GSoC2010 global search application proposal discussion
Dear Maemo developers, now that Maemo has been accepted as a mentoring organization for GSoC 2010, I'd like to ask about joining the action. I've been pondering on what project to look into, as quite a few of them interest me (including porting Remmina, the Ebook reader idea and an idea of my own, which revolves around recovering data from your device if it gets stolen and getting a GPS lock upon sending it an SMS or similar things), but I think I'll settle for the Global Search Application Project. What is it about? I'll just rephrase the proposal from the project ideas page and then share my ideas on how to implement and present it: == Global Search Application An application which indexes local files and content and supports searching for files and content in a convenient and appealing fashion as well as opening results in their respective application. == Thoughts about presentation and implementation The search application should respond quickly and present results in an appealing fashion while the user is typing - possibly supported by thumbnails - and should open them in the appropriate application if possible. It should also include searching for contacts (including contact fields, e.g. jabber nicknames and such), SMS, emails and other content commonly stored on the device. Indexing shouldn't waste battery life or impact performance while the user is actively using the device. While looking around for clues, I've seen that this idea has been around for a while with lots of discussions happening around the 2008 Maemo summit ( https://wiki.maemo.org/Desktop_Search_Hackfest ). Solutions discussed included Strigi, Beagle and Xesam, but it seems like there hasn't been a final product of sorts. I myself have been thinking about adapting some more streamlined solutions like locate (updatedb), and building around those. Quite interesting for such a device would be rlocate because it's small, fast and will know about file changes, but that needs it's own kernel module which might not be feasable for the platform. If using one of these, there'a also another component needed for indexing file content, and of course there need to be special components for accessing the maemo-specific data like contacts and so on. There's definitely lots of way we could do this, and the preferred method would need to be found. Concrete tasks would include collecting the existing bits and pieces, choosing the grounds to build upon, writing or adapting, porting and packaging a backend indexer, writing a frontend, preferrably as a widget, and conduction reliability and speed tests. == About myself My name is Carol Alexandru and I'm studying Computer Sciences at the University of Zurich, Switzerland (UZH) in the 4th semester. I've been a Linux user for around 5 years and have lots of experience regarding the CLI and the underlying mechanisms of Linux. During my study I've acquired skills in modeling, project planning and of course programming - the usual stuff. I'm highly motivated to work for the community and to write open source software. I've been working on a larger project of my own (creating a multi-user music streaming server, which I have not yet released; It needs a few finishing touches) and lots of minor hacks and tools, mostly for the CLI. I'm also the author of the Alarmed Qt/CLI wrapper around alarmd/libalarm for Maemo 5 ( http://talk.maemo.org/showthread.php?t=46594 ). Now, I *love* Python but I also know the more common languages like C/C++. During my spare time, I like to just hack small scripts'n'stuff, maintain our student clubs servers and infrastructure and have a couple of beers with friends. And of course I've got an N900 ;-) Regarding other commitments during summer, I need to pass a few crucial exams in the first half of June. Apart from that, I'm 100% available. == Contact info I'm always on IRC in irc.freenode.net#maemo, nickname Shapeshifter, and of course you can reach me through this email address: shapeshif...@archlinux.us. So, I know it's a bit early for a full-blown application, as the period starts on March 28th, so please consider this an expression of my interest. :-) I'd really like for possible mentors on this to get in touch with me to discuss details and see if this is a project fitting into Maemo and if I'm the right person for the job. I think quite a few people would like the idea and profit from such an application. Thanks for your time. Regards, Carol Alexandru ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Re: GSoC2010 global search application proposal discussion
