Arch Linux User Repository

I've created an issue upstream to track requirements for this build to make it to [extra]. As soon as those are met I'll make an official build for this. The project is a lot closer than it was even the last time I reviewed the status. If you want this to happen I'm sure the upstream project could use a hand with a few of the remaining details for devendoring.

@schnedan I have no plans to cater to broken AUR helpers. This package builds cleanly with no interaction and passing checks in its current form in a clean chroot using Arch tooling. I also publish the built packages in my user repository, signed with official Arch Linux keys, so if you are so worried about just getting work done just use that.

Hi,

For some reason, I have been getting the behaviour described here: https://github.com/Murmele/Gittyup/issues/543 (my issue) which is something to do with libssh2 and accepted signature algorithms, but when I build Gittyup myself following instructions from the README, the issue is no longer there. Hence, I wonder if something in the PKGBUILD is introducing this bug. I'm happy to help debug if needed.

Best,

Mark.

Again, please remove the check from makepkg, so AUR-helpers run without interaction (again). In the last 3 month packages which do not build with default setting is accumulating. This situation is horrible. Please note: people use Arch to do work and not to play developer.

@schnedan @denzil --nocheck option exist for makepkg

Yah, commenting out is obvios. But if the checks are added to a build on a specific machine, there must be a reason. And if they fail, I guess something is wrong or broken.

So I think a.) Checks not needed? then the maintainer should remove em b.) tests needed? then the maintainer should ensure they do not fail in normal enviroments.

plus: editing PKGBUILDs by hand is a pia. It makes AUR Helpers fail and If you have several running systems it costs a lot of time. Maintainers should always have the correct mindset, and Arch as one of the largest distros around is also used to earn money. Thats no freetime toy! People depend on that stuff just works!

If you're having issues with failing tests.. easy workaround is just skipping the tests in general. Just edit your PKGBUILD and comment out line 82: ninja -C build check, or remove it completely with the surrounding IF

Tested on 2 Systems, Build fails with failed unittests:

QWARN : TestEditorLineInfo::discardCompleteAddedContent() This plugin does not support propagateSizeHints()
QWARN : TestEditorLineInfo::discardCompleteAddedContent() This plugin does not support propagateSizeHints()
QWARN : TestEditorLineInfo::discardCompleteAddedContent() This plugin does not support propagateSizeHints()
QWARN : TestEditorLineInfo::discardCompleteAddedContent() QProcess: Destroyed while process ("/tmp/trizen-danny/gittyup/src/build/test/indexer") is still running.
PASS : TestEditorLineInfo::discardCompleteAddedContent()
PASS : TestEditorLineInfo::cleanupTestCase()
Totals: 14 passed, 7 failed, 0 skipped, 0 blacklisted, 9482ms
* Finished testing of TestEditorLineInfo *

  Start 15: CommitAuthorCommitter

15/23 Test #15: CommitAuthorCommitter ............ Passed 0.59 sec
Start 16: rebase
16/23 Test #16: rebase ........................... Passed 4.42 sec
Start 17: TreeView
17/23 Test #17: TreeView ......................... Passed 2.61 sec
Start 18: Submodule
18/23 Test #18: Submodule ........................ Passed 2.45 sec
Start 19: referencelist
19/23 Test #19: referencelist .................... Passed 0.91 sec
Start 20: amend
20/23 Test #20: amend ............................ Passed 0.91 sec
Start 21: SshConfig
21/23 Test #21: SshConfig ........................ Passed 0.16 sec
Start 22: fileContextMenu
22/23 Test #22: fileContextMenu .................. Passed 2.71 sec
Start 23: Setting
23/23 Test #23: Setting .......................... Passed 0.14 sec

96% tests passed, 1 tests failed out of 23

Label Time Summary:
NO_WIN32_OFFSCREEN = 5.11 sec*proc (4 tests)

Total Test time (real) = 40.23 sec

The following tests FAILED:
14 - EditorLineInfos (Failed)
Errors while running CTest
FAILED: test/CMakeFiles/check /tmp/trizen-danny/gittyup/src/build/test/CMakeFiles/check
cd /tmp/trizen-danny/gittyup/src/build/test && /usr/bin/ctest --force-new-ctest-process --output-on-failure --timeout 300
ninja: build stopped: subcommand failed.
==> FEHLER: Ein Fehler geschah in check().
Breche ab...
:: Unable to build gittyup - makepkg exited with code: 4
=>> Try again? [y/N]:

@teras: As I've mentioned in the comments before, try my PKGBUILD.

I have the same problem with @denzil

Still two errors.

FAIL!  : TestIndex::stageDirectory() Compared values are not the same
   Actual   (unstagedModel->rowCount()): 2
   Expected (1)                        : 1
   Loc: [/home/teras/.cache/paru/clone/gittyup/src/Gittyup/test/index.cpp(157)]

FAIL!  : TestTreeView::restoreStagedFileAfterCommit() Compared values are not the same
   Actual   (unstagedModel->data(file_txt).toString()): "folder/subfolder/file.txt"
   Expected (QString("file.txt"))                     : "file.txt"
   Loc: [/home/teras/.cache/paru/clone/gittyup/src/Gittyup/test/TreeView.cpp(59)]

The funny thing is, I installed it two days ago for the first time (version 1.2.0). Now it updates to 1.2.1 and the error appears. Could it be, that what it "finds' is the older build version?

I actually tried uninstalling and reinstalling , even removing the cache, but the only result was not to have gittyup any more installed on my system :(

For me it's building normally.

TIP: If you want to get rid of the "recent files" that comes with the installation open the file nano ~/.config/gittyup.github.com/Gittyup.conf and delete what is in "recent"...

[General]
recent=
[...]

.