tint2 at https://gitlab.com/o9000/tint2 is sentenced to end of life by recent maintainer Chris Lee, right at this year - version 17.0.2 is told to be final. I took liberty to adopt it, for now at https://gitlab.com/nick87720z/tint2. There were some visible problems, set aside that its code could be more optimal. ChangeLog: https://gitlab.com/nick87720z/tint2/-/blob/master/ChangeLog For shott - visible changes affect tooltip behavior, taskbar color tinting and tint2conf behavior. I'm less sure about optimizations, because there's still too much to grind. Unfortunally, one of my changes makes impossible to build with cmake->ninja generator, though regular make is fine. Not specific to portage or cmake.eclass. I managed to make complete build with unix makefile generator, but not sure if it's satisfying for portage package. If someone can help to make it work with ninja again - would be awesome. For now - I have live ebuild to submit, using makefile generator. Also - my time is too limited, probably not much better, than for Chris Lee. So I don't expect my pace to be even moderate. Any help with upstream is welcome. Reproducible: Always
Created attachment 764844 [details] tint2-9999.ebuild Forgot the ebuild
Created attachment 765700 [details] tint2-9999.ebuild Solved ninja problem at the cost of manual sources regeneration. 17.1 is current last release (I would propose version bump).
Created attachment 765701 [details] tint2-17.1.ebuild And ebuild for release. Besides changed URI, it's also enhanced a bit, based on ebuild for 17.0.1 from main portage tree.
I understand, it's too early to write version bump request usually, but old upstream is officially dead.
My appologies for too frequent annoying manipulations there.
Created attachment 767723 [details] tint2-17.1.3.ebuild There's new release (bug fixes). Previous one had some nasty bugs, introduced after 17.0.2, unfixed for long time. E.g., regression, causing sigfault with linear gradients. Just if I forgot earlier, there's changelog (plain text): https://gitlab.com/nick87720z/tint2/-/raw/master/ChangeLog
Hello, Sorry for long waiting time. I will check it out now.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=fa48f080467639fc11ac7f8feb2ddee3dadbf901 commit fa48f080467639fc11ac7f8feb2ddee3dadbf901 Author: Amy Liffey <amynka@gentoo.org> AuthorDate: 2022-12-14 11:34:50 +0000 Commit: Amy Liffey <amynka@gentoo.org> CommitDate: 2022-12-14 11:57:42 +0000 x11-misc/tint2: version bump 17.1.3 - Change upstream Bug: https://bugs.gentoo.org/833135 Submitted-by: Nikita Zlobin <nick87720z@gmail.com> Signed-off-by: Amy Liffey <amynka@gentoo.org> x11-misc/tint2/Manifest | 1 + x11-misc/tint2/metadata.xml | 1 + x11-misc/tint2/tint2-17.1.3.ebuild | 68 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 70 insertions(+)
Do we still want the live ebuild or is this enough for you? For me the new version we have now works well..
Nice, I'm glad if it's really better :) I can't tell its development is finished, so live ebuild would be not bad (but even if removed, I could just use my own from local overlay). There are still possible rare redraw issues with tasks or taskbar due to some pixmap resize misses. Also some good features still not implemented or incomplete. _____________ BTW, there's question in return: what about "CMAKE_MAKEFILE_GENERATOR = emake" ? When I tried to ensure, that it's fine with ninja generator, this broke my custom rules for autogenerated sources, using srcpp, forcing to revert to manual update (you could check how it works in git history). This would allow to exclude one redundant maintainance step - everything would be done in just one make. Imho, tint2 is just not enough big to suffer from underefficiency with amount of source files.
Oh, and could you please check, if there are missing icons? I got missing icons after attempt to upgrade system after long delay (with toolchain). Could be that my system needs full reinstall from zero. I remember, this problem exactly appeared with tray, buttons and executors (though it told like can't find or load some files in the log).
