Sending SMS via Telepathy: Error Channel Dispatcher .. does not support interface CD.I.Messages
Hi Mat Am 2015-08-10 02:15, schrieb Matthew Vogt: Mission control appears to implement the org.freedesktop.Telepathy.ChannelDispatcher.Interface.Messages1 interface now - see: https://github.com/nemomobile-packages/telepathy-qt/commit/6f0177bc024f3f7bce94c0abc283fba1c3b8a0d1 Perhaps your telepathy-qt packages and your telepathy-mission-control packages are not in sync? Thanks, Matt Sounds interesting. These are the telepathy packages I have installed on the device: [root@Jolla nemo]# pkcon search name telepathy | grep Installed Installedcontactsd-1.2.65.3-1.30.2.armv7hl Installedtelepathy-accounts-signon-0.1.27.2-1.12.3.armv7hl Installedtelepathy-farstream-0.4.0-1.1.6.armv7hl Installedtelepathy-gabble-0.18.3+git1-1.7.1.armv7hl Installedtelepathy-glib-0.20.3-1.2.3.armv7hl Installedtelepathy-mission-control-5.15.0+git9-1.6.3.armv7hl Installedtelepathy-qt5-0.9.4+git2-1.6.5.armv7hl Installedtelepathy-qt5-farstream-0.9.4+git2-1.6.5.armv7hl Installedtelepathy-ring-2.4.0-1.16.1.armv7hl Installedvoicecall-qt5-plugin-telepathy-0.6.10-1.22.1.armv7hl I just tried to do a pkcon update on the telepathy-qt5 and telepathy-mission-control packages, but there are no update candidates (currently) available. b.t.w I am on SailfishOS 1.1.7.25 (Björnträsket) (armv7hl) Chris ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] pageStack.pop() - send paramater to prev page
Sali Markus I would also love to see a pop() method with parameters so I can pass info back to the previous page, jus like the push() As a workaround I just set the page's properties directly immediately before the pop() call e.g: AreaSelectionPage { id: areaSelectionPage onBackPageWithInfo: { mainPage.areaSet = true; mainPage.area_id = area_id; pageStack.pop(mainPage); } onCancelled: pageStack.pop(mainPage); } mfg Chris On 05.06.14 13:48, i...@flyingfischer.ch wrote: Beginners follow up question Hello while transfering parameters works fine with pageStack.push(Qt.resolvedUrl(nextPage.qml, {topic: myTopic})) there seems not to be any method like this in pageStack.pop() How can I give back a parameter from the top of the stack to the previous page? Thanks for you inputs and patience! Markus Fischer -- I tried: Page { id: onTopOfStack property string topic ... Column { id: column ... Button { id: myTopic text: qsTr(MyTopic) anchors.horizontalCenter: parent.horizontalCenter onClicked: { topic: MyTopic pageStack.pop(); } } -- Page { id: previousPage property string topic ... SilicaFlickable { id: mainFlickable ... Column { id: mainColumn ... PageHeader { id: header title: qsTr(topic) } ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] pageStack.pop() - send paramater to prev page
Hi Markus In my app I chose to instantiate all my Pages in the same place. Each of the pages has a declaration in a separate qml file (e.g. MainPage.qml, AreaSelectionPage.qml). With this architecture none of the declarations need to know about the previous or next page. It is not their concern. Instead they emit signals NextPage, BackPageWithInfo, and Cancelled. In the signal handler in the page instantation I do the actual pushing and popping, and because all pages are instantiated in the same scope, they are all available to be pushed / popped to. (see code example below). The downside is that all the pages are created on startup (regardless of if they are ever visited). For lightweight pages this is not a problem, but for some pages this might be a problem. For these I dynamically create the page content when the page is visible. Grüsse Chris ApplicationWindow { id: appWindow ... initialPage: mainPage MainPage { id: mainPage onNextPage: { if (pageType ==SMS) { console.log(smsType is:+ smsType) if (smsType ==Default) pageStack.push(smsPage, {lati: mainPage.getLati(), longi: mainPage.getLongi(), alti: mainPage.getAlti(), area_id: area_id, template_id: template_id, msg_status: msg_status, lastPage: mainPage}) } else { pageStack.push(areaSelectionPage) } } } AreaSelectionPage { id: areaSelectionPage onBackPageWithInfo: { mainPage.areaSet = true; mainPage.area_id = area_id; pageStack.pop(mainPage); } onCancelled: pageStack.pop(mainPage); } SMSPage { id: smsPage onNextPage: { pageStack.push(contactSelectionPage, {area_id: area_id, template_id: template_id}) } } ContactSelectionPage { id: contactSelectionPage onBackPageWithInfo: { console.log(About to pop smsPage; contactName:+ contactName + , contactPhone:+ contactPhone); smsPage.contactName = contactName; smsPage.contactPhone = contactPhone; smsPage.lastPage = contactSelectionPage; pageStack.pop(smsPage); } onCancelled: pageStack.pop(smsPage); } } On 05.06.14 14:43, i...@flyingfischer.ch wrote: Thanks for your inputs! Calling PreviousPage.topic = MyTopic; pageStack.pop(); gives me a ReferenceError: PreviousPage is not defined. PreviousPage.qml and property string topic do exist. Thanks for your patience! Markus ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Jarko and Juha Right now QtCreator is deploying as RPM without a problem - whereas last week it was consistently giving the error. I am sure it will rear its ugly head soon, so I will report back then following your advice. In the meantime, some more background info. I am on OSX 10.8.5 on a MBA. I am running the very latest SailfishOS SDK, but have had this problem with previous versions to. I have lots of applicaitons running on the MBA, including QtCreator Lotus Notes, MS Office, Thunderbird, Chrome, Safari, Terminal, a VPN etc. I almost never shutdown / restart unless forced to. . I move between various networks - at home, office, hotspot on Jolla, VPN to office etc. I do a lot of Sailfish Development mobile - on the train, and thus often open and shut the laptop lid during development The problem can occur out of the blue: i.e. 30 seconds previously it successfully deployed - I make a small one line change to the code, try to redeploy, and bingo: the problem occurs. The progress bar for the build was at about 50 % complete or more. The timeout comes after seconds rather than minutes (but I need to reproduce it again to give more accurate timings). During the problem I have been able to ssh into the SDK and Emulator with no noticeable problem. Chris On 02.06.14 10:23, Jarko Vihriala wrote: when that happens, have you tried to SSH into the MerSDK from command line. We have seen issues where the SSH key exchange takes long time for some reason , currently that's under investigation. If you want to help in that, - you could do a SSH session inside MerSDK and run journalctl -fa in terminal - run deploy commands from Qt Creator - observe the output, is there a +10s timeout in the log somewhere (looking at timestamps)? thanks, Jarko *From:* devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Tone Kastlunger [users.giulie...@gmail.com] *Sent:* Saturday, May 31, 2014 11:28 AM *To:* Sailfish OS Developers *Subject:* Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server. Hi I for once happen to get the issue every once in a while even during binary deployment. Best, tk On Thursday, May 29, 2014, Christopher Lamb christopher.l...@thurweb.ch mailto:christopher.l...@thurweb.ch wrote: Hi All I am still suffering from this issue, and getting more and more frustrated! Right now it is refusing to deploy as RPM to the Emulator. Is anything I can do to better debug / understand this issue? The timeout settings previously suggested are probably red-herrings, as they apply to the device (phone / Emulator), whereas the error is a timeout to the SDK. All help greatly appreciated. Grüsse Chris On 28.04.14 08:01, christopher.l...@thurweb.ch wrote: Hi Thanks. I found a well hidden {1} timeout setting in the preferences / devices and increased this from the default 2 secs to 10 secs. This does seem to help, but as always with sporadic problems only time will tell if this is a fix, or if the problem is just dormant for a while. What I don't understand is, given the error says Could not connect to MerSDK Virtual Machine why should a timeout for connecting to a Jolla device play any role? Is the error misleading? is the problem actually in connecting FROM the MerSDK to the Jolla? Chris {1} when the preferences pane opens, at least on my OSX host the timeout setting is offscreen in an embedded pane with scroll bars in the middle of the preferences pane. Only if you scroll down in the embedded pane are further settings including the timeout revealed. Truly a wonderful bit of UI design 8-) Zitat von fasza2mob...@gmail.com: I believe the timeout setting can be adjusted in QtCreator where you set up your mer device(ssh password, etc). By increasing this value you can probably prevent the timeout issue altogether. I don't have the SDK in front of me to give you the exact menu you need to open, but it isn't too hard to find once you know what you're looking for. On Sat Apr 26 2014 13:41:20 GMT+0100 (BST), christopher.l...@thurweb.ch wrote: Hi All I am suffering from sporadic bouts of the error: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server One moment I can be happily hacking and deploying code to my Jolla without error. Then I make a small change to the code, try to deploy
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Jarko Have got on the train, started deploying, and get the problem. Below is output from the SDK and QtCreator. I am starting to wonder if the problem is related to networking in some way. This morning I successfully deployed at home - connected to the home wireless network. Now I am on the train - when I deployed / got the error) I was not connected to any network. (I have just opened the Jolla hotspot to send this mai). Chris SDK journalctl -fa output Apple-Pip:~ christopherlamb$ ssh -p -i ~/SailfishOS/vmshare/ssh/private_keys/engine/root root@localhost Last login: Wed Apr 30 07:00:35 2014 [root@SailfishSDK ~]# journalctl -fa Logs begin at Tue, 03 Jun 2014 06:00:29 +. Jun 03 06:48:07 SailfishSDK sshd[10191]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:48:07 SailfishSDK sshd[9470]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:48:07 SailfishSDK sshd[9447]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:48:07 SailfishSDK sshd[9447]: syslogin_perform_logout: logout() returned an error Jun 03 06:48:07 SailfishSDK systemd-logind[143]: Removed session c36. Jun 03 06:48:07 SailfishSDK systemd-logind[143]: Removed session c32. Jun 03 06:48:07 SailfishSDK systemd-logind[143]: Removed session c33. Jun 03 06:48:25 SailfishSDK sshd[10683]: Accepted publickey for root from 10.0.2.2 port 59953 ssh2 Jun 03 06:48:25 SailfishSDK systemd-logind[143]: New session c37 of user root. Jun 03 06:48:25 SailfishSDK sshd[10683]: pam_unix(sshd:session): session opened for user root by (uid=0) Jun 03 06:48:54 SailfishSDK sshd[10702]: Accepted publickey for mersdk from 10.0.2.2 port 59955 ssh2 Jun 03 06:48:54 SailfishSDK systemd-logind[143]: New session c38 of user mersdk. Jun 03 06:48:54 SailfishSDK sshd[10702]: pam_unix(sshd:session): session opened for user mersdk by (uid=0) Jun 03 06:49:00 SailfishSDK sshd[10707]: Accepted publickey for mersdk from 10.0.2.2 port 59958 ssh2 Jun 03 06:49:00 SailfishSDK systemd-logind[143]: New session c39 of user mersdk. Jun 03 06:49:00 SailfishSDK sshd[10707]: pam_unix(sshd:session): session opened for user mersdk by (uid=0) Jun 03 06:49:01 SailfishSDK sshd[10707]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:49:01 SailfishSDK systemd-logind[143]: Removed session c39. Jun 03 06:49:01 SailfishSDK sshd[11124]: Accepted publickey for mersdk from 10.0.2.2 port 59960 ssh2 Jun 03 06:49:01 SailfishSDK systemd-logind[143]: New session c40 of user mersdk. Jun 03 06:49:01 SailfishSDK sshd[11124]: pam_unix(sshd:session): session opened for user mersdk by (uid=0) Jun 03 06:49:04 SailfishSDK sshd[11124]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:49:04 SailfishSDK systemd-logind[143]: Removed session c40. Jun 03 06:49:04 SailfishSDK sshd[11423]: Accepted publickey for mersdk from 10.0.2.2 port 59962 ssh2 Jun 03 06:49:04 SailfishSDK systemd-logind[143]: New session c41 of user mersdk. Jun 03 06:49:04 SailfishSDK sshd[11423]: pam_unix(sshd:session): session opened for user mersdk by (uid=0) Jun 03 06:49:04 SailfishSDK sshd[10705]: dispatch_protocol_error: type 128 seq 5 Jun 03 06:50:07 SailfishSDK sshd[11426]: dispatch_protocol_error: type 128 seq 8 Jun 03 06:50:07 SailfishSDK sshd[10705]: dispatch_protocol_error: type 128 seq 6 Jun 03 06:50:07 SailfishSDK sshd[10705]: Received disconnect from 10.0.2.2: 11: Jun 03 06:50:07 SailfishSDK sshd[11423]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:50:07 SailfishSDK sshd[10702]: pam_unix(sshd:session): session closed for user mersdk Jun 03 06:50:07 SailfishSDK systemd-logind[143]: Removed session c41. Jun 03 06:50:07 SailfishSDK systemd-logind[143]: Removed session c38. //Qt Creator Output 08:49:00: Running steps for project beertent... 08:49:00: Starting: /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-i486/qmake /Users/christopherlamb/SailfishProjects/beertent/beertent/beertent.pro -r -spec linux-g++ CONFIG+=debug CONFIG+=declarative_debug CONFIG+=qml_debug Executing(%build): /bin/sh -e /var/tmp/rpm-tmp.Abqi8B 08:49:01: The process /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-i486/qmake exited normally. 08:49:01: Starting: /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-i486/make Executing(%build): /bin/sh -e /var/tmp/rpm-tmp.vaeuWx mkdir -p translations lupdate /home/mersdk/share/SailfishProjects/beertent/beertent/src /home/mersdk/share/SailfishProjects/beertent/beertent/qml -ts /home/mersdk/share/SailfishProjects/beertent/beertent/translations/beertent.ts Scanning directory '/home/mersdk/share/SailfishProjects/beertent/beertent/src'... Scanning directory '/home/mersdk/share/SailfishProjects/beertent/beertent/qml'... Updating '../beertent/translations/beertent.ts'... Found 7 source text(s) (0 new and 7 already existing) lrelease
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Jarko Still on the train, now with the Jolla Wifi Hotspot active, and I can build without problems. This indicates that I get the problem when no network is connected. In this case I am deploying to the Emulator, so no network should be required Grüsse Chris On 02.06.14 10:23, Jarko Vihriala wrote: journalctl -fa ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Saija In my case restarting the VMs does not help. I think (from memory) that rebooting the whole laptop does help. It is possible that the same reported error may have a number of different causes - in my case it seems to be when my development host loses its network connection (though as with all sporadic problems it is difficult to be sure). Thanks Chris On 03.06.14 09:02, Saija Saarenpää wrote: I get the same error often when developing with Windows 8 64-bit, never on Mac OSX 10. They are both always connected to the (same) network. It helps to close and open the virtual machine, the it stays reachable for another while once it does the same again. - Saija Lähetetty iPadista Christopher Lamb christopher.l...@thurweb.ch kirjoitti 3.6.2014 kello 9.58: Hi Jarko Still on the train, now with the Jolla Wifi Hotspot active, and I can build without problems. This indicates that I get the problem when no network is connected. In this case I am deploying to the Emulator, so no network should be required Grüsse Chris On 02.06.14 10:23, Jarko Vihriala wrote: journalctl -fa ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Juha I have edited my /etc/hosts to look like this: 127.0.0.1 localhost SailfishSDK 255.255.255.255 broadcasthost ::1 localhost fe80::1%lo0 localhost Now in the office, with WIFI turned off, and no ethernet plugged in, I can deploy as RPM successfully! So for the moment it looks like editing the /etc/hosts may be enough. Later this evening on train on the way home I will try with WiFi activated but no proper connection (which means it will probably default to the SBB/Swisscom WiFi which requires a logon to properly connect). Thanks Chris On 03.06.14 11:24, Juha Kallioinen wrote: Hi Christopher, the changing networks may play a part in this problem. Did I already sometime earlier ask you to modify your Mac's /etc/hosts file? It has helped in some cases to add the name for the MerSDK virtual machine to /etc/hosts like this: 127.0.0.1 SailfishSDK Or if you already have the 127.0.0.1 there, just add the SailfishSDK to the end of the existing line like this: 127.0.0.1localhost.localdomain localhost SailfishSDK Could you try this and let me know if there's any improvement? ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Juha I am now on the train. Deployment also works with WiFi active, but not properly connected. This means that - at least in my case - adding SailfishSDK to the /etc/hosts is a good fix for this error. I think this is one for the Known Issues page. I guess that (without the entry in /etc/hosts) active networking is required for SailfishSDK to be resolved. Of course I could SSH in, because I was using localhost as the machine name to connect (ssh -p -i ~/SailfishOS/vmshare/ssh/private_keys/engine/root root@localhost). Thanks for your help and forebearance, Chris On 03.06.14 17:15, Christopher Lamb wrote: Hi Juha I have edited my /etc/hosts to look like this: 127.0.0.1 localhost SailfishSDK 255.255.255.255 broadcasthost ::1 localhost fe80::1%lo0 localhost Now in the office, with WIFI turned off, and no ethernet plugged in, I can deploy as RPM successfully! So for the moment it looks like editing the /etc/hosts may be enough. Later this evening on train on the way home I will try with WiFi activated but no proper connection (which means it will probably default to the SBB/Swisscom WiFi which requires a logon to properly connect). Thanks Chris On 03.06.14 11:24, Juha Kallioinen wrote: Hi Christopher, the changing networks may play a part in this problem. Did I already sometime earlier ask you to modify your Mac's /etc/hosts file? It has helped in some cases to add the name for the MerSDK virtual machine to /etc/hosts like this: 127.0.0.1 SailfishSDK Or if you already have the 127.0.0.1 there, just add the SailfishSDK to the end of the existing line like this: 127.0.0.1localhost.localdomain localhost SailfishSDK Could you try this and let me know if there's any improvement? ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi All I am still suffering from this issue, and getting more and more frustrated! Right now it is refusing to deploy as RPM to the Emulator. Is anything I can do to better debug / understand this issue? The timeout settings previously suggested are probably red-herrings, as they apply to the device (phone / Emulator), whereas the error is a timeout to the SDK. All help greatly appreciated. Grüsse Chris On 28.04.14 08:01, christopher.l...@thurweb.ch wrote: Hi Thanks. I found a well hidden {1} timeout setting in the preferences / devices and increased this from the default 2 secs to 10 secs. This does seem to help, but as always with sporadic problems only time will tell if this is a fix, or if the problem is just dormant for a while. What I don't understand is, given the error says Could not connect to MerSDK Virtual Machine why should a timeout for connecting to a Jolla device play any role? Is the error misleading? is the problem actually in connecting FROM the MerSDK to the Jolla? Chris {1} when the preferences pane opens, at least on my OSX host the timeout setting is offscreen in an embedded pane with scroll bars in the middle of the preferences pane. Only if you scroll down in the embedded pane are further settings including the timeout revealed. Truly a wonderful bit of UI design 8-) Zitat von fasza2mob...@gmail.com: I believe the timeout setting can be adjusted in QtCreator where you set up your mer device(ssh password, etc). By increasing this value you can probably prevent the timeout issue altogether. I don't have the SDK in front of me to give you the exact menu you need to open, but it isn't too hard to find once you know what you're looking for. On Sat Apr 26 2014 13:41:20 GMT+0100 (BST), christopher.l...@thurweb.ch wrote: Hi All I am suffering from sporadic bouts of the error: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server One moment I can be happily hacking and deploying code to my Jolla without error. Then I make a small change to the code, try to deploy, and bing! the error rears its ugly head once again like a scandinavian troll from under a bridge. Project ERROR: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server. 14:02:28: The process /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-armv7hl/deploy exited with code 1. Error while building/deploying project landed26_QT5 (kit: MerSDK-SailfishOS-armv7hl) When executing step 'Rpm' 14:02:28: Elapsed time: 00:27. When the error occurs, there seem to be at least 2 variants: 1) It will successfully deploy As binaries without error, but will give the error on deploy as RPM. 2) Any kind of deploy gives the error. The strange thing is, that despite the error, it does seem to (sometimes) deploy (or possibly partially deploy) code. Right now I am getting the error in constellation 1), so according to the error a deploy as RPM should fail. Yet the test below indicates that it is at least partially deploying code. [root@Jolla javascript]# pwd /usr/share/landed26_QT5/qml/javascript [root@Jolla javascript]# [root@Jolla javascript]# ls -ahl total 52K drwxr-xr-x 1 root root 180 2014-04-26 13:56 . drwxr-xr-x 1 root root 110 2014-04-26 13:56 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 13:55 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 13:55 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 13:55 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 13:55 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 13:55 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-26 13:55 writeDataModel.js //create new file testDeploy.js in QtCreator, then attempt deploy as RPM. Error is generated. Repeat attempt to deploy as RPM. Error is generated again. //but testDeploy.js is now on the Jolla! [root@Jolla javascript]# ls -ahl total 56K drwxr-xr-x 1 root root 206 2014-04-26 14:01 . drwxr-xr-x 1 root root 110 2014-04-26 14:01 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 14:00 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 14:00 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 14:00 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 14:00 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 14:00 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 38 2014-04-26 14:00 testDeploy.js -rw-r--r-- 1 root root 5.2K 2014-04-26 14:00 writeDataModel.js So it looks like that after a second failed deploy, the file is actually deployed! During bouts of the error, QtCreator seems to get out of sync the true status of the SDK VM (indicated by the VirtualBox GUI). So QtCreator may show the green Start SDK status, even though the SDK is already running. Or QtCreator shows the same button grey. So far I have not found any lasting solution to prevent the error in the first place. Sometimes it goes away on its own
Re: [SailfishDevel] was Acceptable Behaviour.. -- Forum
Hi all I am pleasantly surprised at the volume of traffic resulting from a thread that I innocently kicked of that the weekend. Thanks for the active and balanced discussion. I probably won't be able to take an active part in the IRC meeting this afternoon, as it is during the Swiss working day, so here is a summary of some key points from the thread so far. There is a strong wish to avoid fragmentation of channels: That implies that anything new should be instead of / integrated into / a refinement of an existing channel (e.g. TJC) We currently have the following meeting channels of some interest to developers, each with a different focus. 1) Together.Jolla.com (TJC): Q/A site mainly concerned on the phone + bugs and flames. So far no dedciated area for development issues. 2) Talk.Maemo.Org (TMO): A forum, strongly used by app developers and their users. Unlike other channels. this is Jolla independent. 3) This mailing list: used by developers to developers (both inside and outside Jolla). 4) IRC Chat. similar purpose to 3) above. 5) Other non-Sailfish-dedicated developer forums (StackOverflow, Qt Project etc.) We heard from both supporters of mailing lists and forums. The one side will probably never convince the other. I suspect that I am used too… and personally I prefer … are more important as anything else. Both tools have their plus points and weakness. Here are main ones.. Mailing lists (with the appropriate mail client) appeal to those who like the structure of tree views. There is a certain charm in the simplicity of mailing lists (KISS) which do not have all the extraneous functionality / bells and whistles / baggage / bullshit (you choose) that forum or QA sites do. Mailing lists do require a powerful properly configured mail client to be used properly. In a Forum you get that for free in a web client (but can't do much to customise it). Mailing lists aggregate communication in your mail client - avoiding the yet another thing I need to visit .. syndrome. Forums allow editing of previous posts to correct typos / make clarifications. Forums typically have extra functionality: e.g.like, formatting, stickies, the dreaded karma… Privacy: Some forums use avatars, and allow you to suppress your personal email. Forums support sub-forums. At the moment we are talking about splitting the list into development and other, but what happens when we want to split it further? (Jobs, C++, QML, Silica, OpenSource etc.). Cheers Chris On 26.05.14 21:43, fasza2mob...@gmail.com wrote: Hi all, I think Norman is bang on with the fragmentation issue. To address this I propose a possible solution. Why not convert this mailing list to a 'virtual' one; By that I mean move all discussion to a devel section on TJC and have the mailing list be another frontend or interface if you like for the same content(yet better, keep all mails and posts with their metadata in the same database and create an API that both ML and TJC can call thus eases applification too, kind of like MVC). To accommodate this TJC (or at least the proposed devel section) would have to be changed/improved to have a treelike structure alongside its QA nature so that OT answers and its children can easily be tagged and filtered out in both TJC and ML interfaces. OT tag should be available for every poster to tag their answer/comment. To achieve this there should be 2 scripts one that converts each mail to mailing list to a new question/answer/comment retaining the treelike structure and one other sc ript that posts entries from TJC to ML setting the subject correctly as it is now; Perhaps with an x minutes latency to alllow for editing post. If one wants to use a different email address for the mailing list that is set for TJC a setting should be available. Further to this subsections(Qt, Qml, news, politics, ads, jobs etc) could be introduced making it easier to filter or subscribe to selected subsections only; This could benefit both interfaces. Having this approach I believe would give users/developers the flexibility to choose, mix and match the best way(for them) to interact with fellow developers whilst not fragmenting the community. The obvious tradeoff is some developer hours, but doing it in the open could reduce that somewhat especially when it comes to maintenance. Please do comment on what you all think about my proposition, be it positive or negative. Thanks Kris Ps: I'd prefer QA, but mailing list has been more effective in my experience so far. ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide
Hi Stefano Sorry If I distracted the line of argument with the forum vs. ML vs. TJC thing... To directly answer your / David's question: I have been in the ML since pretty early on, and I have never felt oh I wish this list was moderated, get me out of here Indeed I have welcomed the wide range of topics from intensely technical through to Jolla's take on Open Source. Equally I have never felt he / she should be summarily booted off the M (though I can't exclude the possibility that others may think that in my case ...). The past few days have been very different from the norm, and we should not make massive changes based on them alone. However it is possible that as this community increases, and becomes less of a small intimate family, then the need for moderation may grow. Chris Zitat von Stefano Mosconi stefano.mosc...@jolla.com: It seems that the original topic was a bit forgotten in favor of the forum vs. ML vs. TJC (which is a good and intersting topic but no matter what tool we use we will have to go back to the main question). Just to try to steer the boat back on the original journey this the question that David was asking at the beginning: On 22/05/14 17:18, David Greaves wrote: We need to be careful about over-policing discussions but there is such a thing as under-policing too. How would we (community, not Jolla) determine the line? and what measures do we think should be taken? Lorn pointed to this as a useful document: http://www.kde.org/code-of-conduct/ Stefano ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] was Acceptable Behaviour.. -- Forum
Hi Iekku Zitat von Iekku Pylkka iekku.pyl...@jolla.com: Ahoy, We have been pondering the idea about forum. Truth to be told, we are too small company to follow and contribute to that as well. We are following/contributing to several IRC channels, together.Jolla.com, this mailing list and I'm bombing our technical developers with developer-c...@jolla.com questions I can't answer and can't find easily (below 30 minutes). And we have huge amount of work to do :) Yes, we could have trusted community members as a moderators in forum, but we would love to contribute and _read_ those posts coming to forum. Br, Iekku Pylkkä From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Martin Kolman [martin.kol...@gmail.com] Sent: Friday, May 23, 2014 10:39 PM To: devel@lists.sailfishos.org Subject: Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide 23.5.2014 21:21, christopher.l...@thurweb.ch: Hi Jarko before we jump to splitting into multiple SF mailing lists, I wonder has any thought been given to replacing the lists with a developer' forum? I fear that multiple lists would not really help - I would end up subscribed to all of them (both those that I am very interested in, and those that I am sometimes interested in), and thus my In-box would be just as full as it is now. A forum would be divided into sub-forums, and being thread oriented (as opposed to post-oriented) I could freely choose which sub-forums and which threads I chose to dive into, and which to ignore. Well, roughly the same thing can (and usually is by people subscribing to mailing list) achieved by setting up mail filters and redirecting email to per-mailing-list folders. I can't even think about all those email directly ending in my main inbox... Regarding a forum - well, talk.maemo.org is currently basically filling the role of the main Sailfish forum and this seems to work just fine, so I don't really see the need for yet another forum instance (anybody still remembers forum.meego.com ?). Chris Zitat von Jarko Vihriala jarko.vihri...@jolla.com: We're looking into splitting the discussions on SF mailing lists. But, let's not make hasty moves and keep the technological part alive. thanks, Jarko From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Ville M. Vainio [vivai...@gmail.com] Sent: Friday, May 23, 2014 9:45 PM To: Sailfish OS Developers Subject: Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide Some of the discussions here have been borderline illegal; discouraging such behavior even with absence of 'technical' moderator tools (like you have even on barbaric environments like phpbb) is not necessarily a bad idea. On May 22, 2014 5:53 PM, christopher.l...@thurweb.chmailto:christopher.l...@thurweb.ch wrote: Seconded. I could not have put it better myself. Since the early days I have enjoyed the very open nature of this forum where pretty much anything goes that is vaguely Jolla / Qt / Open Source / technical. From my recollection the number of threads that struck me as definitely better elsewhere has been refreshingly small. Many of the threads have been helpful or and or insightful. Keep up the good work Chris Zitat von David Greaves david.grea...@jolla.commailto:david.grea...@jolla.com: There's been enough noise on this mailing list recently that some people have felt they don't want to participate. We need to be careful about over-policing discussions but there is such a thing as under-policing too. How would we (community, not Jolla) determine the line? and what measures do we think should be taken? Lorn pointed to this as a useful document: http://www.kde.org/code-of-conduct/ FWIW I personally don't think there's anything happened recently that I would actually take action over. My delete key works fine and history shows that sometimes cries for help come in strange forms. David/lbt PS This thread is for generic guidelines - please keep any specific issues out of it. ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.orgmailto:devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.orgmailto:devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to
Re: [SailfishDevel] was Acceptable Behaviour.. -- Forum
Ignore my last - I absentmindedly clicked Send before writing the content! My suggestion is that rather than splitting into multiple mailing lists, which does not really cure anything, you replace the mailing lists with a forum. I agree it would be crazy to have both. Martin points out that I could simulate the same with rules and filters: but that is post-fact, so if somebody starts spamming the list from different threads and sock-puppets, then I need to continually add new rules / filters. I prefer to spend my free time programming Sailfish apps, not rules in my mail client! While there is talk.maemo.org, that serves a different role: it is an independent forum (as opposed to one owned by Jolla); and the Sailfish part has not divided into more focused sub-forums yet. Chris Zitat von christopher.l...@thurweb.ch: Hi Iekku Zitat von Iekku Pylkka iekku.pyl...@jolla.com: Ahoy, We have been pondering the idea about forum. Truth to be told, we are too small company to follow and contribute to that as well. We are following/contributing to several IRC channels, together.Jolla.com, this mailing list and I'm bombing our technical developers with developer-c...@jolla.com questions I can't answer and can't find easily (below 30 minutes). And we have huge amount of work to do :) Yes, we could have trusted community members as a moderators in forum, but we would love to contribute and _read_ those posts coming to forum. Br, Iekku Pylkkä From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Martin Kolman [martin.kol...@gmail.com] Sent: Friday, May 23, 2014 10:39 PM To: devel@lists.sailfishos.org Subject: Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide 23.5.2014 21:21, christopher.l...@thurweb.ch: Hi Jarko before we jump to splitting into multiple SF mailing lists, I wonder has any thought been given to replacing the lists with a developer' forum? I fear that multiple lists would not really help - I would end up subscribed to all of them (both those that I am very interested in, and those that I am sometimes interested in), and thus my In-box would be just as full as it is now. A forum would be divided into sub-forums, and being thread oriented (as opposed to post-oriented) I could freely choose which sub-forums and which threads I chose to dive into, and which to ignore. Well, roughly the same thing can (and usually is by people subscribing to mailing list) achieved by setting up mail filters and redirecting email to per-mailing-list folders. I can't even think about all those email directly ending in my main inbox... Regarding a forum - well, talk.maemo.org is currently basically filling the role of the main Sailfish forum and this seems to work just fine, so I don't really see the need for yet another forum instance (anybody still remembers forum.meego.com ?). Chris Zitat von Jarko Vihriala jarko.vihri...@jolla.com: We're looking into splitting the discussions on SF mailing lists. But, let's not make hasty moves and keep the technological part alive. thanks, Jarko From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Ville M. Vainio [vivai...@gmail.com] Sent: Friday, May 23, 2014 9:45 PM To: Sailfish OS Developers Subject: Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide Some of the discussions here have been borderline illegal; discouraging such behavior even with absence of 'technical' moderator tools (like you have even on barbaric environments like phpbb) is not necessarily a bad idea. On May 22, 2014 5:53 PM, christopher.l...@thurweb.chmailto:christopher.l...@thurweb.ch wrote: Seconded. I could not have put it better myself. Since the early days I have enjoyed the very open nature of this forum where pretty much anything goes that is vaguely Jolla / Qt / Open Source / technical. From my recollection the number of threads that struck me as definitely better elsewhere has been refreshingly small. Many of the threads have been helpful or and or insightful. Keep up the good work Chris Zitat von David Greaves david.grea...@jolla.commailto:david.grea...@jolla.com: There's been enough noise on this mailing list recently that some people have felt they don't want to participate. We need to be careful about over-policing discussions but there is such a thing as under-policing too. How would we (community, not Jolla) determine the line? and what measures do we think should be taken? Lorn pointed to this as a useful document: http://www.kde.org/code-of-conduct/ FWIW I personally don't think there's anything happened recently that I would actually take action over. My delete key works fine and history shows that sometimes cries for help come in strange forms. David/lbt PS This thread is for generic guidelines - please keep any specific issues out of it.
Re: [SailfishDevel] was Acceptable Behaviour.. -- Forum
Hi Lauri That is an interesting perspective: my feeling / experience was exactly the opposite: the chief weakness of mailing lists is that they have no usability whatsoever: I get just a flat list of unorganised mails swamping my inbox. That of course may be down to the weaknesses of the web-mail client I use. I would welcome advice on best-practice tools for consuming mailing lists. I am (mostly) on OSX, sometimes on Windows and Linux. To quote Churchill give us the tools and we will do the job. From my point-of-view, plus points for usability of forums over mailing lists are (in no special order): * Stickies * Sub forums * Search * Private Messages * Edit previous posts (I could have used that today when I inadvertently clicked send too early) * Thread based perspective * Does not fill my mail file * Visible metadata such as number of replies, date of most recent reply, number of posts etc. Downsides of forums are: * yet another user/password to manage (although a forum could be integrated into together.jolla.com * yet another site to visit regularly (in addition to my mail client) * karma: so distracting, I prefer the simple egalitarianism / meritocracy of mailing lists Grüsse Chris Zitat von Lauri Nurmi lanu...@iki.fi: 24.5.2014 13:56, christopher.l...@thurweb.ch kirjoitti: My suggestion is that rather than splitting into multiple mailing lists, which does not really cure anything, you replace the mailing lists with a forum. I agree it would be crazy to have both. -1 Is there any forum platform whose usability is not totally terrible compared to mailing lists? First of all, to be able to keep track of who has replied to whose post, and which post, I want to see the posts as a tree. But offering such a tree view seems to be very rare on forums. LN ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Right to silence / General Mailing List
+1 Zitat von Panu Artimo tuntema...@gmail.com: I'm starting to think this whole thread is another instance of the problem that caused the initial problem in the first place. +1 plus for creating a general mailing list for subjects not directly related to software development and technical issues. ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] qt5 ?
The more accurate question is when will SailfishOS be based on Qt 5.2 - or dare I say it, 5.3 8-) ? If I recollect correctly the current Qt level is not quite Qt 5.2 Zitat von mikete...@gmail.com mikete...@gmail.com: Wow thanks for the info, plenty of good news! I'm impressed by how much u r bleeding edge .. kernel 3, qt5 based UI, btrfs, wayland!!! It is a strange feeling to have much newer technology on the smartphone than on the laptop (debian wheezy) and still it's very stable, rock solid! Thanks for your hard work! On Fri, 23 May 2014 15:38:52 +0200 (CEST) sfietkonstan...@free.fr wrote: - Mail original - De: mikete...@gmail.com À: devel@lists.sailfishos.org Envoyé: Vendredi 23 Mai 2014 15:29:34 Objet: [SailfishDevel] qt5 ? Hi sorry if missed this previously on some announcement or mailing list, are there any plans / ETA to move the sailfishos SDK to qt version 5 ? Hello ! Trying to not post with webmail client (not sure of the result) The SDK uses Qt 4 to display Qt Creator, however you are already using Qt 5 to develop. Indeed SailfishOS is based on Qt 5. So technically, SailfishOS SDK is already on Qt 5. Also, do you know if QML on qt 5 is much different from the current one ? Developing with QtQuick in Qt 5 should be quite close to Qt 4. However, you won't be able to develop apps like on Symbian or the N9. The API used on SailfishOS for their UI components is called Silica, and you have the documentation here: https://sailfishos.org/sailfish-silica/ btw I love the phone and I love to develop on it, QML is very relaxing! I have a neo, my primary phone is an n900 from years, had a firefox os (cheap htc) but is currently, sadly, very flawled by hardware and software problems and the jolla, I think, is the natural evolution of the neo and the n900. I loved it from the first moment, I'm very impressed by it, everything worked out of the box, which for a techie power user is not to be taken for granted. Also, the battery duration is from another planet, I don't know how you did it :) True that I don't use it with a sim card but it lasts ages! Btw number 2, sorry .. what about lte support ? :) LTE support is coming with the next update, as stated in this mailing list already: https://lists.sailfishos.org/pipermail/devel/2014-May/004253.html I work with phones, as many of you I had dozens, from the first StarTac ones :) The Jolla phone and Sailfish OS beats, destroys, pulverizes, hands down, everything else I've used and I use now. The usability (gestures) is genius, very quick and natural, and still power user. And it is OPEN .. so fresh and inspiring!! Thanks and have a nice day, pls stop the drama! Have a nice day ! Mike ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide
Hi Jarko before we jump to splitting into multiple SF mailing lists, I wonder has any thought been given to replacing the lists with a developer' forum? I fear that multiple lists would not really help - I would end up subscribed to all of them (both those that I am very interested in, and those that I am sometimes interested in), and thus my In-box would be just as full as it is now. A forum would be divided into sub-forums, and being thread oriented (as opposed to post-oriented) I could freely choose which sub-forums and which threads I chose to dive into, and which to ignore. Chris Zitat von Jarko Vihriala jarko.vihri...@jolla.com: We're looking into splitting the discussions on SF mailing lists. But, let's not make hasty moves and keep the technological part alive. thanks, Jarko From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of Ville M. Vainio [vivai...@gmail.com] Sent: Friday, May 23, 2014 9:45 PM To: Sailfish OS Developers Subject: Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide Some of the discussions here have been borderline illegal; discouraging such behavior even with absence of 'technical' moderator tools (like you have even on barbaric environments like phpbb) is not necessarily a bad idea. On May 22, 2014 5:53 PM, christopher.l...@thurweb.chmailto:christopher.l...@thurweb.ch wrote: Seconded. I could not have put it better myself. Since the early days I have enjoyed the very open nature of this forum where pretty much anything goes that is vaguely Jolla / Qt / Open Source / technical. From my recollection the number of threads that struck me as definitely better elsewhere has been refreshingly small. Many of the threads have been helpful or and or insightful. Keep up the good work Chris Zitat von David Greaves david.grea...@jolla.commailto:david.grea...@jolla.com: There's been enough noise on this mailing list recently that some people have felt they don't want to participate. We need to be careful about over-policing discussions but there is such a thing as under-policing too. How would we (community, not Jolla) determine the line? and what measures do we think should be taken? Lorn pointed to this as a useful document: http://www.kde.org/code-of-conduct/ FWIW I personally don't think there's anything happened recently that I would actually take action over. My delete key works fine and history shows that sometimes cries for help come in strange forms. David/lbt PS This thread is for generic guidelines - please keep any specific issues out of it. ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.orgmailto:devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.orgmailto:devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] Acceptable Behaviour Guidelines - you decide
Seconded. I could not have put it better myself. Since the early days I have enjoyed the very open nature of this forum where pretty much anything goes that is vaguely Jolla / Qt / Open Source / technical. From my recollection the number of threads that struck me as definitely better elsewhere has been refreshingly small. Many of the threads have been helpful or and or insightful. Keep up the good work Chris Zitat von David Greaves david.grea...@jolla.com: There's been enough noise on this mailing list recently that some people have felt they don't want to participate. We need to be careful about over-policing discussions but there is such a thing as under-policing too. How would we (community, not Jolla) determine the line? and what measures do we think should be taken? Lorn pointed to this as a useful document: http://www.kde.org/code-of-conduct/ FWIW I personally don't think there's anything happened recently that I would actually take action over. My delete key works fine and history shows that sometimes cries for help come in strange forms. David/lbt PS This thread is for generic guidelines - please keep any specific issues out of it. ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org ___ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
Re: [SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi Thanks. I found a well hidden {1} timeout setting in the preferences / devices and increased this from the default 2 secs to 10 secs. This does seem to help, but as always with sporadic problems only time will tell if this is a fix, or if the problem is just dormant for a while. What I don't understand is, given the error says Could not connect to MerSDK Virtual Machine why should a timeout for connecting to a Jolla device play any role? Is the error misleading? is the problem actually in connecting FROM the MerSDK to the Jolla? Chris {1} when the preferences pane opens, at least on my OSX host the timeout setting is offscreen in an embedded pane with scroll bars in the middle of the preferences pane. Only if you scroll down in the embedded pane are further settings including the timeout revealed. Truly a wonderful bit of UI design 8-) Zitat von fasza2mob...@gmail.com: I believe the timeout setting can be adjusted in QtCreator where you set up your mer device(ssh password, etc). By increasing this value you can probably prevent the timeout issue altogether. I don't have the SDK in front of me to give you the exact menu you need to open, but it isn't too hard to find once you know what you're looking for. On Sat Apr 26 2014 13:41:20 GMT+0100 (BST), christopher.l...@thurweb.ch wrote: Hi All I am suffering from sporadic bouts of the error: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server One moment I can be happily hacking and deploying code to my Jolla without error. Then I make a small change to the code, try to deploy, and bing! the error rears its ugly head once again like a scandinavian troll from under a bridge. Project ERROR: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server. 14:02:28: The process /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-armv7hl/deploy exited with code 1. Error while building/deploying project landed26_QT5 (kit: MerSDK-SailfishOS-armv7hl) When executing step 'Rpm' 14:02:28: Elapsed time: 00:27. When the error occurs, there seem to be at least 2 variants: 1) It will successfully deploy As binaries without error, but will give the error on deploy as RPM. 2) Any kind of deploy gives the error. The strange thing is, that despite the error, it does seem to (sometimes) deploy (or possibly partially deploy) code. Right now I am getting the error in constellation 1), so according to the error a deploy as RPM should fail. Yet the test below indicates that it is at least partially deploying code. [root@Jolla javascript]# pwd /usr/share/landed26_QT5/qml/javascript [root@Jolla javascript]# [root@Jolla javascript]# ls -ahl total 52K drwxr-xr-x 1 root root 180 2014-04-26 13:56 . drwxr-xr-x 1 root root 110 2014-04-26 13:56 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 13:55 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 13:55 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 13:55 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 13:55 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 13:55 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-26 13:55 writeDataModel.js //create new file testDeploy.js in QtCreator, then attempt deploy as RPM. Error is generated. Repeat attempt to deploy as RPM. Error is generated again. //but testDeploy.js is now on the Jolla! [root@Jolla javascript]# ls -ahl total 56K drwxr-xr-x 1 root root 206 2014-04-26 14:01 . drwxr-xr-x 1 root root 110 2014-04-26 14:01 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 14:00 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 14:00 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 14:00 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 14:00 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 14:00 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 38 2014-04-26 14:00 testDeploy.js -rw-r--r-- 1 root root 5.2K 2014-04-26 14:00 writeDataModel.js So it looks like that after a second failed deploy, the file is actually deployed! During bouts of the error, QtCreator seems to get out of sync the true status of the SDK VM (indicated by the VirtualBox GUI). So QtCreator may show the green Start SDK status, even though the SDK is already running. Or QtCreator shows the same button grey. So far I have not found any lasting solution to prevent the error in the first place. Sometimes it goes away on its own accord, and sometimes various combinations of restarting VirtualBox, the VM, QtCreator or my development host help for a while (but rarely for long). What exactly is the timeout? Is this a property that I can increase? The known issues page of the Alpha (Qt4.8) SDK suggests: https://sailfishos.org/wiki/SDK_Alpha_Known_Issues VBoxManage modifyvm MerSDK --natdnshostresolver1 on But that has not helped. I am running the very latest version of the
Re: [SailfishDevel] Some problems since 1.038+ update.