Hi Carol, good idea, an index is not bad, but what about starting with a website index? www.yacy.net it allows to search for webpages in a browser, a mameo compatible browser is http://dooble.sf.net Maybe you want to compile it? as it has a file explorer integrated, you do not need a new indexer, as they want launchy.sf.net in the desktop, that is the indexer you can look for? Regards On Sun, Mar 21, 2010 at 4:21 PM, Shapeshifter shapeshif...@archlinux.us wrote: Dear Maemo developers, now that Maemo has been accepted as a mentoring organization for GSoC 2010, I'd like to ask about joining the action. I've been pondering on what project to look into, as quite a few of them interest me (including porting Remmina, the Ebook reader idea and an idea of my own, which revolves around recovering data from your device if it gets stolen and getting a GPS lock upon sending it an SMS or similar things), but I think I'll settle for the Global Search Application Project. What is it about? I'll just rephrase the proposal from the project ideas page and then share my ideas on how to implement and present it: == Global Search Application An application which indexes local files and content and supports searching for files and content in a convenient and appealing fashion as well as opening results in their respective application. == Thoughts about presentation and implementation The search application should respond quickly and present results in an appealing fashion while the user is typing - possibly supported by thumbnails - and should open them in the appropriate application if possible. It should also include searching for contacts (including contact fields, e.g. jabber nicknames and such), SMS, emails and other content commonly stored on the device. Indexing shouldn't waste battery life or impact performance while the user is actively using the device. While looking around for clues, I've seen that this idea has been around for a while with lots of discussions happening around the 2008 Maemo summit ( https://wiki.maemo.org/Desktop_Search_Hackfest ). Solutions discussed included Strigi, Beagle and Xesam, but it seems like there hasn't been a final product of sorts. I myself have been thinking about adapting some more streamlined solutions like locate (updatedb), and building around those. Quite interesting for such a device would be rlocate because it's small, fast and will know about file changes, but that needs it's own kernel module which might not be feasable for the platform. If using one of these, there'a also another component needed for indexing file content, and of course there need to be special components for accessing the maemo-specific data like contacts and so on. There's definitely lots of way we could do this, and the preferred method would need to be found. Concrete tasks would include collecting the existing bits and pieces, choosing the grounds to build upon, writing or adapting, porting and packaging a backend indexer, writing a frontend, preferrably as a widget, and conduction reliability and speed tests. == About myself My name is Carol Alexandru and I'm studying Computer Sciences at the University of Zurich, Switzerland (UZH) in the 4th semester. I've been a Linux user for around 5 years and have lots of experience regarding the CLI and the underlying mechanisms of Linux. During my study I've acquired skills in modeling, project planning and of course programming - the usual stuff. I'm highly motivated to work for the community and to write open source software. I've been working on a larger project of my own (creating a multi-user music streaming server, which I have not yet released; It needs a few finishing touches) and lots of minor hacks and tools, mostly for the CLI. I'm also the author of the Alarmed Qt/CLI wrapper around alarmd/libalarm for Maemo 5 ( http://talk.maemo.org/showthread.php?t=46594 ). Now, I *love* Python but I also know the more common languages like C/C++. During my spare time, I like to just hack small scripts'n'stuff, maintain our student clubs servers and infrastructure and have a couple of beers with friends. And of course I've got an N900 ;-) Regarding other commitments during summer, I need to pass a few crucial exams in the first half of June. Apart from that, I'm 100% available. == Contact info I'm always on IRC in irc.freenode.net#maemo, nickname Shapeshifter, and of course you can reach me through this email address: shapeshif...@archlinux.us. So, I know it's a bit early for a full-blown application, as the period starts on March 28th, so please consider this an expression of my interest. :-) I'd really like for possible mentors on this to get in touch with me to discuss details and see if this is a project fitting into Maemo and if I'm the right person for the job. I think quite a few people would like the idea and profit from such an application. Thanks for your time. Regards
Re: GSoC2010 global search application proposal discussion
