Re: Making Purpose part of KF5
Am 10.11.2017 um 12:50 schrieb Aleix Pol: > I share the concern in fact, part of the reason why I didn't push it > earlier was to be able to have it proper tier2. My reasoning for > leaving it as this is that it just adds an extra step users will have > to take to install the plugins and it's not a very useful one (to pull > the plugins). Without plugins, purpose is useless. This means: purpose is the framework and ought to be in"frameworks". Maybe it should be called "purpose-framework" to emphasis this. The plugins have more dependencies and should not go into the framework, though. If purpose would introduce higher level dependencies, how should this be build then? You'll end up with cyclic dependencies, which are a horror to build. Please move the plugins into a separate repository or at least make them to be build *completely* separate (like a subproject.) -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel added a comment. My aim is to bring Plasma into GuixSD, not to develop ktexteditor. So please understand that I'm not going to work on ktexteditor. REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D8577 To: hgoebel, #ktexteditor, dfaure Cc: cullmann, dhaumann, #frameworks, kevinapavew, demsking, sars
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel added a comment. Well, the currently the dependency is still needed, see https://cgit.kde.org/ktexteditor.git/tree/autotests/src/scripting_test.cpp#n51: #include If the code is not compiled in an isolated and controlled environment, and QtScript is installed on the system, the compiler might silently get the includes installed elsewhere. The same is true, if some QT package requires QtScript and QtScript is installed alongside that package. REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D8577 To: hgoebel, #ktexteditor, dfaure Cc: cullmann, dhaumann, #frameworks, kevinapavew, demsking, sars
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel added a comment. QtScript is only required for the tests. Maybe just a left-over #include? REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D8577 To: hgoebel, #ktexteditor, dfaure Cc: dhaumann, #frameworks, kevinapavew, demsking, cullmann, sars
Test failures with krunner 5.39.0
Hallo, I'm packaging krunner 5.39.0 for GNU Guix [1]. GNU Guix builds all software in an isolated environment (a container), which includes only the packages and tools specified explicitly. When running the tests, dbusrunnertest faills, with this notable warning and the error show below: DBusRunnerTest::testMatch() org.kde.krunner: Invalid entry: "DBus runner test" Any idea what may be missing, what needs to be set up (env-var, directory, package, paths)? Any idea hos to debug this further? All required packages and all optional/feature packages (except of QCH) are installed. I tried running the tests with both dbus-launch and without. The external program "testremoterunner" gets started. strace -e trace=file" shows nothing relevant (as far as I can tell). * Start testing of DBusRunnerTest * Config: Using QtTest library 5.9.2, Qt 5.9.2 (x86_64-little_endian-lp64 shared (dynamic) release build; by GCC 5.4.0) PASS : DBusRunnerTest::initTestCase() QWARN : DBusRunnerTest::testMatch() inotify_add_watch("/etc/mtab") failed: "No such file or directory" QWARN : DBusRunnerTest::testMatch() inotify_add_watch("/etc/fstab") failed: "No such file or directory" QWARN : DBusRunnerTest::testMatch() org.kde.krunner: Invalid entry: "DBus runner test" FAIL! : DBusRunnerTest::testMatch() 'spy.wait()' returned FALSE. () Loc: [/tmp/guix-build-krunner-5.39.0.drv-0/krunner-5.39.0/autotests/dbusrunnertest.cpp(82)] PASS : DBusRunnerTest::cleanupTestCase() Totals: 2 passed, 1 failed, 0 skipped, 0 blacklisted, 4931ms ***** Finished testing of DBusRunnerTest * -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel added a comment. Please note that I do not have commit rights, so please somebody else commit. REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D8577 To: hgoebel, #ktexteditor Cc: #frameworks, kevinapavew, demsking, cullmann, sars, dhaumann
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel added a reviewer: KTextEditor. REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D8577 To: hgoebel, #ktexteditor Cc: #frameworks, kevinapavew, demsking, cullmann, sars, dhaumann
D8577: Fix: Missing dependencies for ktexeditor autotests
hgoebel updated this revision to Diff 21656. hgoebel changed the visibility from "No One" to "Public (No Login Required)". REPOSITORY R39 KTextEditor CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D8577?vs=21652=21656 REVISION DETAIL https://phabricator.kde.org/D8577 AFFECTED FILES ktexteditor-5.39.0/CMakeLists.txt ktexteditor-5.39.0/autotests/CMakeLists.txt ktexteditor-5.39.0/autotests/src/vimode/CMakeLists.txt To: hgoebel Cc: #frameworks, kevinapavew, demsking, cullmann, sars, dhaumann
Test failures with kdelibs4support