Hi all There was a thread on this way back in December last year: https://lists.sailfishos.org/pipermail/devel/2013-December/001804.html HtH Chris Zitat von Luca Donaggio donag...@gmail.com: I think you should use QScopedPointers in your main: QScopedPointerQGuiApplication app(SailfishApp::application(argc, argv)); QScopedPointerQQuickView view(SailfishApp::createView()); Covers (and pages as well) can see all your ApplicationWindow (ie: their parent) properties. On Fri, Apr 25, 2014 at 10:14 AM, Gabriel Böhme m.gabrielboe...@googlemail.com wrote: Hi sailors, I have released my app Freiertag in the days Jolla started. It worked pretty well, but now some updates later (since 1.038 and later) I'm facing a problem, I can't find a solution for. Starting the app first time works, but on the second try it crashes. After some inspection it seems the problem is, that the app is don't quit correctly. I can use the gesture in Emulator or press and click the X, in both cases it's still running, after a kill via SSH, it can be started again. The strange thing is, that it worked earlier. Thats the code from main: #include sailfishapp.h#include FreiertagCom.hpp int main(int argc, char *argv[]) { QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); //create a new ListModel for the results FreiListModel *freiModel = new FreiListModel(); //create a new instance of our object that is doing th e most work FreiertagCom *frei = new FreiertagCom(freiModel); view-rootContext()-setContextProperty(frei, frei); view-rootContext()-setContextProperty(freiModel, freiModel); view-setSource(SailfishApp::pathTo(qml/harbour-freiertag.qml)); view-showFullScreen(); return app-exec(); } Another thing is, that I've set my cover and QML pages in this way: import QtQuick 2.0import Sailfish.Silica 1.0 import pages import cover ApplicationWindow { id: mainApplicationWindow MainPage {id: mainPage} StartCover {id: startCover} ResultPage {id: resultPage} initialPage: mainPage cover: startCover } I was doing it, to set text and other informations on the cover(s). But now that seems not to work anymore. So how can I talk to a cover, if I'm setting it via Qt.resolvedUrl for example? Thank you very much! :) Gabriel. ___ SailfishOS.org Devel mailing list -- Luca Donaggio ___ SailfishOS.org Devel mailing list
[SailfishDevel] sporadic bouts of Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server.
Hi All I am suffering from sporadic bouts of the error: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server One moment I can be happily hacking and deploying code to my Jolla without error. Then I make a small change to the code, try to deploy, and bing! the error rears its ugly head once again like a scandinavian troll from under a bridge. Project ERROR: Could not connect to MerSDK Virtual Machine. Timeout waiting for reply from server. 14:02:28: The process /Users/christopherlamb/.config/SailfishAlpha4/mer-sdk-tools/MerSDK/SailfishOS-armv7hl/deploy exited with code 1. Error while building/deploying project landed26_QT5 (kit: MerSDK-SailfishOS-armv7hl) When executing step 'Rpm' 14:02:28: Elapsed time: 00:27. When the error occurs, there seem to be at least 2 variants: 1) It will successfully deploy As binaries without error, but will give the error on deploy as RPM. 2) Any kind of deploy gives the error. The strange thing is, that despite the error, it does seem to (sometimes) deploy (or possibly partially deploy) code. Right now I am getting the error in constellation 1), so according to the error a deploy as RPM should fail. Yet the test below indicates that it is at least partially deploying code. [root@Jolla javascript]# pwd /usr/share/landed26_QT5/qml/javascript [root@Jolla javascript]# [root@Jolla javascript]# ls -ahl total 52K drwxr-xr-x 1 root root 180 2014-04-26 13:56 . drwxr-xr-x 1 root root 110 2014-04-26 13:56 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 13:55 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 13:55 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 13:55 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 13:55 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 13:55 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-26 13:55 writeDataModel.js //create new file testDeploy.js in QtCreator, then attempt deploy as RPM. Error is generated. Repeat attempt to deploy as RPM. Error is generated again. //but testDeploy.js is now on the Jolla! [root@Jolla javascript]# ls -ahl total 56K drwxr-xr-x 1 root root 206 2014-04-26 14:01 . drwxr-xr-x 1 root root 110 2014-04-26 14:01 .. -rw-r--r-- 1 root root 4.2K 2014-04-26 14:00 jsonpath.js -rw-r--r-- 1 root root 1.3K 2014-04-26 14:00 jsonSupport.js -rw-r--r-- 1 root root 349 2014-04-26 14:00 landed.js -rw-r--r-- 1 root root 1.2K 2014-04-26 14:00 message.js -rw-r--r-- 1 root root 8.4K 2014-04-26 14:00 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 38 2014-04-26 14:00 testDeploy.js -rw-r--r-- 1 root root 5.2K 2014-04-26 14:00 writeDataModel.js So it looks like that after a second failed deploy, the file is actually deployed! During bouts of the error, QtCreator seems to get out of sync the true status of the SDK VM (indicated by the VirtualBox GUI). So QtCreator may show the green Start SDK status, even though the SDK is already running. Or QtCreator shows the same button grey. So far I have not found any lasting solution to prevent the error in the first place. Sometimes it goes away on its own accord, and sometimes various combinations of restarting VirtualBox, the VM, QtCreator or my development host help for a while (but rarely for long). What exactly is the timeout? Is this a property that I can increase? The known issues page of the Alpha (Qt4.8) SDK suggests: https://sailfishos.org/wiki/SDK_Alpha_Known_Issues VBoxManage modifyvm MerSDK --natdnshostresolver1 on But that has not helped. I am running the very latest version of the SailfishOS SDK, but had the same error with previous versions. Any ideas? Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] SailfishOS SDK April update available tomorrow. [update:all installers and repositories are online]
Hi Juha and Mikael For what it is worth, to kill some time while a batch of beer is brewing, I installed Kubuntu 14.04 in a VM, and then the April SDK. In my case the Emulator does work in Kubuntu 14.04 - though it does take inordinately long to do so (maybe down to the multiple layers of VMs. The mother host is an Mac Pro 2008 running Mavericks and Professional Version 6.0.3 (1747349). In that I did I fresh install of Kubuntu 14_04 64 bit. Into this I installed VirtualBox 4.3.10 for Linux (package for Ubuntu 13.04 (Raring Ringtail) / 13.10 (Saucy Salamander) AMD), and the April SailfishOS SDK. Having installed everything, the Emulator starts, and I eventually get a GUI, and am able to deploy a demo Hello Sailors app. Cheers Chris Zitat von Juha Kallioinen juha.kallioi...@jolla.com: On 17.04.2014 23:55, Mikael Hermansson wrote: Reinstall from scratch did not solve the issue. I guess kubuntu 14.04 have to be blacklisted for now since emulator don't work :/ Let me know if someone else got similar issues. Btw. the touchpointer is shown. Hi, I installed ubuntu 14.04 (not KDE) to my Mac vmware Fusion and the SDK into that - I'm seeing the same problem here now. What's your host machine and how do you run the SDK in it? I'm just trying to collect more info to figure out what's common in our systems. If I set all the environment variables correctly (from /var/lib/eviconment/compositor/*.conf) and run lipstick from nemo's console it crashes to a SIGILL in the end. Best regards, Juha ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] OSX sdk issue (was maces sdk issue)
Note on OSX, even if you do a clean uninstall reinstall as recommended in the release notes, the old SDK and Emulator VMS are not deleted. However this is not a real problem, the installer installs the new VMs next to the old, and the old VMs can be easily removed using the VirtualBox GUI Cheers Chris Zitat von Juha Kallioinen juha.kallioi...@jolla.com: On 16.04.2014 19:52, Nicola De Filippo wrote: Hi, updating the sdk on MacOS i have 2 issue: 1) when the maintenance tool execute/usr/bin/VBoxManage createvm --name MerSDK --register --basefolder /Users/nicola/SailfishOS/mersdk and i resolved removing MerSDK.vbox and pressing retry in the dialog 2) when the maintenance tool execute /usr/bin/VBoxManage createvm --name SailfishOS Emulator --register --basefolder /Users/nicola/SailfishOS/emulator and i resolved removing Emulator.vbox and pressing retry in the dialog Hi I think this can happen if you chose to update your SDK (instead of a reinstallation) and the MerSDK and SailfishOS Emulator virtual machines were in running state. I need to add more checking to the installer in the update case. In the worst case you will be required to uninstall and download the full installer and reinstall from that. Same goes for Mikael. Best regards, Juha ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] OSX sdk issue (was maces sdk issue)
Hi Juha before you ask, I am fairly sure that the VMs were not running during the uninstall ... I probably would not even have noticed, had I not got into the habit of having the VirtualBox GUI app open all the time (due to sporadic SSH Timeout problems that will be a subject of a later thread ...) I am using an MBA on 10.8.4 Chris Zitat von Juha Kallioinen juha.kallioi...@jolla.com: On 17.04.2014 12:08, christopher.l...@thurweb.ch wrote: Note on OSX, even if you do a clean uninstall reinstall as recommended in the release notes, the old SDK and Emulator VMS are not deleted. However this is not a real problem, the installer installs the new VMs next to the old, and the old VMs can be easily removed using the VirtualBox GUI Hi, I have not seen this happen in any mac where I've tested (three different macs). The old virtual machines definitely should get deleted in the process. The uninstaller does need some work to make it more robust. Best regards, Juha ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] problem with package dependencies
Iosif Have you tried creating a new (throwaway) project that has nothing other than the dependencies in the .yaml / spec files? I have just done this, with only BuildRequires: pkgconfig(sdl2), and the app builds and deploys without error to the emulator. Both SDL2 SDL2-devel packages are installed. Chris Zitat von Iosif Hamlatzis i.hamlat...@gmail.com: Yes I have for all libs (sdl2* and glesv*). And by logging into the engine either as root or mersdk user and checking of installed packages I see an i (installed) left of the libraries' names I also have shutdown the engine and the QtCreator IDE and tried it with same results, even when I rebooted my system. On 17 April 2014 14:12, Kalle Vahlman kalle.vahl...@movial.com wrote: 2014-04-17 14:03 GMT+03:00 Iosif Hamlatzis i.hamlat...@gmail.com: I have un-installed previous SDK and did a clean installation of the latest SDK. I then went inside the engine machine and installed (zypper install ) the packages I need for my game such as SDL2, SDL2_image, SDL2_ttf, SDL2_mixer, GLESv1 GLESv2 Did you install the -devel packages too? Though I think the SDK build should automatically install the deps if they are in the spec... -- Kalle Vahlman, Movial Creative Technologies Inc. Porkkalankatu 20, FI-00180 Helsinki Tel +358 9 8567 6400 Fax +358 9 8567 6401 www.movial.com ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] After April SDK Update Existing App fails on RPM deploy to Jolla Device: Possible rights problem
Hi All I have just installed the April SDK update, and have opened an existing app, and am trying to deploy to my Jolla. The app worked with the previous SDK release, and is already installed on the phone. If I deploy as Deploy by copying Binaries then thinks work as expected, and the app successfully deploys and starts on the phone. If I deploy as Deploy as RPM Package, then I get the error below in QtCreator, and a white screen of death on the phone. [W] unknown:72 - file:///usr/share/landed26_QT5/qml/landed26_QT5.qml:72:5: Type MainPage unavailable MainPage { ^ [W] unknown:9 - file:///usr/share/landed26_QT5/qml/gui/MainPage.qml:9:1: Script file:///usr/share/landed26_QT5/qml/javascript/landed.js unavailable import ../javascript/landed.js as LJS ^ [W] unknown:-1 - file:///usr/share/landed26_QT5/qml/javascript/landed.js: File not found I also get the white screen when trying to open the app. Via SSH on the jolla device I can find the file is present: [nemo@Jolla javascript]$ pwd /usr/share/landed26_QT5/qml/javascript [nemo@Jolla javascript]$ ls -ahl total 48K drwxr-xr-x 1 root root 152 2014-04-16 14:31 . drwxr-xr-x 1 root root 110 2014-04-16 14:31 .. -rw-r--r-- 1 root root 4.2K 2014-04-16 14:31 jsonpath.js -rw--- 1 root root 349 2014-04-16 14:31 landed.js -rw--- 1 root root 1.2K 2014-04-16 14:31 message.js -rwxr-xr-x 1 root root 9.4K 2014-04-16 14:31 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-16 14:31 writeDataModel.js However the access permissions look very strange. I suspect that all the js files should have rights like -rwxr-xr-x Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] After April SDK Update Existing App fails on RPM deploy to Jolla Device: Possible rights problem
Hi Andrey Thanks for the fast answer Given that we are talking about cross-platform development - developing on one and deploying to another I am not sure that is always possible. What if my development host was Windows? And as a developer I want to worry about writing code, not all the arcane rights that all the files need to work when deployed In fact my main development host is OSX which has unix like permissions, the same as Sailfish. Having a look at the files on my host, it looks like the files are now deployed with the original (strange mix of) rights from my host. However in previous SDK releases this was not an issue, As far as I can see the SDK took care of permissions. In the meantime I will chmod the files on the host and see if that helps... Cheers Chris Zitat von Andrey Kozhevnikov coderusin...@gmail.com: IMHO you should manage file permissions yourself before deploying? 16.04.2014 18:39, christopher.l...@thurweb.ch пишет: Hi All I have just installed the April SDK update, and have opened an existing app, and am trying to deploy to my Jolla. The app worked with the previous SDK release, and is already installed on the phone. If I deploy as Deploy by copying Binaries then thinks work as expected, and the app successfully deploys and starts on the phone. If I deploy as Deploy as RPM Package, then I get the error below in QtCreator, and a white screen of death on the phone. [W] unknown:72 - file:///usr/share/landed26_QT5/qml/landed26_QT5.qml:72:5: Type MainPage unavailable MainPage { ^ [W] unknown:9 - file:///usr/share/landed26_QT5/qml/gui/MainPage.qml:9:1: Script file:///usr/share/landed26_QT5/qml/javascript/landed.js unavailable import ../javascript/landed.js as LJS ^ [W] unknown:-1 - file:///usr/share/landed26_QT5/qml/javascript/landed.js: File not found I also get the white screen when trying to open the app. Via SSH on the jolla device I can find the file is present: [nemo@Jolla javascript]$ pwd /usr/share/landed26_QT5/qml/javascript [nemo@Jolla javascript]$ ls -ahl total 48K drwxr-xr-x 1 root root 152 2014-04-16 14:31 . drwxr-xr-x 1 root root 110 2014-04-16 14:31 .. -rw-r--r-- 1 root root 4.2K 2014-04-16 14:31 jsonpath.js -rw--- 1 root root 349 2014-04-16 14:31 landed.js -rw--- 1 root root 1.2K 2014-04-16 14:31 message.js -rwxr-xr-x 1 root root 9.4K 2014-04-16 14:31 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-16 14:31 writeDataModel.js However the access permissions look very strange. I suspect that all the js files should have rights like -rwxr-xr-x Chris ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] After April SDK Update Existing App fails on RPM deploy to Jolla Device: Possible rights problem
Zitat von Juha Kallioinen juha.kallioi...@jolla.com: Hi Juha On my host the access permissions were apple-pip:javascript christopherlamb$ cd /Users/christopherlamb/QTProjects/landed/landed26_QT5/landed26_QT5/qml/javascript apple-pip:javascript christopherlamb$ ls -ahl total 72 drwxr-xr-x 7 christopherlamb staff 238B 22 Feb 22:01 . drwxr-xr-x 9 christopherlamb staff 306B 8 Feb 19:57 .. -rw-r--r-- 1 christopherlamb staff 4,1K 22 Feb 22:01 jsonpath.js -rw--- 1 christopherlamb staff 349B 8 Nov 15:56 landed.js -rw--- 1 christopherlamb staff 1,1K 11 Feb 19:33 message.js -rwxrwxrwx 1 christopherlamb staff 9,4K 22 Feb 19:47 readDataModel.js -rw-r--r-- 1 christopherlamb staff 5,2K 8 Feb 17:29 writeDataModel.js I am not sure why the files on my host have such wildly different permissions. This probably reflects the different ages / sources of the files. Until now the permissions set on the host was not an issue I had to think about. I have now done a chmod 777 *.js That got me a little further, it next complained about missing qml files, so we have a similar issue with the permissions of .qml files. Having gone through all my qml file in the project on my host with chmod 755 *.qml the project now deploys successfully to the device. So it looks like, that (at least on OSX) the permissions of files on the host are important and need to be set correctly as expected on the end device. Something for the known issues page? Grüsse Chris Hi Chris, I've changed the way the shared directories are mounted to the virtual machine. This might be caused by that. Previously the shared folders were mounted with 777 rights and all files were shown as read/write/exec for all inside the virtual machine. I've changed it so that the rights are as they are on your host filesystem. But there's really no reason for those two files to have only root access and the others have more access unless there's some root cause with your source files. There's no code anywhere in the SDK that checks if a file is named 'landed.js' and modifies its access rights :-) What are the source files' permissions in your Mac's filesystem when you list them? Best regards, Juha ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] After April SDK Update Existing App fails on RPM deploy to Jolla Device: Possible rights problem
Hi Mohammed thanks The varied permission on the qml / js files on my host probably reflect that the project originated in pre-sailfish days, and thus the files have come from various sources / versions of Qt Creator / edited with different text editors etc. since then. I have now set all qml and js files in the project on my development host to chmod 644 e.g -rw-r--r-- 1 christopherlamb staff 7,7K 24 Feb 17:20 InitialCharacterPicker.qml Are these the correct permissions? It is how a new demo project sets the permissions. Chris Zitat von Mohammed Hassan mohammed.has...@jolla.com: You don't need executable permission (x) for js files. Read bit is enough so the SDK is correct here. I noticed that the missing scripts do not have read permission set which is the problem. I have no idea how it happened though. Cheers, -- Sent from my Nokia N9On 16.4.2014 15:39 christopher.l...@thurweb.ch wrote: Hi All I have just installed the April SDK update, and have opened an existing app, and am trying to deploy to my Jolla. The app worked with the previous SDK release, and is already installed on the phone. If I deploy as Deploy by copying Binaries then thinks work as expected, and the app successfully deploys and starts on the phone. If I deploy as Deploy as RPM Package, then I get the error below in QtCreator, and a white screen of death on the phone. [W] unknown:72 - file:///usr/share/landed26_QT5/qml/landed26_QT5.qml:72:5: Type MainPage unavailable MainPage { ^ [W] unknown:9 - file:///usr/share/landed26_QT5/qml/gui/MainPage.qml:9:1: Script file:///usr/share/landed26_QT5/qml/javascript/landed.js unavailable import ../javascript/landed.js as LJS ^ [W] unknown:-1 - file:///usr/share/landed26_QT5/qml/javascript/landed.js: File not found I also get the white screen when trying to open the app. Via SSH on the jolla device I can find the file is present: [nemo@Jolla javascript]$ pwd /usr/share/landed26_QT5/qml/javascript [nemo@Jolla javascript]$ ls -ahl total 48K drwxr-xr-x 1 root root 152 2014-04-16 14:31 . drwxr-xr-x 1 root root 110 2014-04-16 14:31 .. -rw-r--r-- 1 root root 4.2K 2014-04-16 14:31 jsonpath.js -rw--- 1 root root 349 2014-04-16 14:31 landed.js -rw--- 1 root root 1.2K 2014-04-16 14:31 message.js -rwxr-xr-x 1 root root 9.4K 2014-04-16 14:31 readDataModel.js -rwxr-xr-x 1 root root 8.8K 2014-02-03 08:32 settingsDB.js -rw-r--r-- 1 root root 5.2K 2014-04-16 14:31 writeDataModel.js However the access permissions look very strange. I suspect that all the js files should have rights like -rwxr-xr-x Chris ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Two prolblems: #1 Telepathy qt5 and #2 OpenSSL
Hi UV Many moons ago I wrote a blog on installing additional packages http://flyingsheeponsailfish.blogspot.ch/2013/11/deploying-additional-packages-to.html My project also uses Telepathy on Sailfish, so this may also help. Here is a link to the yaml file which included Telepathy stuff. https://github.com/sailfishapps/landed/blob/master/landed26_QT5/rpm/landed26_QT5.yaml HtH Chris Zitat von Yuvraaj Kelkar yuvr...@gmail.com: Hello all. I'm trying to port my application from Maemo/Harmattan to Sailfish. Problem 1: One of the dependencies is telepathy-qt. In Maemo/Harmattan this used to be installable/searchable as telepathy-qt4-devel From what I understand, in Sailfish, I'm supposed to use telepathy-qt5. I ssh'd into the mersdk VM and did a sudo zypper install telepathy-qt5 As I had assumed, this doesn't install it into the sb2 environment of course. I added a PKGCONFIG += TelepathyQt5 to my pro file (found out from Google). However that didn't work either. I then entered the sb2 environment and tried to install telepathy-qt5. This too didn't work: No root privileges. What do I have to do to be able to use libtelepathy-qt5 ? Problem 2: Based on more Googling, I figured out that Sailfish does NOT provide its own openssl library and expects users to use the statically compiled version on their own. I have since then gotten the non-heartbleed openssl, compiled it in sailfish sb2 for static compilation and managed to link it against my app. The question is is this really the correct way to use the openssl library? It seems unfriendly. I really hope the answer to #1 isn't compile it yourself and use it. I can do that, but then I'm not sure thats the right way of doing things... Thanks, Uv ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Warning: File `Makefile' has modification time 0.51 s in the future