Hello there, Max I just installed launchy on my desktop and wow, is pretty cool! Who are you refering to saying they want launchy in the desktop? It seems to be a pretty nice solution, and it seems like it only depends on Qt which fits perfectly to the MeeGo/Harmattan future. So, the task would simply be to widgetize launchy and compile it for ARM and package it? Doesn't seem like such a big task, but maybe I'm underestimating this? I see it supports plugins (maybe even QtScript in the future). Maybe porting it _and_ writing a plugin for it? I'm not so sure about dooble, though. A browser wich a file manager and a desktop? I'm pretty much for the old linux philosophy do only one thing but do it right. Seems a bit odd to me to mix them. And I see they next want to add support for tagging mp3s even! There's also loads of dedicated linux webkit browsers around already, and for example midori is already ported. YaCy seems like an interesting thing but I'm not sure how many people would be using it. Regards, Carol On 21/03/10 17:51, Max wrote: Hi Carol, good idea, an index is not bad, but what about starting with a website index? www.yacy.net it allows to search for webpages in a browser, a mameo compatible browser is http://dooble.sf.net Maybe you want to compile it? as it has a file explorer integrated, you do not need a new indexer, as they want launchy.sf.net in the desktop, that is the indexer you can look for? Regards On Sun, Mar 21, 2010 at 4:21 PM, Shapeshiftershapeshif...@archlinux.us wrote: Dear Maemo developers, now that Maemo has been accepted as a mentoring organization for GSoC 2010, I'd like to ask about joining the action. I've been pondering on what project to look into, as quite a few of them interest me (including porting Remmina, the Ebook reader idea and an idea of my own, which revolves around recovering data from your device if it gets stolen and getting a GPS lock upon sending it an SMS or similar things), but I think I'll settle for the Global Search Application Project. What is it about? I'll just rephrase the proposal from the project ideas page and then share my ideas on how to implement and present it: == Global Search Application An application which indexes local files and content and supports searching for files and content in a convenient and appealing fashion as well as opening results in their respective application. == Thoughts about presentation and implementation The search application should respond quickly and present results in an appealing fashion while the user is typing - possibly supported by thumbnails - and should open them in the appropriate application if possible. It should also include searching for contacts (including contact fields, e.g. jabber nicknames and such), SMS, emails and other content commonly stored on the device. Indexing shouldn't waste battery life or impact performance while the user is actively using the device. While looking around for clues, I've seen that this idea has been around for a while with lots of discussions happening around the 2008 Maemo summit ( https://wiki.maemo.org/Desktop_Search_Hackfest ). Solutions discussed included Strigi, Beagle and Xesam, but it seems like there hasn't been a final product of sorts. I myself have been thinking about adapting some more streamlined solutions like locate (updatedb), and building around those. Quite interesting for such a device would be rlocate because it's small, fast and will know about file changes, but that needs it's own kernel module which might not be feasable for the platform. If using one of these, there'a also another component needed for indexing file content, and of course there need to be special components for accessing the maemo-specific data like contacts and so on. There's definitely lots of way we could do this, and the preferred method would need to be found. Concrete tasks would include collecting the existing bits and pieces, choosing the grounds to build upon, writing or adapting, porting and packaging a backend indexer, writing a frontend, preferrably as a widget, and conduction reliability and speed tests. == About myself My name is Carol Alexandru and I'm studying Computer Sciences at the University of Zurich, Switzerland (UZH) in the 4th semester. I've been a Linux user for around 5 years and have lots of experience regarding the CLI and the underlying mechanisms of Linux. During my study I've acquired skills in modeling, project planning and of course programming - the usual stuff. I'm highly motivated to work for the community and to write open source software. I've been working on a larger project of my own (creating a multi-user music streaming server, which I have not yet released; It needs a few finishing touches) and lots of minor hacks and tools, mostly for the CLI. I'm also the author of the Alarmed Qt/CLI wrapper around alarmd/libalarm for Maemo 5 ( http://talk.maemo.org/showthread.php?t=46594 ). Now, I *love* Python but I also know
Re: GSoC2010 global search application proposal discussion
On Sun, Mar 21, 2010 at 5:21 PM, Shapeshifter shapeshif...