Hi, I try building kdelibs4support 5.32.0 for GNU Guix <https://www.gnu.org/software/guix/>. GNU Guix builds all software within a container, which does not have any network interface nor any service running except if explicitly started. kdelibs4support builds fine, but some tests are failing. Any hints what could cause this or how to debug it? Where can I find the CI logs for this package to compare the results? 1) KStandarddirsTest::testFindResource FAIL! : KStandarddirsTest::testFindResource() '!data.isEmpty()' returned FALSE. () Loc: [/tmp/guix-build-kdelibs4support-5.32.0.drv-0/kdelibs4support-5.32.0/autotests/kstandarddirstest.cpp(160)] stracing showed that the requested config dir does not exist: stat("/tmp/guix-build-kdelibs4support-5.32.0.drv-0/build/.kde-unit-test/xdg/config/foorc", 0x7ffc0658bf30) = -1 ENOENT (No such file or directory) 2) KMimeTypeTest::testMimeTypeTraderForTextPlain (and others) - QWARN : KMimeTypeTest::testMimeTypeTraderForTextPlain() KMimeTypeTrader: mimeType "text/plain" not found FAIL! : KMimeTypeTest::testMimeTypeTraderForTextPlain() 'offerListHasService(offers, "faketextpart.desktop")' returned FALSE. () Loc: [/tmp/guix-build-kdelibs4support-5.32.0.drv-0/kdelibs4support-5.32.0/autotests/kmimetypetest.cpp(669)] I could not find the cause for this one: ksycoca is reading the .desktop-files generated by the test-suite, but the tests fail. Adding "shared-mime-info" to the container did not help. -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
Re: Test failures with networkmanager-qt
Am 03.04.2017 um 14:16 schrieb Jan Grulich: > Pushed two fixes into master branch, hopefully they fix all issues with the > failing unit test. Applying both patches make the tests pass. Thanks for the fix. -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
Re: Test failures with networkmanager-qt
Hi Jan, > could you please try attached patch? For me it was working even without the > patch, but I think I spotted a potential problem and this hopefully will fix > it. Thanks for the patch, unfortunately it did not solve the problem. I'll send you the log-output via personal mail to avoid cluttering the mainling-list. (Sorry for not answering earlier, I bas busy with my day-job.) -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
Re: Test failures with networkmanager-qt
Am 28.03.2017 um 11:43 schrieb Hartmut Goebel: > Which versions are you referring to in "the only relevant change". Guix > is currently lacking behind and still has 5.28, which we updated from > 5.27 (which passed tests). Update: The error occurs even with 5.32. -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
Re: Test failures with networkmanager-qt
Am 28.03.2017 um 08:10 schrieb Jan Grulich: > I can't think of any reason why the test fails for you. Looking to the git > log > the only relevant change is fixed uninitialized state variable in > ActiveConnection, which may affect ActiveConnection::stateChanged() signal. So you suggest to not try to find the cause but simply ignore this failure? Which versions are you referring to in "the only relevant change". Guix is currently lacking behind and still has 5.28, which we updated from 5.27 (which passed tests). Between 5.27 and 5.28 here is one change that might be relevant: "Fix unit tests" [1]. [1] <https://cgit.kde.org/networkmanager-qt.git/commit/?id=cc9d198676e8c8afae9e55286bd401f40634ce6b> -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |
Test failures with networkmanager-qt
Hi, I try building networkmanager-qt for GNU Guix [1]. Gnu Guix build all software within a container, which does not have any network interface. networkmanager-qt build fine, but one test (ActiveConnectionTest) fails with activeConnectionStateChangedSpy.count()) being 0 instead of 1. There are also some non-fatal "Unhandled property" errors. Any hints what could cause this or how to debug it? networkmanager-qt is 5.28.0, compared to the up-to-date version I could not spot any related change. I tried with both (gnome) network-manger is 1.4.4 and 1.6.2 [1] https://www.gnu.org/software/guix/ --8<---cut here---start->8--- 3: Test command: /tmp/guix-build-networkmanager-qt-5.28.0.drv-0/build/autotests/activeconnectiontest 3: Test timeout computed to be: 9.99988e+06 3: * Start testing of ActiveConnectionTest * 3: Config: Using QtTest library 5.7.1, Qt 5.7.1 (x86_64-little_endian-lp64 shared (dynamic) release build; by GCC 4.9.4) 3: QWARN : ActiveConnectionTest::initTestCase() networkmanager-qt: void NetworkManager::NetworkManagerPrivate::propertiesChanged(const QVariantMap&) Unhandled property "Devices" 3: QWARN : ActiveConnectionTest::initTestCase() networkmanager-qt: virtual void NetworkManager::DevicePrivate::propertyChanged(const QString&, const QVariant&) Unhandled property "Carrier" 3: QWARN : ActiveConnectionTest::initTestCase() networkmanager-qt: virtual void NetworkManager::DevicePrivate::propertyChanged(const QString&, const QVariant&) Unhandled property "HwAddress" 3: QWARN : ActiveConnectionTest::initTestCase() networkmanager-qt: virtual void NetworkManager::DevicePrivate::propertyChanged(const QString&, const QVariant&) Unhandled property "PermHwAddress" 3: QWARN : ActiveConnectionTest::initTestCase() networkmanager-qt: virtual void NetworkManager::DevicePrivate::propertyChanged(const QString&, const QVariant&) Unhandled property "Speed" 3: PASS : ActiveConnectionTest::initTestCase() 3: FAIL! : ActiveConnectionTest::testActiveConnection() Compared values are not the same 3:Actual (activeConnectionStateChangedSpy.count()): 0 3:Expected (1) : 1 3:Loc: [/tmp/guix-build-networkmanager-qt-5.28.0.drv-0/networkmanager-qt-5.28.0/autotests/activeconnectiontest.cpp(118)] 3: PASS : ActiveConnectionTest::cleanupTestCase() 3: Totals: 2 passed, 1 failed, 0 skipped, 0 blacklisted, 805ms 3: * Finished testing of ActiveConnectionTest * 3/25 Test #3: activeconnectiontest .***Failed0.81 sec --8<---cut here---end--->8--- -- Regards Hartmut Goebel | Hartmut Goebel | h.goe...@crazy-compilers.com | | www.crazy-compilers.com | compilers which you thought are impossible |