Hi Thomas Earlier this year in addition to my normal day job I took over responsibility for a server farm of over 50 servers both real and virtual. Their clocks, system and hardware were all over the place. This lead to strange knock on effects, like Samba not being able to authenticate users via SSSD / LDAP. In order to git the clocks under control again I was forced to find out much more about clock-skew {1} and NTP then I cared for. As a result I can be very boring on the subject. 8-) While ntpdate is a valid pragmatic workaround, it remains a workaround, as a properly configured ntp daemon should automatically keep your system clock in sub millisecond sync. Be aware also that NTDATE is deprecated {2}. Instead you should use ntpd -gq However if the time is already too far off the NTD may never be able to catch up (as normally it only makes small jumps {3), or even give up altogether. When I encountered a server with clock(s) way off I used the set of commands below to get it back in line. hwclock --show date service ntpd stop ntpd -gq hwclock --systohc --localtime service ntpd start hwclock --show date After that, if NTPD is properly configured, then NTP should be able to keep the server's system clock in line. HtH Chris {1} the root of all evil {2} http://support.ntp.org/bin/view/Dev/DeprecatingNtpdate {3} by my measurements about 1.7 mins per day Zitat von Thomas Tanghus tho...@tanghus.net: On Monday 14 April 2014 14:49 Chris Walker wrote: I installed the ntp client and it now picks up network time. I'm assuming therefore that there is some time 'slip' between the host machine and VBox. I had the same problem and now run ntpdate from cron.daily. Turns out my PCs clock loses ~5 seconds(sic!) for every 24h :( -- Med venlig hilsen / Best Regards Thomas Tanghus ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Warning: File `Makefile' has modification time 0.51 s in the future
Hi David As we are in danger of taking this thread in a direction that the OP CDW probably did not intend I'll keep it short(ish) 8-) You are right that NTP is very tricky when VMs are involved. VMs need exactly the right NTP config to work properly (e.g. to jump the time quickly after a suspend). The ultimate crime is to run an NTPD Server in a VM. The less said about that the better. For those in the need of some nerdy bedtime reading there is an excellent document on the topic from VMWare: http://www.vmware.com/files/pdf/Timekeeping-In-VirtualMachines.pdf While it specific for VMWare, much of the content is relevant to other virtualisers (e.g. VirtualBox) Cheers Chris Zitat von David Greaves david.grea...@jolla.com: On 15/04/14 07:07, christopher.l...@thurweb.ch wrote: Hi Thomas Earlier this year in addition to my normal day job I took over responsibility for a server farm of over 50 servers both real and virtual. Their clocks, system and hardware were all over the place. This lead to strange knock on effects, like Samba not being able to authenticate users via SSSD / LDAP. In order to git the clocks under control again I was forced to find out much more about clock-skew {1} and NTP then I cared for. As a result I can be very boring on the subject. 8-) While ntpdate is a valid pragmatic workaround, it remains a workaround, as a properly configured ntp daemon should automatically keep your system clock in sub millisecond sync. Be aware also that NTDATE is deprecated {2}. Instead you should use ntpd -gq Personally I think you should use a virtualisation specific solution to time sync if possible - it tends to cope better with virt specific issues like suspend and migrate. eg VMs running on laptops. In this case the VM may not even know that it was suspended since the virt layer does it, not VM pm layer and hence cannot trigger special-case timesync handling. Running ntp on all guests is required in some situations though (eg older xen). Happy to hear any counter-arguments though nb - in farm situations this also means that you simply run ntp on the physical hosts and that's one less config pita on the guests :) However if the time is already too far off the NTD may never be able to catch up (as normally it only makes small jumps {3), or even give up altogether. When I encountered a server with clock(s) way off I used the set of commands below to get it back in line. hwclock --show date service ntpd stop ntpd -gq hwclock --systohc --localtime service ntpd start hwclock --show date After that, if NTPD is properly configured, then NTP should be able to keep the server's system clock in line. HtH Chris {1} the root of all evil {2} http://support.ntp.org/bin/view/Dev/DeprecatingNtpdate {3} by my measurements about 1.7 mins per day Zitat von Thomas Tanghus tho...@tanghus.net: On Monday 14 April 2014 14:49 Chris Walker wrote: I installed the ntp client and it now picks up network time. I'm assuming therefore that there is some time 'slip' between the host machine and VBox. I had the same problem and now run ntpdate from cron.daily. Turns out my PCs clock loses ~5 seconds(sic!) for every 24h :( -- Med venlig hilsen / Best Regards Thomas Tanghus ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Warning: File `Makefile' has modification time 0.51 s in the future
Chris you should clarify what you installed ntp on: your rebuilt workstation or in the sdk? I am assuming the former, but from David's suboptimal answer I assume he assumes the latter ...8-) Cheers Chris Zitat von David Greaves david.grea...@jolla.com: On 14/04/14 14:49, Chris Walker wrote: On Mon, 14 Apr 2014 14:26:58 +0100 Chris Walker cdw_noki...@the-walker-household.co.uk wrote: Can somebody help me with this one please? I've just rebuilt my machine and have reinstalled QtCreator etc. I never had this problem before so I'm guessing it's something I've done but each time I rebuild my project, I get this error. I've googled it and several people suggest that it's because the build system is on another machine. I suppose it is here as the build is a virtual machine, but as I said, it didn't happen until I rebuilt the box (64 bit Mageia 4) and I'm struggling to fix it. hmmm I can't move on with my project until I solve it so can anybody suggest what I could check please? Physician heal thyself! I installed the ntp client and it now picks up network time. I'm assuming therefore that there is some time 'slip' between the host machine and VBox. This works but is suboptimal. Make sure /usr/bin/VBoxService -f is running in the guest; it manages time sync. Also https://www.virtualbox.org/manual/ch09.html Finally check your host is running the correct services and has all the right modules built. David ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] No QSerialPort
Hi Philippe From this and other threads going back to the very early days of this mailing list it is very clear that Wim wants to connect an external GPS unit to his Jolla via cable (usb / serial) for use in the cockpit of a glider. HtH Chris Zitat von Philippe De Swert philippe.desw...@jolla.com: Hi, As this is getting very confused and we are mixing connecting devices to the Jolla and connecting the Jolla to the pc. So which one is it? On 26/03/14 22:11, Wim de Vries wrote: Thank you guys. Indeed devel-su works. I have attached the devices, but the first problem is that the connection does not supply power (the devices stay dead). So you connected some device to your Jolla? As far as I know, a usb connection should always have power, but it is a bit beyond my kowledge. Do I need another plug? This is not true. All depends on context, what you plug into what etc... In general if you plug in a device into a USB host (like your average USB port on your pc/laptop) you should be able to expect power from it. In the case of a phone this is usually a USB device, so it expects power. And in the OTG case it gets complicated. Here the device is able to switch between host and device mode. And will supply power if it is playing host. On other linux/MeeGo systems there was always power... and dmesg would give: [ 31.959468] usb 3-2.1.4: new full-speed USB device number 7 using xhci_hcd [ 31.977423] usb 3-2.1.4: New USB device found, idVendor=10c4, idProduct=ea60 [ 31.977434] usb 3-2.1.4: New USB device strings: Mfr=1, Product=2, SerialNumber=3 [ 31.977440] usb 3-2.1.4: Product: CP2102 USB to UART Bridge Controller [ 31.977445] usb 3-2.1.4: Manufacturer: Silicon Labs [ 31.977450] usb 3-2.1.4: SerialNumber: 0001 [ 31.992330] usbcore: registered new interface driver usbserial [ 31.992343] usbcore: registered new interface driver usbserial_generic [ 31.992355] USB Serial support registered for generic [ 31.992361] usbserial: USB Serial Driver core [ 31.992942] usbcore: registered new interface driver cp210x [ 31.992955] USB Serial support registered for cp210x [ 31.992991] cp210x 3-2.1.4:1.0: cp210x converter detected [ 32.063473] usb 3-2.1.4: reset full-speed USB device number 7 using xhci_hcd [ 32.079934] xhci_hcd :00:14.0: xHCI xhci_drop_endpoint called with disabled ep 8801debc38c0 [ 32.079947] xhci_hcd :00:14.0: xHCI xhci_drop_endpoint called with disabled ep 8801debc3880 [ 32.080450] usb 3-2.1.4: cp210x converter now attached to ttyUSB0 Use Qserialport on ttyUSB0 and ready you are. This looks like plugging in a serial device into a PC. Do you want your Jolla to expose a serial port? That is definitely possible and easy. Do you want your Jolla to read data from some serial port with a serial to USB adapter then you're getting in a very complicated world. Unless you can explain what you actually want there is little we can do to help. Regards, Philippe ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] No QSerialPort
Hi Wim Sailfisch made me smile ... Grüsse Chris Zitat von Wim de Vries wsvr...@xs4all.nl: Thanks Chris and others, Indeed I need to use an external USB GPS and some other USB sensors. All communicate via serial ports. I gave the dmesg output txt as it is on other linux/MeeGo devices to show that I am not talking rocket science (simply plug in and use ttyUSB0 for (qt)serial communication). Never heard about OTG, and never had to use it for my app. Don't know why is there a difference between Sailfisch and MeeGo. My questions are now: Can I enable power on the USB port? (escape would be a battery powered USB hub) Will Sailfish automatically set a up a /dev/ttyUSBsomething when the device (like an external GPS) is plugged in? I will rephrase my question about adding the Qtserialport as a more general question on howto add Qt modules (also in the final package). Thanks. r wim On 03/27/2014 07:06 AM, christopher.l...@thurweb.ch wrote: Hi Philippe From this and other threads going back to the very early days of this mailing list it is very clear that Wim wants to connect an external GPS unit to his Jolla via cable (usb / serial) for use in the cockpit of a glider. HtH Chris Zitat von Philippe De Swert philippe.desw...@jolla.com: Hi, As this is getting very confused and we are mixing connecting devices to the Jolla and connecting the Jolla to the pc. So which one is it? On 26/03/14 22:11, Wim de Vries wrote: Thank you guys. Indeed devel-su works. I have attached the devices, but the first problem is that the connection does not supply power (the devices stay dead). So you connected some device to your Jolla? As far as I know, a usb connection should always have power, but it is a bit beyond my kowledge. Do I need another plug? This is not true. All depends on context, what you plug into what etc... In general if you plug in a device into a USB host (like your average USB port on your pc/laptop) you should be able to expect power from it. In the case of a phone this is usually a USB device, so it expects power. And in the OTG case it gets complicated. Here the device is able to switch between host and device mode. And will supply power if it is playing host. On other linux/MeeGo systems there was always power... and dmesg would give: [ 31.959468] usb 3-2.1.4: new full-speed USB device number 7 using xhci_hcd [ 31.977423] usb 3-2.1.4: New USB device found, idVendor=10c4, idProduct=ea60 [ 31.977434] usb 3-2.1.4: New USB device strings: Mfr=1, Product=2, SerialNumber=3 [ 31.977440] usb 3-2.1.4: Product: CP2102 USB to UART Bridge Controller [ 31.977445] usb 3-2.1.4: Manufacturer: Silicon Labs [ 31.977450] usb 3-2.1.4: SerialNumber: 0001 [ 31.992330] usbcore: registered new interface driver usbserial [ 31.992343] usbcore: registered new interface driver usbserial_generic [ 31.992355] USB Serial support registered for generic [ 31.992361] usbserial: USB Serial Driver core [ 31.992942] usbcore: registered new interface driver cp210x [ 31.992955] USB Serial support registered for cp210x [ 31.992991] cp210x 3-2.1.4:1.0: cp210x converter detected [ 32.063473] usb 3-2.1.4: reset full-speed USB device number 7 using xhci_hcd [ 32.079934] xhci_hcd :00:14.0: xHCI xhci_drop_endpoint called with disabled ep 8801debc38c0 [ 32.079947] xhci_hcd :00:14.0: xHCI xhci_drop_endpoint called with disabled ep 8801debc3880 [ 32.080450] usb 3-2.1.4: cp210x converter now attached to ttyUSB0 Use Qserialport on ttyUSB0 and ready you are. This looks like plugging in a serial device into a PC. Do you want your Jolla to expose a serial port? That is definitely possible and easy. Do you want your Jolla to read data from some serial port with a serial to USB adapter then you're getting in a very complicated world. Unless you can explain what you actually want there is little we can do to help. Regards, Philippe ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] show/hide pulley menu items
It can be very difficult for the beginner to distinguish if a given QML element if pure Qt, or from a component set like Silica. I remember struggling with this in the days of Harmattan. It would be very helpful if the Silica documentation linked the Qt Documentation, and showed inherited properties etc. Yes there is a danger of link rot, and it is not that hard to find the same information via google, but a direct link to the correct version of the Qt Docs would make life a little easier. Zitat von Andrey Kozhevnikov coderusin...@gmail.com: Its QML basics Silica reference cant explain all standard properties and functions, you should learn it yourself. On 11.03.2014 01:43, Vivian Brégier wrote: Thank you, this works just fine. And I always have at least one menuitem visible, so there is no need to hide the menu itself, but it is a good thing to keep in mind. Strangely, https://sailfishos.org/sailfish-silica/qml-sailfishsilica-menuitem-members.html does not reference this attribute. 2014-03-10 13:05 GMT+01:00 dcali...@free.fr mailto:dcali...@free.fr: There is the visible attribute that can be set of each menuitem. In addition, don't forget to put the visible attribute of the pulley menu itself to false if there is no visible menuitem, to avoid empty pulley menu. Damien. À Mon Mar 10 2014 11:29:13 GMT+0100 (CET), Vivian Brégier a écrit : Hi, I am trying to have the pulley menu of my app depend on the app context. I want some menuitem to be hidden when a certain condition is false. I did not find any attribute in MenuItem class to do this. Is it possible to do it in qml ? Maybe using javascript ? Thank you for your help. -- Vivian Brégier http://www.are-ata.org/Vivian.vcf ___ SailfishOS.org Devel mailing list -- Vivian Brégier http://www.are-ata.org/Vivian.vcf ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] How to debug Julla
Hi Tony I initially had great problems upgrading my Jolla, installing stuff from the store, and even activating developer mode. After lots of hunting around, experimentation and measuring, I concluded that the problem was high latency in my home network, and the Jolla's poor tolerance for this. (Other similar devices could upgrade no problem). I stripped my home network down to the bare essentials, the latency issues vanished, and hey presto I could install developer mode, upgrade my Jolla etc. Maybe you have a similar issue? Chris Zitat von Tony tony.chend...@gmail.com: Hi, My Julla ?s system version is 1.0.0.5. I tried to upgrade the system , but system mentioned that is the newest version. Besides , I also tried to start using development mode. The system noticed that development package install failed. Do you know how to debug it ? tony.chend...@gmail.com ___ SailfishOS.org Devel mailing list
[SailfishDevel] QtContacts: phoneNumbers.length attribute gets lost when copied to ListModel
Hi All I have come across some strangeness using QtContacts 5.0 on Sailfish. For various reasons partially lost in the mists of time, I copy contacts (or rather a subset of contract roles) from the ContactModel to a a normal ListModel. One of the roles copied is phoneNumbers. This has a child property length indicating the number of phoneNumbers available for the contact. phoneNumbers.length is available (defined) direct from the ContactModel, but is undefined when I try to use it from the ListModel. So the console.log in the function appendContact() gives a valid result, but the console.log in populate() gives undefined. Other properties, including the phone numbers in phoneNumbers are defined when accessed from the ListModel. Equivalent code works in Haramtatan / Qt 4.7.4 Mobility 1.2 import QtContacts 5.0 ContactModel { id: phoneContactsModelInternal } ListModel { id: localContactModelInternal function populate() { clear(); for (var i = 0; i phoneContactsModelInternal.contacts.length; i ++) { appendContact(phoneContactsModelInternal.contacts[i]); console.log(checking number of phoneNumbers: + localContactModel.get(i).phoneNumbers.length); } } function appendContact(contact) { localContactModel.append({contactId: contact.contactId, displayLabel: contact.displayLabel.label, phoneNumbers: contact.phoneNumbers, phoneNumbersCount: contact.phoneNumbers.length}); console.log(appending: + contact.displayLabel.label + , numbers: + contact.phoneNumbers.length) } } Any ideas? Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] How to debug Julla
Hi Jonni In my case getting rid of the latency in my home network seemed to do the trick. Having got the latency down to normal levels, I was able to upgrade the phone, use the store, activate developer mode etc. without a hitch. Interestingly though, the phone is more than 1.5 mins behind real time. This raises the question, what is supposed to regulate time on a Jolla? Should not reasonably accurate time be available from the GSM signal? PkCon shows that ntpdate is available, but not ntpd (and given that the phone will not always have an active udp network, ntpd would probably not be much help anyway). grüsse Chris Zitat von Jonni Rainisto jonni.raini...@jolla.com: Few things to check: - date and clock is current. https connection will fail if clock is even a minute behind real time. - you haven't touched SSU button in developer mode, if you have then you need to switch your domain back to sales. If nothing else is helps then setting time and doing factory reset might do the trick as last resort. re, Jonni From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of christopher.l...@thurweb.ch [christopher.l...@thurweb.ch] Sent: Thursday, February 27, 2014 11:48 AM To: Sailfish OS Developers; Tony Cc: devel@lists.sailfishos.org Subject: Re: [SailfishDevel] How to debug Julla Hi Tony I initially had great problems upgrading my Jolla, installing stuff from the store, and even activating developer mode. After lots of hunting around, experimentation and measuring, I concluded that the problem was high latency in my home network, and the Jolla's poor tolerance for this. (Other similar devices could upgrade no problem). I stripped my home network down to the bare essentials, the latency issues vanished, and hey presto I could install developer mode, upgrade my Jolla etc. Maybe you have a similar issue? Chris Zitat von Tony tony.chend...@gmail.com: Hi, My Julla ?s system version is 1.0.0.5. I tried to upgrade the system , but system mentioned that is the newest version. Besides , I also tried to start using development mode. The system noticed that development package install failed. Do you know how to debug it ? tony.chend...@gmail.com ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] How to debug Julla
Hi Jonni I have just found an option Automatic update under Settings/Time and Date. Selecting that has jumped the time to be within seconds of real time. Chris Zitat von Jonni Rainisto jonni.raini...@jolla.com: Yes, usually you get accurate time from your GSM network depending on your operator. Sometimes users do start the phone without SIM attached, and in that case people can forget 1970 or some other wrong year in the device. In future updates there are improvements to automaticly fetch time also from ntp servers (don't remember if that was already included in 1.0.3.8). re, Jonni From: christopher.l...@thurweb.ch [christopher.l...@thurweb.ch] Sent: Thursday, February 27, 2014 2:26 PM To: Sailfish OS Developers; Jonni Rainisto Subject: Re: [SailfishDevel] How to debug Julla Hi Jonni In my case getting rid of the latency in my home network seemed to do the trick. Having got the latency down to normal levels, I was able to upgrade the phone, use the store, activate developer mode etc. without a hitch. Interestingly though, the phone is more than 1.5 mins behind real time. This raises the question, what is supposed to regulate time on a Jolla? Should not reasonably accurate time be available from the GSM signal? PkCon shows that ntpdate is available, but not ntpd (and given that the phone will not always have an active udp network, ntpd would probably not be much help anyway). grüsse Chris Zitat von Jonni Rainisto jonni.raini...@jolla.com: Few things to check: - date and clock is current. https connection will fail if clock is even a minute behind real time. - you haven't touched SSU button in developer mode, if you have then you need to switch your domain back to sales. If nothing else is helps then setting time and doing factory reset might do the trick as last resort. re, Jonni From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of christopher.l...@thurweb.ch [christopher.l...@thurweb.ch] Sent: Thursday, February 27, 2014 11:48 AM To: Sailfish OS Developers; Tony Cc: devel@lists.sailfishos.org Subject: Re: [SailfishDevel] How to debug Julla Hi Tony I initially had great problems upgrading my Jolla, installing stuff from the store, and even activating developer mode. After lots of hunting around, experimentation and measuring, I concluded that the problem was high latency in my home network, and the Jolla's poor tolerance for this. (Other similar devices could upgrade no problem). I stripped my home network down to the bare essentials, the latency issues vanished, and hey presto I could install developer mode, upgrade my Jolla etc. Maybe you have a similar issue? Chris Zitat von Tony tony.chend...@gmail.com: Hi, My Julla ?s system version is 1.0.0.5. I tried to upgrade the system , but system mentioned that is the newest version. Besides , I also tried to start using development mode. The system noticed that development package install failed. Do you know how to debug it ? tony.chend...@gmail.com ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QML C++ Integration not working in Sailfish