(In reply to Larry the Git Cow from comment #8) > The bug has been referenced in the following commit(s): > > https://gitweb.gentoo.org/repo/gentoo.git/commit/ > ?id=fa48f080467639fc11ac7f8feb2ddee3dadbf901 > > commit fa48f080467639fc11ac7f8feb2ddee3dadbf901 > Author: Amy Liffey <amynka@gentoo.org> > AuthorDate: 2022-12-14 11:34:50 +0000 > Commit: Amy Liffey <amynka@gentoo.org> > CommitDate: 2022-12-14 11:57:42 +0000 > > x11-misc/tint2: version bump 17.1.3 > > - Change upstream > > Bug: https://bugs.gentoo.org/833135 > Submitted-by: Nikita Zlobin <nick87720z@gmail.com> > Signed-off-by: Amy Liffey <amynka@gentoo.org> > > x11-misc/tint2/Manifest | 1 + > x11-misc/tint2/metadata.xml | 1 + > x11-misc/tint2/tint2-17.1.3.ebuild | 68 > ++++++++++++++++++++++++++++++++++++++ > 3 files changed, 70 insertions(+) Are not pkg_postinst and pkg_postrm exported from xdg.eclass, just like pkg_preinst?
(In reply to Nikita Zlobin from comment #12) > (In reply to Larry the Git Cow from comment #8) > > The bug has been referenced in the following commit(s): > > > > https://gitweb.gentoo.org/repo/gentoo.git/commit/ > > ?id=fa48f080467639fc11ac7f8feb2ddee3dadbf901 > > > > commit fa48f080467639fc11ac7f8feb2ddee3dadbf901 > > Author: Amy Liffey <amynka@gentoo.org> > > AuthorDate: 2022-12-14 11:34:50 +0000 > > Commit: Amy Liffey <amynka@gentoo.org> > > CommitDate: 2022-12-14 11:57:42 +0000 > > > > x11-misc/tint2: version bump 17.1.3 > > > > - Change upstream > > > > Bug: https://bugs.gentoo.org/833135 > > Submitted-by: Nikita Zlobin <nick87720z@gmail.com> > > Signed-off-by: Amy Liffey <amynka@gentoo.org> > > > > x11-misc/tint2/Manifest | 1 + > > x11-misc/tint2/metadata.xml | 1 + > > x11-misc/tint2/tint2-17.1.3.ebuild | 68 > > ++++++++++++++++++++++++++++++++++++++ > > 3 files changed, 70 insertions(+) > > Are not pkg_postinst and pkg_postrm exported from xdg.eclass, just like > pkg_preinst? I thats actually good point.. It seems that when i used only xdg eclass I had problems with the reseting environment not being supported for EAPI 8 but maybe I can put also xdg and xdg-utils which could fix the icons.. However maybe my environment is not the best to test because it worked for me fine..
(In reply to Nikita Zlobin from comment #11) > Oh, and could you please check, if there are missing icons? > > I got missing icons after attempt to upgrade system after long delay (with > toolchain). Could be that my system needs full reinstall from zero. > > I remember, this problem exactly appeared with tray, buttons and executors > (though it told like can't find or load some files in the log). https://gitlab.com/Zelena/hell-overlay/-/commit/676b26b808cf702970ab4f231e00a2880d8dac43 Could you maybe test this one for me? If its better with the icons?
Before test: I see no differences from my ebuild besides xdg calls in src_prepare, however there's typo - version is 17.1.3, not 17.3. Need to sort out system before upgrading portage (before some masked packages are droped, if it's not yet).
(In reply to Nikita Zlobin from comment #15) > Before test: > I see no differences from my ebuild besides xdg calls in src_prepare, > however there's typo - version is 17.1.3, not 17.3. > > Need to sort out system before upgrading portage (before some masked > packages are droped, if it's not yet). Yes, Thank you! Not considering the typo in the version (fixed now, it would show up before committing to gentoo). So the difference is some minor styling " / things and line 40 https://gitlab.com/Zelena/hell-overlay/-/commit/676b26b808cf702970ab4f231e00a2880d8dac43#a93b4c4b8ffd957c51e9052ae6d5f74e5f3fa3a1_0_40 Since the xdg_src_prepare is not for EAPI 8.