@archlinux.us wrote: == Global Search Application An application which indexes local files and content and supports searching for files and content in a convenient and appealing fashion as well as opening results in their respective application. == Thoughts about presentation and implementation The search application should respond quickly and present results in an appealing fashion while the user is typing - possibly supported by thumbnails - and should open them in the appropriate application if possible. It should also include searching for contacts (including contact fields, e.g. jabber nicknames and such), SMS, emails and other content commonly stored on the device. Indexing shouldn't waste battery life or impact performance while the user is actively using the device. While looking around for clues, I've seen that this idea has been around for a while with lots of discussions happening around the 2008 Maemo summit ( https://wiki.maemo.org/Desktop_Search_Hackfest ). Solutions discussed included Strigi, Beagle and Xesam, but it seems like there hasn't been a final product of sorts. I myself have been thinking about adapting some more streamlined solutions like locate (updatedb), and building around those. Quite interesting for such a device would be rlocate because it's small, fast and will know about file changes, but that needs it's own kernel module which might not be feasable for the platform. If using one of these, there'a also another component needed for indexing file content, and of course there need to be special components for accessing the maemo-specific data like contacts and so on. There's definitely lots of way we could do this, and the preferred method would need to be found. Concrete tasks would include collecting the existing bits and pieces, choosing the grounds to build upon, writing or adapting, porting and packaging a backend indexer, writing a frontend, preferrably as a widget, and conduction reliability and speed tests. Maemo 5 uses Tracker 0.6 to consolidate most of metadata of files indexed already. Harmattan (Maemo 6) is extensively using current Tracker master branch and utilizing SparQL language as its query mechanism. Tracker master (0.7.x) and Tracker 0.6 use different storage backend and query mechanisms, it is quite a big difference, and it seems most of non-file based data is going to be indexed as well, according to the ontologies being developed. You can find current maemo-specific modifications of Tracker at http://maemo.gitorious.org/tracker/, these are synced well with Tracker master at gnome.org. If you want to develop something useful for Maemo 5 and 6 indexing, it would logically be good to at least learn what is happening in this area already and co-operate with existing code running on the device. -- / Alexander Bokovoy ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Re: GSoC2010 global search application proposal discussion
On 21/03/10 21:43, Alexander Bokovoy wrote: On Sun, Mar 21, 2010 at 5:21 PM, Shapeshiftershapeshif...@archlinux.us wrote: == Global Search Application An application which indexes local files and content and supports searching for files and content in a convenient and appealing fashion as well as opening results in their respective application. Maemo 5 uses Tracker 0.6 to consolidate most of metadata of files indexed already. Harmattan (Maemo 6) is extensively using current Tracker master branch and utilizing SparQL language as its query mechanism. Tracker master (0.7.x) and Tracker 0.6 use different storage backend and query mechanisms, it is quite a big difference, and it seems most of non-file based data is going to be indexed as well, according to the ontologies being developed. You can find current maemo-specific modifications of Tracker at http://maemo.gitorious.org/tracker/, these are synced well with Tracker master at gnome.org. If you want to develop something useful for Maemo 5 and 6 indexing, it would logically be good to at least learn what is happening in this area already and co-operate with existing code running on the device. I see, thanks for the hint! I saw this tracker on the project proposal page and it didn't catch my attention because I thought it was somehow refering to the bug tracker or something. I was wondering what it did there and I didn't realize it was a project itself! That makes a lot of sense, no doubt. I'm reading through the Tracker wiki on gnome.org now to learn more about how the whole thing works. Building on tracker for a desktop search app would of course be the most obvious thing to do! Thanks again. Regards, Carol ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Global Search
Hi, trying to follow the maemo tutorial[1], I have difficulties finding the libogs-dev package. Could you give me a hint how to obtain it? Kind regards T. 1. http://maemo.org/development/documentation/tutorials/maemo_4-0_tutorial.html#ogs -- Thomas Viehmann, http://thomas.viehmann.net/ ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Re: Global Search
Thomas Viehmann wrote: Hi, trying to follow the maemo tutorial[1], I have difficulties finding the libogs-dev package. Could you give me a hint how to obtain it? Kind regards T. 1. http://maemo.org/development/documentation/tutorials/maemo_4-0_tutorial.html#ogs Maemo documentation never seems to mention when software packages are not actually part of (free) maemo, but Nokia binary packages. The chinook installation instuctions do cover installing the binary packages, even if it's a bit confusing in places. The installer script should have downloaded the packages to ~/maemo-sdk-nokia-binaries_4.0/ (inside scratchbox) and added the correct source line to /etc/apt/sources.list. After this it's just a matter of fakeroot apt-get update fakeroot apt-get install maemo-explicit HTH, Jussi ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