Antonio Can you provide a code example, and the exact error(s) that you get. Chris Zitat von antonio.cano.go...@ovi.com: Yes, it helped thanks. My problem now is different. I have some private attributes in my C++ object. When I call the first tiem I can modify the attributes, but in the second call the attributes dessapear. Is there a way to solve this problem? (The attributes a new C++ Object not declared as QObject) El Jueves 27 de febrero de 2014 12:32, Stefan Brand stefan.br...@seiichiro0185.org escribió: Hi, On 25.02.2014 15:06, antonio.cano.go...@ovi.com wrote: An I try to connect this with my sailfish application with the code: QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); Bloomfilters * bloomfilters = new Bloomfilters(); view-rootContext()-setContextProperty(Bloomfilters, bloomfilters); view-setSource(SailfishApp::pathTo(qml/bloomfilter.qml)); view-showFullScreen(); return app-exec(); But this is not working. Someone have any idea about what am I doing wrong? You can use qmlRegisterType to pass the Info of the C++ class to QML like this in the cpp: #include bloomfilters.h int main(int argc, char *argv[]) { // Get App and QML-View objects QScopedPointerQGuiApplication app(SailfishApp::application(argc, argv)); QScopedPointerQQuickView view(SailfishApp::createView()); // Register Bloomfilter Class qmlRegisterTypeBloomfilter, 1(harbour.myapp.Bloomfilter, 1, 0, Bloomfilter); // Prepare the QML view-setSource(SailfishApp::pathTo(qml/bloomfilter.qml)); view-show(); // Run the app return app-exec(); } Afterwards you can create a Bloomfilter-Object in QML: Bloomfilter { id: bloom } and use the methods with bloom.insertElement(myelement) I hope this helps. Regards Stefan Brand ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Starting point for building packages
Sali Tobias The Sailfish SDK / Qt Creator does much of the hard work for you. Dependencies on other libraries are specified in the .yaml file. In the archives of this forum you will find many posts on this issue. I suggest you start with a default Sailfish project, then gradually move in code from you ownCloud app. mfg Chris Zitat von Tobias Brunner tob...@tobru.ch: Hi, I'm new to the SailfishOS SDK. I'd like to build the ownCloud client for SailfishOS and create a RPM package for easy installation at the Jolla. The documentation on how to build the ownCloud sync client is here: http://doc.owncloud.org/desktop/1.5/building.html But now I'm looking for a hint on how to build software for SailfisOS. The SDK is up and running, and now? The howto's are all about how to create a new software, but not about how to build and package an existing one. I know how to build software on linux: ./configure; make; make install. (And resolve needed dependencies). How can this three commands be mapped to the SailfishOS world? Thanks for any hints on getting started. Cheers, Tobias ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QtContacts: phoneNumbers.length attribute gets lost when copied to ListModel
Hoi Matthew Thanks. I have now raised this bug report: https://bugreports.qt-project.org/browse/QTBUG-37139 Funnily, at least one of reasons lost in the mists of time is a previous bug report I raised at the start of the year: https://bugreports.qt-project.org/browse/QTBUG-35276 Grüsse Chris Zitat von Matthew Vogt matthew.v...@jolla.com: Hi Chris. This sounds like a regression in the (unfinished) qtpim module of Qt 5, which reimplements the functions provided by QtContacts in QtMobility. You can report bugs against this module here: https://bugreports.qt-project.org/browse/QTBUG/component/19726 Thanks, Matt From: devel-boun...@lists.sailfishos.org [devel-boun...@lists.sailfishos.org] on behalf of christopher.l...@thurweb.ch [christopher.l...@thurweb.ch] Sent: Thursday, February 27, 2014 8:09 PM To: Sailfish OS Developers Subject: [SailfishDevel] QtContacts: phoneNumbers.length attribute gets lostwhen copied to ListModel Hi All I have come across some strangeness using QtContacts 5.0 on Sailfish. For various reasons partially lost in the mists of time, I copy contacts (or rather a subset of contract roles) from the ContactModel to a a normal ListModel. One of the roles copied is phoneNumbers. This has a child property length indicating the number of phoneNumbers available for the contact. phoneNumbers.length is available (defined) direct from the ContactModel, but is undefined when I try to use it from the ListModel. So the console.log in the function appendContact() gives a valid result, but the console.log in populate() gives undefined. Other properties, including the phone numbers in phoneNumbers are defined when accessed from the ListModel. Equivalent code works in Haramtatan / Qt 4.7.4 Mobility 1.2 import QtContacts 5.0 ContactModel { id: phoneContactsModelInternal } ListModel { id: localContactModelInternal function populate() { clear(); for (var i = 0; i phoneContactsModelInternal.contacts.length; i ++) { appendContact(phoneContactsModelInternal.contacts[i]); console.log(checking number of phoneNumbers: + localContactModel.get(i).phoneNumbers.length); } } function appendContact(contact) { localContactModel.append({contactId: contact.contactId, displayLabel: contact.displayLabel.label, phoneNumbers: contact.phoneNumbers, phoneNumbersCount: contact.phoneNumbers.length}); console.log(appending: + contact.displayLabel.label + , numbers: + contact.phoneNumbers.length) } } Any ideas? Chris ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Starting point for building packages
Jolla Harbour currently has very strict rules would be even more correct ;) Zitat von Thomas Tanghus tho...@tanghus.net: On Friday 28 February 2014 00:20 Artem Marchenko wrote: Also sailfish has quite strict rules on what can be used from where and what can be deployed where Jolla Harbour has very strict rules would be more correct ;) -- Med venlig hilsen / Best Regards Thomas Tanghus ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QML C++ Integration not working in Sailfish
Hi Antonio This works for me: //Start BloomfiltersDemo.cpp #ifdef QT_QML_DEBUG #include QtQuick #endif #include sailfishapp.h #include bloomfilters.h int main(int argc, char *argv[]) { QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); Bloomfilters * bloomfilters = new Bloomfilters(); view-rootContext()-setContextProperty(bloomfiltersdata, bloomfilters); view-setSource(SailfishApp::pathTo(qml/BloomFiltersDemo.qml)); view-showFullScreen(); return app-exec(); } //End BloomfiltersDemo.cpp //start bloomfilters.h #ifndef BLOOMFILTERS_H #define BLOOMFILTERS_H #include QObject #include QDebug //#include svn/bloom-read-only/bloom_filter.hpp class Bloomfilters : public QObject { Q_OBJECT public: Q_INVOKABLE void generateFilterInvokable2() const ; explicit Bloomfilters(QObject *parent = 0); void generateFilter() ; private: QString _k; }; #endif // BLOOMFILTERS_H //end bloomfilters.h //start bloomfilters.cpp #include bloomfilters.h Bloomfilters::Bloomfilters(QObject *parent) : QObject(parent) { } void Bloomfilters::generateFilterInvokable2() const { qDebug() Blooming Magic; } //end bloomfilters.cpp //start FirstPage.qml import QtQuick 2.0 import Sailfish.Silica 1.0 Page { id: page Component.onCompleted: { bloomfiltersdata.generateFilterInvokable2() } } //end FirstPage.qm HtH Chris Zitat von antonio.cano.go...@ovi.com: Hi Andrey, The output is: ReferenceError: bloomfilterdata is not defined I am calling it from a .js Some idea about how can I make that this is also defined there. Thanks! El Miércoles 26 de febrero de 2014 12:18, Andrey Kozhevnikov coderusin...@gmail.com escribió: start app in terminal and show output please On 26.02.2014 17:15, antonio.cano.go...@ovi.com wrote: Hi Thomas, I did both modification but the code does not work yet. Here is the new code: QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); Bloomfilters * bloomfilters = new Bloomfilters(); view-rootContext()-setContextProperty(bloomfiltersdata, bloomfilters); view-setSource(SailfishApp::pathTo(qml/bloomfilter.qml)); view-showFullScreen(); return app-exec(); El Miércoles 26 de febrero de 2014 12:01, Thomas Perl th.p...@gmail.com escribió: On 2014-02-26 11:48, antonio.cano.go...@ovi.com wrote: I tried this: QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); Bloomfilters * bloomfilters = new Bloomfilters(); view-setSource(SailfishApp::pathTo(qml/bloomfilter.qml)); view-showFullScreen(); view-rootContext()-setContextProperty(Bloomfilters, bloomfilters); return app-exec(); But it is not working yet. Two things: 1. Call setContextProperty() right before before setSource(), so it's already available when the QML content is loaded. 2. Use lowercase for context property names (bloomfilters instead of Bloomfilters). In some places, QML interprets IDs starting with an upper case letter as a type name, and that won't work then, therefore stick to lowercase. HTH :) Thomas ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QML C++ Integration not working in Sailfish
Hi Antonio Here is a simplified example from one of my projects, showing a number of ways to expose C++ to QML. I tend to use qmlRegister when exposing C++ objects with methods, and SetContextProperty when I want to expose simple stuff like global constants only. I do things that way because it works for me ... Chris #ifdef QT_QML_DEBUG #include QtQuick #endif #include QGuiApplication #include QQuickView #include QQmlContext #include sailfishapp.h //#include landedtorch.h //harmattan #include gsttorch.h //sailfish #include satinfosource.h #include operatingsystem.h #include windowingsystem.h #include telepathyhelper.h #include landedtheme.h #include jsonstorage.h //HELPER Functions getting system information which will be exported to QML bool isSimulator() { bool simulator; #if defined(QT_SIMULATOR) simulator = true; #else // real device or emulator simulator = false; #endif return simulator; } int getOperatingSystem() { //C++ representation of the OperatingSystem enumg const OperatingSystem Os; int OSId; #if defined(Q_OS_SYMBIAN) OSId = Os.Symbian; #elif defined(Q_OS_MAC64) OSId = Os.Mac64; #elif defined(Q_OS_UNIX) OSId = Os.Unix; #elif defined(Q_OS_WIN32) OSId = Os.Win32; #elif defined(Q_OS_WIN64) OSId = Os.Win64; #elif defined(Q_OS_WINCE) OSId = Os.WinCE; #elif defined(Q_OS_SIMULATOR) OSId = Os.Simulator; #else // other OS OSId = Os.Other; #endif return OSId; } //END HELPER Functions static QObject *theme_singletontype_provider(QQmlEngine *engine, QJSEngine *scriptEngine) { Q_UNUSED(engine) Q_UNUSED(scriptEngine) LandedTheme *landedTheme = new LandedTheme(); return landedTheme; } static QObject *jsonStorage_singletontype_provider(QQmlEngine *engine, QJSEngine *scriptEngine) { Q_UNUSED(engine) Q_UNUSED(scriptEngine) JSONStorage *jsonStorage = new JSONStorage(); return jsonStorage; } int main(int argc, char *argv[]) { bool simulator = isSimulator(); //expose an enum of operating systems types to QML qmlRegisterUncreatableTypeOperatingSystem(OperatingSystem, 1, 0, OperatingSystem, ); int OperatingSystemId = getOperatingSystem(); qmlRegisterTypeGstTorch(GstTorch,1,0,GstTorch); qmlRegisterTypeTelepathyHelper(TelepathyHelper,1,0,TelepathyHelper); qmlRegisterTypeSatInfoSource(SatInfoSource,1,0,SatInfoSource); qmlRegisterSingletonTypeLandedTheme(LandedTheme, 1, 0, LandedTheme, theme_singletontype_provider); qmlRegisterSingletonTypeJSONStorage(JSONStorage,1,0,JSONStorage, jsonStorage_singletontype_provider); QScopedPointerQGuiApplication app(SailfishApp::application(argc, argv)); QScopedPointerQQuickView view(SailfishApp::createView()); view-rootContext()-setContextProperty(OperatingSystemId, OperatingSystemId); view-rootContext()-setContextProperty(simulator, simulator); view-setSource(SailfishApp::pathTo(qml/landed26_QT5.qml)); view-show(); view-showFullScreen(); return app-exec(); } Zitat von Andrey Kozhevnikov coderusin...@gmail.com: ??? contextProperty should be set BEFORE loading QML source. On 25.02.2014 21:33, martin.gri...@gmail.com wrote: Hi, you should set the context property after loading the QML source. Then it ought to be available from QML. Martin Am Tue Feb 25 2014 15:06:39 GMT+0100 (CET) schrieb antonio.cano.go...@ovi.com: Hi, I am trying to do a simple Sailfish aplication that uses a c++ library. I try to connect my qml code with the C++ code. First I create QObject extended library: #ifndef BLOOMFILTERS_H #define BLOOMFILTERS_H #include QObject #include svn/bloom-read-only/bloom_filter.hpp class Bloomfilters : public QObject { Q_OBJECT public: Q_INVOKABLE void generateFilterInvokable2() const ; Q_INVOKABLE void generateFilterInvokable(const QString cadenaK, const QString cadenaM) const ; explicit Bloomfilters(QObject *parent = 0); void generateFilter() ; Q_INVOKABLE void insertElement(const QString cadena) ; Q_INVOKABLE bool checkElement(const QString cadena) ; signals: public slots: private: bloom_parameters _parameters ; bloom_filter _filter; bool _filter_charged; QString _k; }; #endif // BLOOMFILTERS_H An I try to connect this with my sailfish application with the code: QGuiApplication *app = SailfishApp::application(argc, argv); QQuickView *view = SailfishApp::createView(); Bloomfilters * bloomfilters = new Bloomfilters(); view-rootContext()-setContextProperty(Bloomfilters, bloomfilters); view-setSource(SailfishApp::pathTo(qml/bloomfilter.qml)); view-showFullScreen(); return app-exec(); But this is not working. Someone have any idea about what am I doing wrong? Best Regards ___
Re: [SailfishDevel] how to get qml debug output to file
Hi This may be related: I haven noticed that using journalctl to view system logs on the device, QML and C++ debug output is reported differently. Below is an example extract: C++ debug is reported with the application name. QML debug is reported as lipstick Feb 23 13:39:39 localhost lipstick[971]: about to getPrimaryContact: area_id: 3, template_id: 1, any: 1 Feb 23 13:39:39 localhost landed26_QT5[27558]: [D] JSONStorage::openDatabase:20 - Document Path: /home/nemo/Documents/ Feb 23 13:39:39 localhost landed26_QT5[27558]: [D] JSONStorage::openDatabase:22 - App Name: landed26_QT5 Feb 23 13:39:39 localhost landed26_QT5[27558]: [D] JSONStorage::openDatabase:24 - Full path: /home/nemo/Documents/landed26_QT5/landeddb.json Feb 23 13:39:39 localhost lipstick[971]: about to pass DB to JSONPath Chris Zitat von Artem Marchenko artem.marche...@gmail.com: I found message handler working just fine (both for QML console and cpp qDebug() messages) is emulator builds, but not for when code is running on device. Go figure. Could it be so that on device exactly QML engine is somehow always used separately from your c++ execution context.. no it doesn't make sense. It would be nice to have some logs-to-file way working in production with both QML and c++ logs indeed. Cheers, Artem. On Wed, Feb 5, 2014 at 4:58 PM, Robin Burchell robin.burch...@jolla.comwrote: On 04 Feb 2014, at 22:37, Tero Siironen tero.siiro...@iki.fi wrote: Andrey Kozhevnikov coderusin...@gmail.com kirjoitti 4.2.2014 kello 23.14: This is messages handler i'm using in my projects: This doesn't seem to make a difference for me, the log file still contains only c++ side debug prints, qml prints (like console.log()) are not handled with messagehandler. Actually I found out that even if set in pro-file: DEFINES +=QT_NO_DEBUG_OUTPUT DEFINES +=QT_NO_WARNING_OUTPUT I still get qml debug prints printed out to console, so it seems that those prints from qml are not handled via normal debug handling at all? I would like to get no debug printing at all, or then just to file. DEFINES in qmake adds additional defines (-D) to the cflags used to build C++ affect code compiled with them. QtDeclarative was not compiled with these defines, so your adding them won't affect console.log (whose C++ implementation lives inside QtDeclarative). If you don't want debug prints, you need to install a message handler (you say you've tried this and it doesn't work, I can't answer why that would be the case, it should work, as it uses the same infrastructure). BR, Robin ___ SailfishOS.org Devel mailing list -- Artem Marchenko http://agilesoftwaredevelopment.com http://twitter.com/AgileArtem ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Hidden information to ListView
Jukka It is very easy to add multiple roles / fields with the same append command. Here is an example culled from one of my apps: var db = DB.DataModel(); var rs = db.getContacts(area_id, template_id); for(var i = 0; i rs.rows.length; i++) { contactModel.append({name: rs.rows.item(i).name, phone: rs.rows.item(i).phone, primary_contact: rs.rows.item(i).primary_contact, contact_id: rs.rows.item(i).id}); } HTH Chris Zitat von Jukka Heikkilä jut...@gmail.com: I'm creating page which shows some information which is returned from local storage. I have successfully created the next code which will work: Page { root Component.onCompleted: { //External Function which return the items from Local Storage items = DB.listItems(); //Adding items to list view for(var i = 0; i items.rows.length; i++){ listItems.model.append({itemValue: items.rows.item(i).itemValue}) } } SilicaListView { id: listItems model: listModel VerticalScrollDecorator {} delegate: ListItem { id: contentListItem Label { text: itemValue } } } ListModel { id: listModel } } I would like to include some hidden values (e.g. DB row ID) for later usage, but is there possibilities to include extra data with same append command to the ListItem? I know the properties, but I haven't managed how to pass the variables in it. I would be grateful for your help. Kind Regards, Jukka @Juukks ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Sideload Native App To Sailfish
No rooting (or jailbreaking) pr verboten-hacks required. Just put the phone into developer mode. In theory any user can do this with a few clicks. Then you can install anything that will run. However this route does imply that the user has some idea of what they are doing, just a a user installing on a Linux desktop will need some idea as well. It might not be a route for a stereotypical grandma. Apologies in a advance to all the non-stereotypical-Linux-savvy-Jolla-wielding-grandmas who are part of this mailing list. Grüsse Chris Zitat von Network Nut sillyst...@gmail.com: Hi All, I have what I imagine to be a very common problem: 1. There will be billions of people who own smartphones. 2. I have a 100% native Linux C++ app that I would like a few of those billions of people to use. These are my future customers. 3. I do not necessarily want to use an app store of any kind, if I choose not to use any. 4. I would like for my customers to decide, at their own discretion, whether to side-load my native app onto their smartphone by going to my web site, and not an app store. 5. I would like to avoid having my customers call my tech-support line and listen on the phone for 30 minutes as one of my tech-support representatives tells him/her how to root their phone so that they can side-load my app. In other words, I would like the same situation that exists now under the desktop model, where anyone who owns a desktop computer has full discretion of what they do with their computer, without (significant) restrictions from the OS vendor. I understand that Jolla allows 100% true native C++ apps, but I was unable to determine, with a quick search on the WWW, whether Jolla allows 100% native C++ apps under the acquisition model above. Can anyone clarify? Is it true that the owner of a Jolla smartphone will be able to determine for himself/herself whether to side-load a third-party native application without jumping through hoops to bypass restrictions created by the OS? Regards, -Nut ___ SailfishOS.org Devel mailing list
[SailfishDevel] Recommended way to populate a LocalStorageDB from an external source
Hi All An architectural question: What is the the recommended approach (and harbour friendly way) to populate a LocalStorageDB on the Jolla with data from an external source? Some background to the question: My app Landed depends on a LocalStorageDB. It was always clear to me that Landed would be a readonly client of the DB, and the business of populating the DB would fall to something else. Until recently that something else was a companion app LandedSettings. But I was never happy with that solution, and it now seems to me that a Qt Desktop app would be much better suited to configuring / populating the DB. (more screen, better keyboard etc.). This raises the question: How best to transfer the data populated in the Desktop app to the DB on the Jolla? The Jolla's Documents folder (or a sub directory thereof) could clearly be used as as staging area. This is the approach that Wim takes with his Harbour-approved Checklists app. But exactly how should this work? I see several possibilities, all with pros and cons. a) The Desktop app creates a LocalStorageDB, and the entire DB is transferred to the Jolla. This means that Landed would have to deploy the DB on a file level, then access it via the LocalStorage api as normal. b) The Desktop app dumps a human readable exchange file (csv, xml, json) which is picked up by Landed, parsed and converted to SQL INSERTS then called via the LocalStorage API c) The Desktop app dumps an SQL script (a list of INSERT INTO statements) which Landed then executes d) The Desktop app dumps an SQL script in the form of a javascript file, which is then dynamically loaded by Landed using dynamic QML e) something I have not thought of ... As further background, I tend to favour QML over C++, but use the latter where necessary (and am beginning to feel uncomfortably comfortable in that language ...). I am aware that for any kind of file manipulation from QML I will probably have to resort to a C++ plugin. (Even this http://www.developer.nokia.com/community/wiki/Reading_and_writing_files_in_QML example from Nokia which sets out to provide file access without a plugin describes an implementation that is to my mind a c++ plugin!) At the moment I favour option b) as the most likely to work and be accepted, with option d) as the wildcard probably wont't work, but might turn out to be the easy option. Thanks in advance for your thoughts, Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QtContacts and Calendar access doesn't work after update.
Matthias I connect via ssh as user nemo by calling the following from a terminal on my development workstation: ssh -p 22 -i ~/.ssh/jolla_rsa nemo@192.168.2.17 See the link below for more details how to connect: http://flyingsheeponsailfish.blogspot.ch/2014/01/connecting-to-jolla-with-rsa-key-file.html Once connected as user nemo, I change to user root by: devel-su The password is the one shown on the Jolla under Settings / Developer Mode / Developer Tools. Once connected as root execute the following commands: cd /usr/share/mapplauncherd cp privileges privileges.old echo /usr/bin/landed26_QT5,p privileges replace landed26_QT5 with the name of your app. The p permission is for people Grüsse Chris Zitat von Matthias Barmeier barme...@barmeier.com: Could you please give me a link hon how to connect to the phone with permissions that will allow me to edit this file. Thanks. Am 01.02.2014 13:02, schrieb Bernd Wachter: Andrey Kozhevnikov coderusin...@gmail.com writes: same here for QtContacts: contactsdatabase.cpp:1028 static QSqlDatabase ContactsDatabase::open(const QString): Unable to create contacts database directory: /home/nemo/.local/share/system//privileged/Contacts/qtcontacts-sqlite Add your application to /usr/share/mapplauncherd/privileges (this will be overwritten on system updates, and only suitable for use on your phone, not for harbour) ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QGeoSatelliteInfoSource in Sailfish
Hi Aaron Adding that information proved easier than I had expected: it just took a few minutes during breakfast! void SatInfoSource::onSatsInViewUpdated(const QListQGeoSatelliteInfo list) { int newInView = list.count(); if (newInView != _satsInView) { qDebug() satInfoSource.cpp: onSatsinViewUpdated: QString::number(newInView, 'g', 2); for (int i = 0; i newInView; i++ ) { qDebug() satsInView: id: list[i].satelliteIdentifier() , system: list[i].satelliteSystem() , strength: + list[i].signalStrength(); } _satsInView = newInView; emit satellitesInViewChanged(newInView); } } The for loop is new, and gives the following output. [D] SatInfoSource::onSatsInViewUpdated:59 - satInfoSource.cpp: onSatsinViewUpdated: 23 [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 1 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 3 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 6 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 11 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 14 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 17 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 19 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 20 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 22 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 27 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 28 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 32 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 74 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 66 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 82 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 73 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 80 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 65 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 88 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 87 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 81 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 67 , system: 0 , strength: [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 72 , system: 0 , strength: Interestingly neither system nor strength give expected output. Sometimes I get garbled output like: [D] SatInfoSource::onSatsInViewUpdated:59 - satInfoSource.cpp: onSatsinViewUpdated: 21 [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 1 , system: 0 [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 3 , system: 0 Updates requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 6 , system: 0 ates requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 11 , system: 0 Stop Updates requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 14 , system: 0 Updates requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 19 , system: 0 [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 22 , system: 0 requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 27 , system: 0 ates requested, but no satellite info source available [D] SatInfoSource::onSatsInViewUpdated:61 - satsInView: id: 28 , system: 0 Stop Updates requested, but no satellite info source available Grüsse Chris Zitat von christopher.l...@thurweb.ch: Aaron I will have to write some more code to spit out the sat ids to the debug console. At the moment my SatInfoSource declarative plugin is ultra-simple, it just exposes the numbers of Sats InView and InUse, nothing more. At the moment my hypothesis is based on the circumstantial evidence that new newer devices (Jolla and Galaxy) show roughly double the number of older GPS only devices (N9, Garmin). Cheers Chris Zitat von Aaron McCarthy aaron.mccar...@jolla.com: Hi, On Sat, 1 Feb 2014 10:37:59 christopher.l...@thurweb.ch wrote: A slight correction to my previous post I think that the high number of Sats in View shown on the Jolla means that it is showing both GPS and Glonass Satellites,
Re: [SailfishDevel] QGeoSatelliteInfoSource in Sailfish
Hi all A slight correction to my previous post I think that the high number of Sats in View shown on the Jolla means that it is showing both GPS and Glonass Satellites, and thus gives a figure almost double of that I was used to from the Nokia N9 and classic GPS devices. Comparing various devices I get the following readings: Sats in use / Sats in View / Device / Technology 6 / 9 / Nokia N9 / GPS only 5 / 20 / Jolla / GPS and Glonass 15 / 21 / Samusung Galaxy Note 3 / GPS and Glonass 9 / Garmin GPSmap 60CSx / GPS only So the Sats in View figure of the Jolla is plausible. Interestingly the Jolla's Sats In Use figure - at least in the tests on my terrace this morning - is consistently lower than that displayed by the other devices. I show 15 for the Galaxy above, but have seen readings of 18 and 19. Chris Zitat von christopher.l...@thurweb.ch: But SatsInView jumps straight to 17 and stays at 17. This figure seems less plausible to me. ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] USB developer mode icon moved on 1.0.3.8
And it is compounded by the fact that the icons are displayed for a very short length of time. They should remain displayed until the user prods one, or unplugs the USB again. mfg Chris Zitat von Tone Kastlunger users.giulie...@gmail.com: +1 to this, this was quite a puzzle for me as well On Fri, Jan 31, 2014 at 10:57 PM, christopher.l...@thurweb.ch wrote: Hi all After 1.0.3.8 I could no longer connect via USB in developer mode. I was in the middle of writing one of those one-step-forwards-two-steps-back emails, when I had another go, and realised that the icons shown on the jolla screen when plugging in the USB have moved. Previously the developer mode icon was in the middle. Now it is on the left. I had got so used the position, that I was just prodding the icon in the middle without reading the label. If I prod the correct icon, developer mode works as before. But why the move? Chris ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QGeoSatelliteInfoSource in Sailfish
Hi Marcel I can confirm that I am getting QGeoSatelliteInfoSource info too. SatsInUse is changing, before levelling out at 4, which is plausible as I am sitting inside. But SatsInView jumps straight to 17 and stays at 17. This figure seems less plausible to me. It maybe down to a bug in my code, but if so I have yet to find it. (Tomorrow I will test outside, to see if that makes a difference). mfg Chris Zitat von Marcel mar...@aliquis.de: Hi, just a small note if everyone else is waiting for this: QGeoSatelliteInfoSource is working since Sailfish 1.0.3.8 The only problem here is what the system (GPS/GLONASS) is not set, it is QGeoSatelliteInfo::Undefined. But this shouldn't be a real problem to nearly all apps. Greetings, Marcel Am Sonntag, 12. Januar 2014 schrieb Marcel mar...@aliquis.de: Hi, I'm trying to get satellite information using QGeoSatelliteInfoSource, but I cannot create a default source. Is it not implemented in Sailfish? I know that Sailfish is using Geoclue, and there is a commit in Qt for supporting the Geoclue Satellite interface: https://qt.gitorious.org/qt/qtlocation/commit/e09d297894a8b4ff9b06317169b5a9eb97aa7162 Is this missing in the Sailfish-Qt-packages? Is the satellite interface missing in the Sailfish-Geoclue? Is it another problem? Can I access the satellite information in another way? Another small partly related problem I found then using the GPS is the following: If you use QGeoPositionInfoSource the VerticalAccuracy and the HorizontalAccuracy are always the same, is this normal behaviour? Greetings, Marcel ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] ListView with SectionDelegate (Strange Behaviour)
Hi Gabriel Thank you for taking the time to reproduce this issue. It is reassuring to know that based on the admittedly slim evidence of this issue alone I am not going completely mad! I agree that the 2 issues sound related (I will have a poke around with your code later today). Based on the assumption that this is probably a Qt issue rather than a Sailfish issue I was about to open an issue on the Qt Bug tracker. Then standing on the platform waiting for my train to arrive I thought: How difficult could it be to fire up my Qt5.2 installation and create a Desktop version of this project to reproduce this issue? By the time the train pulled in I had my first ever Qt Desktop project running! I had expected to see the same sections bug, but unexpectedly I got the expected behaviour! i.e on Qt 5.2.0 desktop I cannot reproduce this bug. My next thought was: Maybe this is still a Qt problem, but has been fixed in Qt 5.2.0? Luckily I have an old Qt 5.0.0 installation lurking on my laptop. So I added a Qt version and kit for that, and launched the project as Qt5.0.0. Again no Section header bug. Then I turned to Android with Qt 5.2.0, and once again got expected behaviour (no bug). So to cut a long story short, this behaviour appears to be limited to Sailfish (or the particular Qt version used by Sailfish) I can reproduce it on the Emulator and real Jolla Device. I cannot reproduce it on Harmattan Qt 4.7.4 (from which I am porting the code) I cannot reproduce it on Qt Desktop (5.0.0 and 5.2.0) I cannot reproduce it on Android Qt 5.2.0 (Emulator and Galaxy Note 3 with KitKat). Grüsse Chris Zitat von Gabriel Böhme m.gabrielboe...@googlemail.com: Hi Chris, I can reproduce this behavior with the code you posted and it seems to be a bug. I have encountered some days ago also a problem with section handling. If you are appending a new Item to the list, while you use FullString section headers, it's creating a new section, even if it is already there. The expected behavior should be, that the new item is sorted under the right section. Especially in my intended use case, it's a problem because I want to fetch some data from an online source and sort it in different sections. But how to do that, without appending? Insert is also not possible at all, because it's varying always. Even if the model is set after, appending (again) - it's not working. Also a code example added at the bottom (just modified your code). The new person/item with name Tom should be displayed in Boss section, but the ListView creates another Boss section at the end of the list. Maybe both are relating to each other? Thank you very much. Gabriel. import QtQuick 2.0 import Sailfish.Silica 1.0 Page { id: page SilicaListView { id: contactList anchors.fill: parent model: contactModel delegate: Label { text: model.name width: page.width } section.property: section section.criteria: ViewSection.FullString section.delegate: Rectangle { height: 50 width: page.width color: Theme.secondaryHighlightColor opacity: 0.5 Label { anchors.horizontalCenter: parent.horizontalCenter text: section } } PullDownMenu { MenuItem { text: Add a person onClicked: { // comment out to set the model again, but thats not working at all. :-/ // contactList.model = emptyList contactModel.insert({name: Tom, section: Boss}) // contactList.model = contactModel } } } } ListModel {id: emptyList} ListModel { id: contactModel ListElement { name: Achim section: Boss } ListElement { name: Ana section: Friends } ListElement { name: Bezelbub section: Friends } ListElement { name: Billy section: Friends } ListElement { name: Boris section: Friends } ListElement { name: Carrie section: Friends } ListElement { name: Cassie section: Friends } ListElement { name: Eziekiel section: Family } ListElement { name: Gargantua section: Family } ListElement { name: Gilbert section: Family } ListElement { name: Giles section: Family } ListElement { name: Gog section: Workgroup } ListElement { name: Mabel section: Workgroup }
Re: [SailfishDevel] ListView with SectionDelegate (Strange Behaviour)
Sali Norbert thanks for your confirmation. I have not filed a bug report with the Qt-Project, as my tests on Desktop, Android etc indicate that this is not a general Qt 5.x bug, but a Sailfish specific one. (Despite my expectations, I got expected behaviour on those platforms). My understanding is that this is currently the best place to report Sailfish development bugs. If there is a more appropriate place, then I will gladly report it there. Chris Zitat von Norbert Wenzel norbert.wenzel.li...@gmail.com: On 01/29/2014 07:56 PM, christopher.l...@thurweb.ch wrote: This time it is a ListView / SilicaListVew with a SectionDelegate based on FirstCharacter. For the items initially displayed this works well, with the section headers exactly where I would expect them[...] But as I flick up, I then a section header between each item (e.g between Gargantua and Gilbert). I can also confirm that behaviour. In your example and also in an app I'm coding myself. Did you already file a bug or are you still investigating? Norbert ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] App own C++ plugin equivalent to Silica Theme
Hi All I would like to create a C++ plugin for my app to export common UX settings as constants to the QML side (Things like font sizes, margin sizes that reoccur throughout the app.) In essence I want something very similar to the functionality offered by by the Silica Theme. After some experimentation I have hit on a solution that both compiles and works. This is pasted at the bottom of this mail. I have 2 questions: 1) Is there a more elegant solution to achieve the same goal? 2) From the C++ side, can I access the Silica them UX constants? e.g to set some of my constants based on Silica constants? Grüsse Chris //start landedtheme.h #ifndef LANDEDTHEME_H #define LANDEDTHEME_H #include QObject class LandedTheme : public QObject { Q_OBJECT Q_PROPERTY(int MarginSmall READ MarginSmall CONSTANT) //many more similar Q_PROPERTY declarations here public: LandedTheme(QObject* parent = 0) : QObject(parent) {} int MarginSmall() const { return 10;} //many more similar variables here }; #endif // LANDEDTHEME_H //end landedtheme.h //in the main .cpp file #include landedtheme.h static QObject *theme_singletontype_provider(QQmlEngine *engine, QJSEngine *scriptEngine) { Q_UNUSED(engine) Q_UNUSED(scriptEngine) LandedTheme *landedTheme = new LandedTheme(); return landedTheme; } //in the main function qmlRegisterSingletonTypeLandedTheme(LandedTheme, 1, 0, LandedTheme, theme_singletontype_provider); ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] SailfishOS SDK January update 2 available today
Hi Juha On the first attempt I got this error: Operation Failed finished: Updating SDK Engine - exited with status 4 History: - File './noarch/sdk-harbour-rpmvalidator-1.3-10.6.1.jolla.noarch.rpm' not found on medium 'http://releases.sailfishos.org/sdk/latest/hotfixes/i486/' - Can't provide ./noarch/sdk-harbour-rpmvalidator-1.3-10.6.1.jolla.noarch.rpm Problem occured during or after installation or removal of packages: Installation aborted by user Please see the above error message for a hint. On the second try a few seconds later, the update worked properly. Chris Zitat von Juha Kallioinen juha.kallioi...@jolla.com: Hello, another minor update to the SDK Build Engine was released today. The update concerns again only the SDK Control Center, which gets some performance improvements, layout changes and provides a Check for updates now button in the Updates page. You should see updates available for the Build Engine in the SDK Control Center Updates tab after your build engine has been running about 10 minutes. Please do not visit the Updates tab before the 10 minutes has passed, otherwise the updates will show up later. The 'Check for updates now' button was put there to avoid this wait period in the future. As always, if you have any questions, you can send email to this mailing list mailto:devel@lists.sailfishos.org or write it up on together.jolla.com and tag your issue with 'sdk'. Happy hacking, Jolla SDK Team. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] White screen and no errors
Hi Marcin How did you deploy? Have you checked the QtCreator compile output as well as the application output? Grüsse Chris Zitat von Marcin M. marmistrz...@gmail.com: You mean unclosed brackets? With the MeeGo components such thing always was being reported in the log... -- Marcin 2014-01-28 Gabriel Böhme m.gabrielboe...@googlemail.com Hi, but an error in the QML file - *should* throw an error message on the console?! Gabriel. Am Dienstag, den 28.01.2014, 15:25 +0100 schrieb Luciano Montanaro: On Tue, Jan 28, 2014 at 2:51 PM, Marcin M. marmistrz...@gmail.com wrote: Why can it happen? Something similar has happened to me when a QML file cannot be parsed. Check those curly brackets! :) Luciano ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] Interacting with the Sailfish Software Input Panel from code
Hi all Is there any way to interact with the Jolla Software Input Panel (Visual Keyboard9 from Code? i.e. Are any signals emitted when it opens / closes? Are there any methods to open and close it? Harmattan had things like panelOpen and closeSoftwareInputPanel(); So far in Sailfish all I have found is that from a QML TextArea element onClicked and onFocusChanged are emitted, giving an indirect indication that the keyboard has opened. mfg Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Interacting with the Sailfish Software Input Panel from code
Hi Pekka thanks, that's it! Now using Qt.inputMethod.show() I have found this documented as a 4.8 to 5.0 change. http://qt.apidoc.info/5.2.0/qtdoc/qtquick-porting-qt5.html I think it would be helpful if the Silica documentation of the TextArea component also referred to the SoftwareInputPanel and Qt.inputMethod Thanks Chris Zitat von Pekka Vuorela pekka.vuor...@jolla.com: On Tue, 2014-01-28 at 17:14 +0100, christopher.l...@thurweb.ch wrote: Hi all Is there any way to interact with the Jolla Software Input Panel (Visual Keyboard9 from Code? i.e. Are any signals emitted when it opens / closes? Are there any methods to open and close it? Harmattan had things like panelOpen and closeSoftwareInputPanel(); So far in Sailfish all I have found is that from a QML TextArea element onClicked and onFocusChanged are emitted, giving an indirect indication that the keyboard has opened. QInputMethod class, QGuiApplication::inputMethod(), and in QML Qt.inputMethod.show()/hide() etc. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] Binding weirdness on Jolla device after deploy with screen locked
Hi all I have come across some weird behaviour on the Jolla: Bindings don't seem to work immediately after an app has deployed when the screen is locked. To demonstrate this I have created the code snippet at the end of this mail. To reproduce, create a default Sailfish project, paste the code into FirstPage.qml. My code has: 1) an Item dummy with an enabled property bound to applicationActive. 2) a BackgroundItem with an enabled property bound to dummy.enabled 3) a Label in the BackgroundItem with with text property bound to dummy.enabled When the app is active, the label should display Correct Label However, if the app is deployed when the screen is locked, on unlocking and activating the app the label text shows Oops Wrong label even though console.log output suggests that the text has changed to Correct Label. Also, when the backgroundItem is clicked, the highlight colour is not shown (but an onClicked event is generated). The app behaves as expected (label correctly displayed, highlight colour shown) when: a) The app is pushed to the left, and then reactivated, b) The app is deployed while the screen is unlocked (immediately shown fullscreen), c) The app is opened from the app icon direct on the Jolla, d) The app is deployed to the Emulator. Grüsse Chris //Start FirstPage.qml import QtQuick 2.0 import Sailfish.Silica 1.0 Page { id: page Item { //represents a much more complex object (e.g. c++ plugin) //reduced to item here for demo purposes id: dummy enabled: applicationActive onEnabledChanged: console.log(QML dummy onEnabledChanged: + enabled); } BackgroundItem { enabled: dummy.enabled onEnabledChanged: console.log(QML BackgroundItem onEnabledChanged: + enabled); width: parent.width height: 160 anchors.horizontalCenter: parent.horizontalCenter anchors.verticalCenter: parent.verticalCenter Label { anchors.horizontalCenter: parent.horizontalCenter anchors.verticalCenter: parent.verticalCenter text: dummy.enabled ? Correct Label : Oops! Wrong label onTextChanged: console.log(QML text onTextChanged: + text); color: Theme.primaryColor } onClicked:console.log(BackgroundItem clicked) } } //End FirstPage.qml ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Python and Qt Widgets development possible?
Hi Dietmar Zitat von Dietmar Schwertberger maill...@schwertberger.de: Switching to QML/Qt Quick seems a major effort and especially, it is very un-pythonic to use this mixture of two languages. There is only one language here, QML, which is an extension of javascript. Qt Quick is a set of components used by QML. This component set is further extended by the Sailfish Silica component set. I did not find QML itself that hard to learn, and there are lots of examples and tutorials on the interweb. The mix of 2 languages comes when need to do some heavy-lifting and have to fall back to C++. Having said that C++ and QML are well integrated. Zitat von Dietmar Schwertberger maill...@schwertberger.de: ...there's quite some effort to stay compatible over multiple platforms You have hit the nail on the head. Multiplatform coding is not easy, especially if you want to keep native interaction, and look and feel. We know that Sailfish is quite a bit different from any other platform. Multi-platform is not impossible, but it does involve compromises. I have been working for some while to port a Harmattan app to Sailfish. For the most part I have got it working, but not without some obstacles. Off the top of my head (and without going into detail), here is a list of differences I found between these 2 platforms 1) Qt Version 4.7.4 -- 5.x 2) QtQuick version 1.0 -- 2.0 3) Mobility 1.2 -- Qt 5.0 equivalents 4) Project Template Files C++ main function, Pro file 5) Project structure 6) Sailfish look and feel, user interaction 7) Sailfish Specials: e.g Cover 8) Missing QML Components / Extra Components, 9) Component Property differences 10) Library differences inc. Bugs 11) Niggly Technical Stuff 11.1) How a javascript file imports a second javascript file 11.2) Location of LocalStorage DB 12) Harbour rules (somethings are technically possible, but not allowed to be harbour compliant) 13) etc Grüsse Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Binding weirdness on Jolla device after deploy with screen locked
Hi Ove That's pretty much the behaviour I get, and the swipe to multitasking and back fixes things. The behaviour 1s 100% reproducible when the screen is locked during deploy, and never occurs when deploying to an unlocked screen. That the problem to do with bindings is only a wild assumption on my part, it could be down to the screen not updating as you suggest. Grüsse Chris Zitat von Ove Kåven o...@arcticnet.no: Den 26. jan. 2014 13:42, skrev christopher.l...@thurweb.ch: Hi all I have come across some weird behaviour on the Jolla: Bindings don't seem to work immediately after an app has deployed when the screen is locked. Are you able to tell the difference between something like this, and the more straightforward issue of the app display simply not updating? I frequently experience the latter immediately after a deploy. If I click anything, or try to use the pulley menu, I can hear the sounds and get the console logs, but the display doesn't start updating before I swipe out to the multitasking view and then reactivate it. Everything works otherwise. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Qt Creator Code AutoCompletion in QML for C++ declarative plugins
Hoi Sven thanks. The problem (and it is a small one only) is limited to the auto-completion. Code builds perfectly well enough. QtCreator's clean function wipes out the moc files, but seems to have no effect on auto-complete. grüsse Chris Zitat von Sven Putze sailfish...@hardcodes.de: Hi, try to build/run qmake from the menu and if that does not help, throw away the Makefile. The property binding depends on the Qt meta system, bound to the Q_OBJECT macro. Often recompiling does not start the needed steps to recreate the moc files and things get strange. BR. Sven The trouble is, I often refactor the C++ code as I add complexity e.g. renaming properties and methods as better names occur to me, or reflect a change in purpose. The trouble is, back on the QML side, QtCreator code-completion still suggests the original names long after these have been renamed or eliminated. Is there any way to flush / update the code completion so it reflects the current state of affairs? Cleaning does not help, nor does Reset Code Model. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Qt Creator Code AutoCompletion in QML for C++ declarative plugins
Hi David Thanks, your guess is correct. I had not seen the option under Tools / C++. Unfortunately it does not help either. During breakfast I have found a solution: closing all projects and QtCreator, then reopening gives autocompletion with the correct updated properties / method names. Chris Zitat von David Greaves david.grea...@jolla.com: On 26/01/14 16:23, christopher.l...@thurweb.ch wrote: Hi All My app exposes chunks of functionality to the QML side from C++ declarative plugins. On the QML side QtCreator gives me automatic code completion for elements exposed from such plugins, suggesting property and method names. So far so good. The trouble is, I often refactor the C++ code as I add complexity e.g. renaming properties and methods as better names occur to me, or reflect a change in purpose. The trouble is, back on the QML side, QtCreator code-completion still suggests the original names long after these have been renamed or eliminated. Is there any way to flush / update the code completion so it reflects the current state of affairs? Cleaning does not help, nor does Reset Code Model. Just guessing but this ^^ option is under Tools = QML/JS ... have you tried Tools = C++ = Update Code Model David ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] LED Flash functionality on a real Jolla
Mohammed et al I have just posted the code of my demo Flashlight app to Github https://github.com/sailfishapps/ThrowawayDemos/tree/master/Working/GSTTorch2 This uses Gstreamer, and in its current form is far from Harbour compliancy. It exposes a torch to QML using the Jolla LED flash that can be turned on and off, and changed from Beam mode to Flash mode. So far the main focus is on the back-end c++ part, as I require this functionality as part of my main app. As and when time permits I will enrich this demo including: * use of resource policy engine to arbitrate access to camera * expose flash rate to QML as property so it can be changed from GUI (e.g. with a slider control). Have fun Chris Zitat von Mohammed Hassan mohammed.has...@jollamobile.com: I am glad it worked for you :-) Perhaps I should start working on an open flash light app. Would anyone like to contribute to such app? I have already 2 apps to maintain and I can't easily fit a 3rd. Cheers, On Fri, Jan 24, 2014 at 09:43:54PM +0100, christopher.l...@thurweb.ch wrote: Hi Mohammed That's it! It's strange how things go. Yesterday was one of those days: At work I spent hours fighting with a recalcitrant Confluence installation that refused to start properly, and the evening I could not get the GST code working properly. Today was very different. I discovered that that the mysql database under the Confluence was corrupt. Once i found that, it more or less fixed itself. Then this evening I see from your mail that my GST code only needs one character changed to work! Zitat von Mohammed Hassan mohammed.has...@jolla.com: g_object_set (G_OBJECT(src), mode, 2, NULL); That should do it for you. Works fine here after changing mode to 2 instead of 1 It really does work, both in developer mode from Qt Creator, and once installed from the Jolla itself. Thank you! I will integrate this functionality into my core app this weekend. Zitat von Mohammed Hassan mohammed.has...@jolla.com: Those errors are harmless. Just ignore them. They decrease with each update as we implement more camera features ;) Looking forward to the update(s)! 8) chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] LED Flash functionality on a real Jolla
Hi Mohammed That's it! It's strange how things go. Yesterday was one of those days: At work I spent hours fighting with a recalcitrant Confluence installation that refused to start properly, and the evening I could not get the GST code working properly. Today was very different. I discovered that that the mysql database under the Confluence was corrupt. Once i found that, it more or less fixed itself. Then this evening I see from your mail that my GST code only needs one character changed to work! Zitat von Mohammed Hassan mohammed.has...@jolla.com: g_object_set (G_OBJECT(src), mode, 2, NULL); That should do it for you. Works fine here after changing mode to 2 instead of 1 It really does work, both in developer mode from Qt Creator, and once installed from the Jolla itself. Thank you! I will integrate this functionality into my core app this weekend. Zitat von Mohammed Hassan mohammed.has...@jolla.com: Those errors are harmless. Just ignore them. They decrease with each update as we implement more camera features ;) Looking forward to the update(s)! 8) chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] LED Flash functionality on a real Jolla
Hi Mohammed thanks. I will try that later (am at work at the moment). I will also need to look into the resource policy engine, as that is new to me. Harbour entry is becoming less and less of a concern for me. My app does too many things that are not considered kosher under the current Harbour rules (sending SMS direct, use of QtContacts, LED access, other Qt Libraries etc.) As an aside, on Harmattan while I got the the Gstreamer route working in developer mode, I never got it working when launching the app direct from the device (and thus chose the QCamera approach). This was probably down to AEGIS, but that should not be a concern on the Jolla. Chris Zitat von Mohammed Hassan mohammed.has...@jollamobile.com: On Wed, Jan 22, 2014 at 08:29:19PM +0100, christopher.l...@thurweb.ch wrote: [...] Solution 3) Via a Gstreamer QML Plugin. I dug out some old code that worked on Harmattan, basically the code from this link to create a qml plugin http://developer.nokia.com/Community/Wiki/How_to_turn_your_camera_flash_into_a_torch_on_Harmattan_using_GStreamer On Sailfish this compiles, but gives the following error when I start the torch. (GSTTorch:708): GStreamer-CRITICAL **: gst_element_set_state: assertion `GST_IS_ELEMENT (element)' failed This is not the recommended way but it's the only way to get flash light to work but I have to warn you that it wil prevent your app from entering harbour. You will also have to make use of resource policy engine to arbitrate access to camera. Failing to do so will lead to breaking jolla camera app. Simply replace subdevsrc with droidcamsrc and use droideglsink as the sink and it should work fine. Cheers, ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Cannot ssh to device in USB developer mode
Hi Damien congratulations on getting in! There is not such thing as a stupid question, only stupid answers 8-) Now that you can connect things by SSH, if you want to do so using an RSA key in the same way that you can connect to the emulator, here is how: http://flyingsheeponsailfish.blogspot.ch/2014/01/connecting-to-jolla-with-rsa-key-file.html Chris Zitat von dcali...@free.fr: Hello, Selon Graham Cobb g+jo...@cobb.uk.net: On 23/01/14 14:01, dcali...@free.fr wrote: It sounds like you are connecting to the ssh on your own machine. To test that you could try logging in to an account which exists on your own machine (or check the hw key, or set up a banner message). Thanks, I've understand my mistake. It's working now. Thanks a lot and sorry for the stupid question#8230; Damien. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] LED Flash functionality on a real Jolla
Hi All As many will know from previous threads, I am well on the way to porting an app for paraglider pilots. A subsidiary yet still important feature {1} that I added to the Harmattan version is an emergency flash, using the phone's LED. Now that I have got the app's main functionality ported and up and running on a real Jolla device, I am now trying to get this feature working too. Unfortunately everything I try fails. But as there is a working flashlight app in the Harbour, it must be possible by hook or crock. So far i have tried 3 solutions: Solution 1) A C++ plugin exposing QtMultimedia QCamera to QML This is the solution that works in the Harmattan version of the app, and was therefore the first solution I tried to port to the Sailfish version. When I start the flash I get the following error: QCameraPrivate::_q_error:171 - Camera error: The camera service is missing Solution 2) QtMultimedia 5.0 Torch QML element As QtMultimedia offers a Torch Element this seemed an easier and cleaner approach. import QtMultimedia 5.0 Page { id: page Torch { id: torch power: 75 // 75% of full power enabled: false // off onEnabledChanged: console.log(onEnabledChanged: + enabled) } However this gives the following errors on app start up. (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'scene-mode' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'noise-reduction' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'image-preview-supported-caps' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'image-capture-supported-caps' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'flicker-mode' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'exposure' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'ev-compensation' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'colour-tone-mode' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'capabilities' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'autofocus' from interface 'GstPhotography' (QMLTorch:2300): GLib-GObject-CRITICAL **: Object class GstDroidCamSrc doesn't implement property 'aperture' from interface 'GstPhotography' QObject::connect: No such signal CameraBinExposure::valueChanged(int) Solution 3) Via a Gstreamer QML Plugin. I dug out some old code that worked on Harmattan, basically the code from this link to create a qml plugin http://developer.nokia.com/Community/Wiki/How_to_turn_your_camera_flash_into_a_torch_on_Harmattan_using_GStreamer On Sailfish this compiles, but gives the following error when I start the torch. (GSTTorch:708): GStreamer-CRITICAL **: gst_element_set_state: assertion `GST_IS_ELEMENT (element)' failed Thanks Chris {1} at about this time last year I discovered that a GPS position is enough to get a helicopter close to you, but on a dark and snowy night you need a bright light to get it exactly over you. ___ SailfishOS.org Devel mailing list
[SailfishDevel] PKGCONFIG entry in project .pro file
Hi All When / why is a PKGCONFIG entry required in a Sailfish .pro file? Is there a better way of specifying the same (e.g. via a .yaml entry)? As an example, I am playing around with GStreamer. If I have the following entry n my .pro file then the project builds and deploys ok. PKGCONFIG += gstreamer-0.10 However if the entry is missing, then I get the error: gst/gstelement.h: No such file or directory Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Project Kits lost on Target Refresh
Hi Juha Thanks. The kits are easily added again, and this is no problem if only one or two projects are open. But I often have 5 or more open ... Another irritating strange behaviour in QtCreator, again with several projects open, Is that If I have one project selected to build and run, but the highlight in the Project Explorer (top left pane of QtCreator) is on a file in a second project, then the build fails because it gets muddled up. e.g below I am building GSTTorch, but the highlight is on a file in QMLTorch. It tries to build GSTTorch using files from QMLTorch! As I am frequently switching between projects I encounter this error many times! Chris RPM build errors: error: File not found: /home/deploy/installroot/usr/share/QMLTorch/qml error: File not found: /home/deploy/installroot/usr/share/applications/QMLTorch.desktop error: File not found: /home/deploy/installroot/usr/share/icons/hicolor/86x86/apps/QMLTorch.png error: File not found: /home/deploy/installroot/usr/share/QMLTorch File not found: /home/deploy/installroot/usr/share/QMLTorch/qml File not found: /home/deploy/installroot/usr/share/applications/QMLTorch.desktop File not found: /home/deploy/installroot/usr/share/icons/hicolor/86x86/apps/QMLTorch.png File not found: /home/deploy/installroot/usr/share/QMLTorch 13:10:48: The process /Users/christopherlamb/.config/SailfishAlpha3/mer-sdk-tools/MerSDK/SailfishOS-i486-x86/deploy exited with code 1. Error while building/deploying project GSTTorch (kit: MerSDK-SailfishOS-i486-x86) Zitat von Juha Kallioinen juha.kallioi...@jolla.com: On 21.01.2014 09:17, christopher.l...@thurweb.ch wrote: Hi All If I do a QtCreator / SailfishOS/ Targets / Manage Targets / Target / Manage / Sync, with project(s) open in QtCreator, then I lose the kits of the open project(s), and have to re-add these. Hi Chris, yes it's an unfortunate side effect of the way the build engine/qt creator interface has been implemented. This should be fixed in some future release, but is not a high priority. Best regards, Juha ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] GPS initially shows old location
Hi Aaron Thanks, I will give that a whorl and report back. I guess something like if the timestamp is older than say 10 minutes they should not be trusted (either not shown, or marked in some way). Chris Zitat von Aaron McCarthy aaron.mccar...@jolla.com: Hi, On Fri, 17 Jan 2014 17:09:37 christopher.l...@thurweb.ch wrote: Is it possible that the Jolla GPS / GPS Software stack initially shows the last location acquired? This is the behaviour I am observing. At the moment I am sitting in an office building, which has poor GPS reception, yet my app immediately shows GPS Coordinates. By peaking I can see that the GPS icon is flashing, and thus no fix has been acquired. Having checked the GPS coordinates with Google Earth, the coords are actually for a location several kilometers away - in the middle of a set of railway tracks that I travel along to and from work. As I do lots of mobile development on the train, this is a plausible location for where the Jolla last got a fix. By comparison the same app running on my old Nokia N9 gets NaN from the GPS for longitude and lattitude until it gets a fix. I can then translate this into a user friendly text on the GUI No valid position yet, and stop the user from proceeding further into the app until a true fix is acquired. If my understanding of the behaviour of the Jolla GPS is correct, is there anyway I can stop it giving me the old fix? It is important to my app that the coords shown are as accurate as possible, otherwise a recovery party / rescue helicopter may be mis-directed. This is a feature in Qt Positioning. Only the coordinates and timestamp are saved. Use the timestamp to check if the location data is old. The accuracy values of the position will be NaN. Cheers, -- Aaron McCarthy ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
[SailfishDevel] Project Kits lost on Target Refresh
Hi All If I do a QtCreator / SailfishOS/ Targets / Manage Targets / Target / Manage / Sync, with project(s) open in QtCreator, then I lose the kits of the open project(s), and have to re-add these. Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Fwd: Re: QGeoSatelliteInfoSource in Sailfish
Salut Damien I am not sure If I have understood your question, but I think you are asking about differences in direction from the QtSensors Compass and the the QtPositioning PostionSource elements. 1) The Compass Azimuth shows the direction the top of the device is pointing to (so it will change as your rotate the device). 2) The PositionSource shows a direction based on the direction the device is travelling (irrespective of where the device is pointing). http://qt-project.org/doc/qt-5/qgeopositioninfo.html Similar to 2) The speed shown by a GPS is absolute, irrespective of how the device is pointing. Some years ago I was coastal soaring on the coast of South Africa when the wind picked up, and got to the point where it was at (and possibly just above) the maximum speed of my paraglider. Flying fully into the wind, pushing maximum speedbar the GPS was showing speeds of 0.5 - 1 Km, and it was very difficult to tell if this was fowrards or backwards - but it was most likely the latter. Cheers Chris Zitat von Caliste Damien dcali...@free.fr: Bonsoir, Le dimanche 12 janvier 2014, christopher.l...@thurweb.ch a écrit : Marcel used the Compass element from QtSensors for azimuth. Ok, I understand, but I'm still wondering : - QtSensors and compas will provide the azimuth of the device, so acting it like a compas, for instance, top of the device is pointing to the north. - QGeoPosition is providing GPS info with latitude and longitude, and I was thinking that the direction attribute was in fact the first derivative of the position, which is different from the compas return value, for instance I'm moving west while pointing top of device to the north. So how to get the first derivative of GPS position, deriving it by hand ? Like calling the azimuth method on the two last GPS fix ? Have a nice week, Damien. ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Detecting if running in emulator/debug mode in QML
Hi Antoine Your point is valid, in the old days of the N9 Simulator I was able to write code to detect the simulator vs real device. But should it be necessary? Why should we need to detect a difference between an Emulator / Simulator, and a real device? Probably because of weaknesses in the Emulator / Simulator. In my humble opinion, it would be much better if the Emulator was so good that from an app developer's point of view there is no difference when running on an Emulator vs a real device. I know that by it's very nature an Emulator cannot do everything a real device can (GPS, Sensors, Camera, LED etc. spring to mind), but the N9 Simulator was a good example of what is possible in this area. Indeed the only reason that I was forced to write code to detect the N9 Simulator was due to differences in font sizing. Therefore I think Jolla's effort is best spent on providing an emulator that closely matches a real device, and in providing appropriate plugins / tools to plug the few remaining gaps (e.g a GPS simulating plugin). Chris Zitat von Artem Marchenko artem.marche...@gmail.com: You can propagate C++ knowledge to QML, Antoine For example setContextProperty #ifdef QT_QML_DEBUG Best regards, Artem. On Sun, Jan 12, 2014 at 1:41 PM, Antoine Reversat a.rever...@gmail.comwrote: Hi, I was wondering if there is a way to know if we're in debug mode or running in the emulator in QML (something like the #ifdef QT_QML_DEBUG in c++). I'd like to be able to act differently whether the app is running on a real device or not. Antoine ___ SailfishOS.org Devel mailing list -- Artem Marchenko http://agilesoftwaredevelopment.com http://twitter.com/AgileArtem ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Sharing a common OfflineStorage DB between multiple Sailfish apps.
Hi Thomas Thank you for your answer. Zitat von Thomas Perl th.p...@gmail.com: If we can fix the offline storage path to be correct everytime, we’ll probably introduce these changes into libsailfishapp, so that application developers don’t have to worry about manually setting this all the time (and you can still override this if you want, but you should not [need to]). Does this not work already? If I did not explicitly call setOfflineStoragePath, the DB was created happily enough in a subdirectory of the app without me having to do anything. Yesterday I experimented further, and arrived at this using QStandardPaths: QString storagePath = QStandardPaths::locate(QStandardPaths::GenericDataLocation, QString(), QStandardPaths::LocateDirectory) + landedcommon; view-engine()-setOfflineStoragePath(storagePath); What are the pros and cons of using QStandardPaths as opposed to XDG Environmental Variables? Zitat von Thomas Perl th.p...@gmail.com: but append the name of your application (harbour-myapp) to the path, so that all your application-specific data ends up there. In my specific case I want this data to be available to a family of apps, which implies that I can't really use the app name as you suggest. I think I could either a) use something like landedcommon as above - which is not the name of any of the apps in the family. b) use the name of the primary app in the family. But this would mean that the secondary apps would have hard-coded references to an external app (not $APPNAME). c) ensure that all apps in the family follow a standard naming pattern (e.g. harbour-landed-pilot, harbour-landed-settings, harbour-landed-recovery), and then create the common directory as a substring of $APPNAME. This raises a greater architectural question: How should families of closely related apps be structured? a) As one package, but with multiple executables b) As separate packages, but able to share stuff within the family of apps: e.g. config data / local storage, shared libraries. The shared libraries is especially important: Given the restrictive harbour rules we have to bundle not-yet-kosher libraries within an app. It would wasteful to have to repeat this for each member of an app family if they use the same libs. In my case each of the apps is a full app in its own right, with a GUI. Others have raised similar questions about app-and-daemon pairs. Grüsse Chris ___ SailfishOS.org Devel mailing list
[SailfishDevel] GPS initially shows old location
Hi all Is it possible that the Jolla GPS / GPS Software stack initially shows the last location acquired? This is the behaviour I am observing. At the moment I am sitting in an office building, which has poor GPS reception, yet my app immediately shows GPS Coordinates. By peaking I can see that the GPS icon is flashing, and thus no fix has been acquired. Having checked the GPS coordinates with Google Earth, the coords are actually for a location several kilometers away - in the middle of a set of railway tracks that I travel along to and from work. As I do lots of mobile development on the train, this is a plausible location for where the Jolla last got a fix. By comparison the same app running on my old Nokia N9 gets NaN from the GPS for longitude and lattitude until it gets a fix. I can then translate this into a user friendly text on the GUI No valid position yet, and stop the user from proceeding further into the app until a true fix is acquired. If my understanding of the behaviour of the Jolla GPS is correct, is there anyway I can stop it giving me the old fix? It is important to my app that the coords shown are as accurate as possible, otherwise a recovery party / rescue helicopter may be mis-directed. Chris ___ SailfishOS.org Devel mailing list
[SailfishDevel] Fwd: Sharing a common OfflineStorage DB between multiple Sailfish apps.
Hi all This morning on the train on the way to work I tried the following to change the location of the offlineStorage DB: The c++ main function of the demo app CreateDBQt5 now looks like this: int main(int argc, char *argv[]) { QScopedPointerQGuiApplication app(SailfishApp::application(argc, argv)); QScopedPointerQQuickView view(SailfishApp::createView()); qDebug() offlineStoragPath orig: view-engine()-offlineStoragePath(); view-engine()-setOfflineStoragePath(QString(/home/nemo/.local/share/landed25_QT5/QML/OfflineStorage )); qDebug() offlineStoragPath new: view-engine()-offlineStoragePath(); view-setSource(SailfishApp::pathTo(qml/CreateDBQt5.qml)); view-show(); view-showFullScreen(); return app-exec(); } This gives the following Application Output: [D] main:42 - offlineStoragPath orig: /home/nemo/.local/share/CreateDBQt5/QML/OfflineStorage [D] main:44 - offlineStoragPath new: /home/nemo/.local/share/landed25_QT5/QML/OfflineStorage The database is indeed created in /home/nemo/.local/share/landed25_QT5/QML/OfflineStorage, and is accessible to the app landed25_QT5 Note the order of the view-engine()-setOfflineStoragePath... and view-setSource... calls is crucial. If the setSource comes first, then the QML is initiated before the storagePath is changed! In my next experiment I will try to change the storageLocation to something like QStandardPaths::GenericDataLocation. The question on harbour acceptability remains ... Chris - Weitergeleitete Nachricht von christopher.l...@thurweb.ch - Datum: Wed, 15 Jan 2014 22:34:40 +0100 Von: christopher.l...@thurweb.ch Antwort an: Sailfish OS Developers devel@lists.sailfishos.org Betreff: [SailfishDevel] Sharing a common OfflineStorage DB between multiple Sailfish apps. An: Sailfish OS Developers devel@lists.sailfishos.org Hi all Is it possible to steer where a QML application stores its OfflineStorage Database? I have a family of applications that share a common OfflineStorage Database. This works well on Harmattan, and on the early Sailfish Alphas without me having had to do anything special at all. In the early Sailfish Alphas OfflineStorage DBs were stored in a common location, and thus could be easily shared. /home/nemo/.local/share/data/QML/OfflineStorage/Databases Now on the real Jolla, I find that the standard DB storage location has changed, and DBs are now put in a subdirectory of the application's installation folder. e.g. /home/nemo/.local/share/landed25_QT5/QML/OfflineStorage /home/nemo/.local/share/CreateDBQt5/QML/OfflineStorage This means without some magic, Sailfish apps can no longer share DBs. In C++ there is a property offlineStoragePath, and a setter method setOfflineStoragePath(). http://qt-project.org/doc/qt-5/qqmlengine.html#offlineStoragePath-prop I might be able to use this to change the location of the DB, e.g to one of the Qt StandardLocations: http://qt-project.org/doc/qt-5.0/qtcore/qstandardpaths.html#StandardLocation-enum But setOfflineStoragePath() appears not to be directly available to QML. http://stackoverflow.com/questions/14209585/how-to-set-custom-offline-storage-path-from-javascript-in-qtquick So I guess I could call setOfflineStoragePath in each apps' main cpp file. Or maybe I could even write a declarative plugin to expose setOfflineStoragepath to QML, but that is probably overkill as it would instantiate QQmlEngine - just t o set a path. Or is there a better way? Are there any Harbour entry considerations to an app creating / accessing a DB somewhere other than in its own folder? Thanks Chris ___ SailfishOS.org Devel mailing list - Ende der weitergeleiteten Nachricht - ___ SailfishOS.org Devel mailing list
[SailfishDevel] Sharing a common OfflineStorage DB between multiple Sailfish apps.
Hi all Is it possible to steer where a QML application stores its OfflineStorage Database? I have a family of applications that share a common OfflineStorage Database. This works well on Harmattan, and on the early Sailfish Alphas without me having had to do anything special at all. In the early Sailfish Alphas OfflineStorage DBs were stored in a common location, and thus could be easily shared. /home/nemo/.local/share/data/QML/OfflineStorage/Databases Now on the real Jolla, I find that the standard DB storage location has changed, and DBs are now put in a subdirectory of the application's installation folder. e.g. /home/nemo/.local/share/landed25_QT5/QML/OfflineStorage /home/nemo/.local/share/CreateDBQt5/QML/OfflineStorage This means without some magic, Sailfish apps can no longer share DBs. In C++ there is a property offlineStoragePath, and a setter method setOfflineStoragePath(). http://qt-project.org/doc/qt-5/qqmlengine.html#offlineStoragePath-prop I might be able to use this to change the location of the DB, e.g to one of the Qt StandardLocations: http://qt-project.org/doc/qt-5.0/qtcore/qstandardpaths.html#StandardLocation-enum But setOfflineStoragePath() appears not to be directly available to QML. http://stackoverflow.com/questions/14209585/how-to-set-custom-offline-storage-path-from-javascript-in-qtquick So I guess I could call setOfflineStoragePath in each apps' main cpp file. Or maybe I could even write a declarative plugin to expose setOfflineStoragepath to QML, but that is probably overkill as it would instantiate QQmlEngine - just t o set a path. Or is there a better way? Are there any Harbour entry considerations to an app creating / accessing a DB somewhere other than in its own folder? Thanks Chris ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] GPS not turned off if app exits without stopping the GPS
Hi Aaron Thanks for you rapid answer. I am on my way to work, I will play around with this further when I get back home this evening (Swiss Time). So far, on both the N9 and Jolla I used the presence of the GPS icon as an indication that it was running / not running. Once I have run the tests you suggest I will update this thread, Cheers Chris Zitat von Aaron McCarthy aaron.mccar...@jolla.com: Hi, On Mon, 13 Jan 2014 08:11:05 christopher.l...@thurweb.ch wrote: I will call this a behaviour for the moment. Yesterday I noticed that if an app that uses the GPS exits without stopping the GPS, then the GPS remains running on the Jolla. How did you confirm this? Presence of the Location icon on the lock screen? I compared this to my Nokia N9. On that device if an app using the GPS exits (e.g. the user closes it), then the GPS stops soon after that app has exited (even if the app does not explicitly shut the GPS down). On the N9 this even works with multiple apps open using the GPS. In this test, when the last GPS-consuming app exits, the GPS shuts down. This is expected to work. Both the Geoclue master and the provider processes monitor DBus disconnections and will clean up any associated interfaces. The provider processes will then close 30 seconds after there are no more connections. How to test that geoclue-master is working: Run /usr/lib/qt5/bin/qdbus org.freedesktop.Geoclue.Master | grep '/org/freedesktop/Geoclue/Master/client' | wc -l to see how many processes have started positioning. Run the above command, expect 0 Run sailfish-maps Run the above command, expect 1 pkill -9 sailfish-maps Run the above command, expect 0 It is harder to test this functionality in the providers as the number of connections are not readily available. Can you check whether there are any client interfaces exported by geoclue- master with the above command. Also what geoclue processes are running (ps aux | grep geoclue). Cheers, -- Aaron McCarthy ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Device deployment problems (SSH connection dies)
hi Jaako As you have the problem with both Wifi and USB, it does sound like something to do with the device. But for what it is worth, have you tried a different USB port? Yesterday evening I was doing quite a bit of development with both my Jolla and my N9 in developer mode connected via USB at the same time. Mostly this worked, but I soon found that while the right-hand USB on my MacBookAir always worked reliably, the lefthand port was a bit iffy - sometimes I could connect, sometimes not, and would have to plug and unplug, swap cables and devices. It may also be that the lefthand port is more sensitive to particular USB cables. Chris Zitat von Jaakko Koskenkorva jaa...@11latoa.com: I've been terribly frustrated by erratically working deployment to device from the SDK. It seems that the SSH connection dies frequently i.e. compile output shows: ssh: connect to host 192.168.2.250 port 22: Connection timed out rsync: connection unexpectedly closed (0 bytes received so far) [sender] rsync error: error in rsync protocol data stream (code 12) at io.c(226) [sender=3.1.0] ...and application output shows: SSH connection failed: Timeout waiting for reply from server. This usually happens after 1-2 deployments, after which device does not respond to ssh or ping. Things that can help in this situation (when using USB): 1) Reconnect USB cable 2) Increase timeout ...but I've yet to find a permanent solution. This also happens over WLAN with same symptoms, in this case re-connecting device to wifi router helps. Initially I thought that Ubuntu connection management was confused by the USB ethernet connection, but since WLAN has this problem too I'm thinking it's something in the device end? I'm using Ubuntu 13.10 64bit, latest SDK and v1.0.2.5 on device. Any ideas? -- Jaakko Koskenkorva Ohjelmistokehittäjä | T:mi 11latoa jaa...@11latoa.com http://11latoa.com ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] MobileNetwork / MobileData automatically on after reboot - what the heck?
Hi kaa could you explain more exactly how you are expecting Assisted GPS to work? It sounds like you are expecting it to show changes in position before a GPS fix has been acquired. My understanding is that A-GPS is intended to get a useable fix faster. If you know roughly where you are, it is easier / quicker to calculate exactly where you are. Then once you have an accurate fix using the GPS signal it is easier to spot and track small changes. mfg Chris Zitat von kaa k...@iol.cz: BTW, Assisted GPS not work. Minimally not work in my country (Czech republic). But I know I am not alone, same problem is in other countries. Positon/location is changed always only with GPS signal on my Jolla. Is required some DB of BTS? Via which provider? PS: I have unlimited O2 mobile data. kaacz Dne 13.1.2014 11:01, christopher.l...@thurweb.ch napsal(a): Hi Simon In what order did you turn mobile date off, and activate the GPS app? If the mobile data was really off, and not accessed or reactivated by the GPS then it potentially could take much longer to get a fix. Where mobile-data is permitted, then the GPS can use the A (assisted) part of A-GPS via mobile date to get a faster fix. Grüsse Chris Zitat von Simon Bolek simon.bo...@googlemail.com: Problem: After i turned on the Location (BTW waited the whole eternity to gain GPS signal) and turned the MobileData OFF, i noticed that after reboot MobileData is activated again! WHY??? I do not have a mobile data plan on my sim card, so i want it to be deactivated, and no app whatsoever has rights to turn it on (for that I could've bought android right away). Is this a BUG or a FEATURE? br simon ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QAccelerometer availability ?
Hi Franck This blog post should go somewhere towards helping you. http://flyingsheeponsailfish.blogspot.ch/2013/11/deploying-additional-packages-to.html What is your target? The Emulator or a real device? Either way, ssh into it, then run the following commands to see what is already installed: export LANG=C pkcon search name sensors On my Jolla I get: [nemo@localhost ~]$ export LANG=C [nemo@localhost ~]$ pkcon search name sensors Searching by name [=] Waiting in queue [=] Starting [=] Refreshing software list [=] Querying [=] Installed qt5-qtdeclarative-import-sensors-5.1.0+git6-1.3.10.armv7hl QtQml sensors import Available qt5-qtdeclarative-import-sensors-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtdeclarative-import-sensors Installed qt5-qtsensors-5.1.0+git6-1.3.10.armv7hl Qt Sensors module Available qt5-qtsensors-5.1.0+git6-1.3.8.source Qt Sensors module Available qt5-qtsensors-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtsensors Available qt5-qtsensors-debugsource-5.1.0+git6-1.3.10.armv7hl Debug sources for package qt5-qtsensors Available qt5-qtsensors-devel-5.1.0+git6-1.3.10.armv7hl Qt sensors - development files Available qt5-qtsensors-plugin-generic-5.1.0+git6-1.3.10.armv7hl Generic sensors plugin Available qt5-qtsensors-plugin-generic-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtsensors-plugin-generic Available qt5-qtsensors-plugin-gestures-sensor-5.1.0+git6-1.3.10.armv7hl Sensor gesture plugin Available qt5-qtsensors-plugin-gestures-sensor-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtsensors-plugin-gestures-sensor Available qt5-qtsensors-plugin-gestures-shake-5.1.0+git6-1.3.10.armv7hl Shake gesture plugin Available qt5-qtsensors-plugin-gestures-shake-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtsensors-plugin-gestures-shake Installed qt5-qtsensors-plugin-sensorfw-5.1.0+git6-1.3.10.armv7hl sensorfw sensors plugin Available qt5-qtsensors-plugin-sensorfw-debuginfo-5.1.0+git6-1.3.10.armv7hl Debug information for package qt5-qtsensors-plugin-sensorfw For qml you will need qt5-qtdeclarative-import-sensors and qt5-qtsensors. The blog post above should give you a number of ways to get these installed, including via the yaml file. If these are missing on your target then (on your target): pkcon install qt5-qtdeclarative-import-sensors This should install both packages. mfg Chris Zitat von Franck Routier (perso) a...@mecadu.org: Hi, I'm still stuck with this. Back to the basics... In my qml, I have added: import QtSensors 5.0 and used Accelerometer. Fine. But On deployment, I get an error: module QtSensors is not installed I think I have to edit something in my .pro file: I tried CONFIG += qt5-qtsensors but it didn't work. Maybe I should also make changes in the rpm yaml, although I think as it is available by default, I shouldn't. Could someone give me a pointer on putting Qt / qml modules in my project classpath ? Thanks in advance, Franck Le 11/01/2014 18:22, Mike Sheldon a écrit : Hi Frank, On Sat, 2014-01-11 at 18:03 +0100, Franck Routier (perso) wrote: So: is QAccelerometer available in Sailfish SDK ? If not, is it planned ? Or is there another way to get this kind of features ? They're part of the QtSensors module which is available in the packages: qt5-qtsensors and qt5-qtdeclarative-import-sensors (for QML support). They're both on the Harbour approved list, so you don't need to worry about bundling them with your app or anything. I'm currently using the QML Accelerometer element in a game, so I can confirm it works correctly. Cheers, Mike. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] qml PositionSource element gives error: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo'
Hi All In an attempt to track down the source of this problem, and possibly even solve it, I have created a throwaway demo posted to Github under: https://github.com/sailfishapps/ThrowawayDemos/tree/master/Kaputt/PositionSourceDemo The demo is basically a standard Sailfish new project, with a QML PositionSource element added to FirstPage, and the start() method called in a Component.onCompleted on that Page. To actually get into the nitty-gritty of the code where the problem is likely to be, I have also copied in C++ code from QtPositioning. I am not sure if this identical to the code used in the Sailfish build of Positioning, but it is probably close enough. This in-line PositionSource element is imported by import PositionSource 5.2 Switching between the standard import and the in-line import is by comment in / out the import statement(s). Of course, the problem may not be in the code at all, maybe I am missing some essential package or plugin on my Jolla .. Grüsse Chris Zitat von christopher.l...@thurweb.ch: Hi Alejandro My understanding is that although the version is quotes as 5.1, it is actually a not quite finished 5.2 In the target there is a 5.2.0 subdirectory: /Users/christopherlamb/SailfishOS/mersdk/targets/SailfishOS-armv7hl/usr/include/qt5/QtPositioning/5.2.0 import QtPositioning 5.2, import QtPositioning 5.1, and import QtPositioning 5.0 all give me the same error. Whereas something like import QtPositioning 5.7 gives module QtPositioning version 5.7 is not installed Chris Zitat von Alejandro Exojo s...@badopi.org: El Saturday 11 January 2014, christopher.l...@thurweb.ch escribió: import QtPositioning 5.2 Isn't 5.1 the newest available? -- Alex (a.k.a. suy) | GPG ID 0x0B8B0BC2 http://barnacity.net/ | http://disperso.net ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] qml PositionSource element gives error: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo'
Hi All The plot thickens! Marcel very kindly posted his code using PositionSource in QML. https://github.com/balta3/sailfish-gpsinfo So I loaded up Marcel's app, and it worked! Then I scanned the application output, and found: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo' Exactly the same error that my app was giving So it looks like this error is a bit or a red herring in that it does not stop the PositionSource from working. But the error had stopped me from further integrating the PositionSource element. Chris Zitat von christopher.l...@thurweb.ch: Hi All In an attempt to track down the source of this problem, and possibly even solve it, I have created a throwaway demo posted to Github under: https://github.com/sailfishapps/ThrowawayDemos/tree/master/Kaputt/PositionSourceDemo The demo is basically a standard Sailfish new project, with a QML PositionSource element added to FirstPage, and the start() method called in a Component.onCompleted on that Page. To actually get into the nitty-gritty of the code where the problem is likely to be, I have also copied in C++ code from QtPositioning. I am not sure if this identical to the code used in the Sailfish build of Positioning, but it is probably close enough. This in-line PositionSource element is imported by import PositionSource 5.2 Switching between the standard import and the in-line import is by comment in / out the import statement(s). Of course, the problem may not be in the code at all, maybe I am missing some essential package or plugin on my Jolla .. Grüsse Chris Zitat von christopher.l...@thurweb.ch: Hi Alejandro My understanding is that although the version is quotes as 5.1, it is actually a not quite finished 5.2 In the target there is a 5.2.0 subdirectory: /Users/christopherlamb/SailfishOS/mersdk/targets/SailfishOS-armv7hl/usr/include/qt5/QtPositioning/5.2.0 import QtPositioning 5.2, import QtPositioning 5.1, and import QtPositioning 5.0 all give me the same error. Whereas something like import QtPositioning 5.7 gives module QtPositioning version 5.7 is not installed Chris Zitat von Alejandro Exojo s...@badopi.org: El Saturday 11 January 2014, christopher.l...@thurweb.ch escribió: import QtPositioning 5.2 Isn't 5.1 the newest available? -- Alex (a.k.a. suy) | GPG ID 0x0B8B0BC2 http://barnacity.net/ | http://disperso.net ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] qml PositionSource element gives error: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo'
Hi Bob Thanks for the confirmation that things work for you. I too now suspect that it is a ghost problem. The trouble is, as soon as I saw the error I just assumed that it was a real problem and did not even bother to try and display Position info. To mix metaphors it led me up the garden path on a wild goose chase ... Chris Zitat von Bob Jelica sailf...@jelica.se: I get the same error/warning in my app, without it affecting the app. At least not enough for me to dig into the error. I still get the GPS fix, location etc, so my app is usable. Would love to find out if it’s a real problem or just a ”ghost” problem, but thus far it seems like it’s not going to give me too much problems :) //bob On 12 Jan 2014, at 20:19, christopher.l...@thurweb.ch wrote: Hi All The plot thickens! Marcel very kindly posted his code using PositionSource in QML. https://github.com/balta3/sailfish-gpsinfo So I loaded up Marcel's app, and it worked! Then I scanned the application output, and found: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo' Exactly the same error that my app was giving So it looks like this error is a bit or a red herring in that it does not stop the PositionSource from working. But the error had stopped me from further integrating the PositionSource element. Chris Zitat von christopher.l...@thurweb.ch: Hi All In an attempt to track down the source of this problem, and possibly even solve it, I have created a throwaway demo posted to Github under: https://github.com/sailfishapps/ThrowawayDemos/tree/master/Kaputt/PositionSourceDemo The demo is basically a standard Sailfish new project, with a QML PositionSource element added to FirstPage, and the start() method called in a Component.onCompleted on that Page. To actually get into the nitty-gritty of the code where the problem is likely to be, I have also copied in C++ code from QtPositioning. I am not sure if this identical to the code used in the Sailfish build of Positioning, but it is probably close enough. This in-line PositionSource element is imported by import PositionSource 5.2 Switching between the standard import and the in-line import is by comment in / out the import statement(s). Of course, the problem may not be in the code at all, maybe I am missing some essential package or plugin on my Jolla .. Grüsse Chris Zitat von christopher.l...@thurweb.ch: Hi Alejandro My understanding is that although the version is quotes as 5.1, it is actually a not quite finished 5.2 In the target there is a 5.2.0 subdirectory: /Users/christopherlamb/SailfishOS/mersdk/targets/SailfishOS-armv7hl/usr/include/qt5/QtPositioning/5.2.0 import QtPositioning 5.2, import QtPositioning 5.1, and import QtPositioning 5.0 all give me the same error. Whereas something like import QtPositioning 5.7 gives module QtPositioning version 5.7 is not installed Chris Zitat von Alejandro Exojo s...@badopi.org: El Saturday 11 January 2014, christopher.l...@thurweb.ch escribió: import QtPositioning 5.2 Isn't 5.1 the newest available? -- Alex (a.k.a. suy) | GPG ID 0x0B8B0BC2 http://barnacity.net/ | http://disperso.net ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] Fwd: Re: QGeoSatelliteInfoSource in Sailfish
Salut Damien Marcel used the Compass element from QtSensors for azimuth. Chris Zitat von dcali...@free.fr: Hello, À Sun Jan 12 2014 17:17:11 GMT+0100 (CET), Marcel a écrit : The QML position source is working here, and you can see the compass access, too Do you have the direction attribute of QGeoPositionInfo that is set ? When using hasAttribute() method on a valid GPS info source, I always get false. Thanks, Damien. ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] qml PositionSource element gives error: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo'
HI Aaron Thanks for this. I am glad to hear a fix is underway. I was about to play around with solution like the one you suggest until others in this mailing list provided me with demo code that indicates that despite the error, the PositionSource element seems to work! If so, is this actually a ghost error that we can ignore for the moment? Grüsse Chris Zitat von Aaron McCarthy aaron.mccar...@jolla.com: Hi, On Sat, 11 Jan 2014 18:08:25 christopher.l...@thurweb.ch wrote: I have added a QML PositionSource element to the FirstPage of the default HelloSailors new project, and call the start() method from Component.onCompleted (see code snippet below mail). When running on the Emulator I get no error: When running on a real Jolla I get the following error: QMetaMethod::invoke: Unable to handle unregistered datatype 'QGeoPositionInfo' I have the GPS enabled, I have used the maps app and agreed to the licence there. This is a problem with the Qt Geoclue positioning plugin, it fails to register the data type. I have submitted a fix upstream [1]. In the meantime you can work around it in you application by calling qRegisterMetaTypeQGeoPositionInfo(); prior to creating a position info source. [1] https://codereview.qt-project.org/75256 Cheers, -- Aaron McCarthy ___ SailfishOS.org Devel mailing list ___ SailfishOS.org Devel mailing list
Re: [SailfishDevel] QGeoSatelliteInfoSource in Sailfish
Hi Aaron Zitat von Aaron McCarthy aaron.mccar...@jolla.com: Hi, Unfortunately, due to a build configuration issue, the satellite interface is not available in the current released version. The issue has already been fixed and will be released with a future update. Thanks for the confirmation that this is not yet working and that a fix is in the works. Just out of interest of how thinks work, could you provide a more detailed explanation of what is missing / broken? I am guessing a geoclue provider, but may be way-off. Cheers Chris ___ SailfishOS.org Devel mailing list