Adélie Linux issueshttps://git.adelielinux.org/groups/adelie/-/issues2019-07-11T14:01:41Zhttps://git.adelielinux.org/adelie/docs/-/issues/1Admin Handbook, Chapter 1, Table 1.2. Policies2019-07-11T14:01:41ZMichael SiegelAdmin Handbook, Chapter 1, Table 1.2. PoliciesThe release cycle for Debian is stated as:
> Release every 2 years; support for 3 years
But Debian also has an LTS program adding 2 additional years of support, with some limitations.
See: https://wiki.debian.org/LTS
The release cycle for Debian is stated as:
> Release every 2 years; support for 3 years
But Debian also has an LTS program adding 2 additional years of support, with some limitations.
See: https://wiki.debian.org/LTS
https://git.adelielinux.org/adelie/docs/-/issues/2Installation Handbook: Chapter 1: Maybe don't use `/dev/sdX` as target device...2020-05-11T13:31:54ZMichael SiegelInstallation Handbook: Chapter 1: Maybe don't use `/dev/sdX` as target device name parameter for `dd` command.I've just realized that, in the instructions on how to create a bootable installation medium on Linux or BSD, I used `/dev/sdX` as the parameter for the target device. However, `sd[[:alpha:]]` seems to be a Linux thing. On BSD, device na...I've just realized that, in the instructions on how to create a bootable installation medium on Linux or BSD, I used `/dev/sdX` as the parameter for the target device. However, `sd[[:alpha:]]` seems to be a Linux thing. On BSD, device names may look quite different. The [chapter on creating installation media](https://www.openbsd.org/faq/faq4.html#MkInsMedia) in the OpenBSD FAQ says a few things about that.https://git.adelielinux.org/adelie/adelie-base/-/issues/1adelie-base should depend on ca-certificates2019-07-11T14:01:41ZSíle Ekaterin Liszkaadelie-base should depend on ca-certificatesCA certificates are a common-enough need that they should be included in the base distribution.CA certificates are a common-enough need that they should be included in the base distribution.https://git.adelielinux.org/adelie/apkkit/-/issues/5Invocation from shell needs to determine correct DBAPI instead of assuming /2019-07-11T14:01:41ZA. WilcoxInvocation from shell needs to determine correct DBAPI instead of assuming /I *believe* this code is wrong when using overlays - currently the Adélie build system is running on my system in a chroot where master=adelie but ::gentoo is still /usr/portage so this somehow cheats the system. Unsure what would happe...I *believe* this code is wrong when using overlays - currently the Adélie build system is running on my system in a chroot where master=adelie but ::gentoo is still /usr/portage so this somehow cheats the system. Unsure what would happen if I were to merge something out of, say, the KDE or Mozilla overlay.
```py
native(os.environ, portage.db['/']['porttree'].dbapi)
```https://git.adelielinux.org/adelie/apkkit/-/issues/4Package name translation is not actually implemented yet2019-07-11T14:01:41ZA. WilcoxPackage name translation is not actually implemented yetI lost the initial prototype on my laptop somewhere. It needs to be done again at some point (before we hit conflicting package names)...
See ``_maybe_xlat`` in portage.py for information on what it is supposed to do.I lost the initial prototype on my laptop somewhere. It needs to be done again at some point (before we hit conflicting package names)...
See ``_maybe_xlat`` in portage.py for information on what it is supposed to do.https://git.adelielinux.org/adelie/apkkit/-/issues/3NEEDED is not parsed for extra dependencies2019-07-11T14:01:42ZA. WilcoxNEEDED is not parsed for extra dependenciesIt would be a very good idea to parse ``NEEDED`` or ``NEEDED.ELF.2`` for additional runtime dependencies, for a few reasons:
* ``RDEPEND`` can be outdated or just plain *wrong* sometimes.
* ``RDEPEND`` by design never contains anythi...It would be a very good idea to parse ``NEEDED`` or ``NEEDED.ELF.2`` for additional runtime dependencies, for a few reasons:
* ``RDEPEND`` can be outdated or just plain *wrong* sometimes.
* ``RDEPEND`` by design never contains anything in the ``@system`` profile or compiler runtimes, including libstdc++ and friends, which could cause issues if someone managed to build an entire system out of simple C apps then tried to install a C++ app without the runtime.
The apk-tools package upstream already seems to support ``so:foo`` where ``foo`` is a SONAME, therefore the apk side should be fairly easy. What we really should do, however, is strip SONAMEs that are provided by packages in ``RDEPEND``.https://git.adelielinux.org/adelie/apkkit/-/issues/2EAPI=2+ weak blocks are not yet supported2019-07-11T14:01:42ZA. WilcoxEAPI=2+ weak blocks are not yet supportedThe docstring in the Portage module is not accurate. ! blocks are not warned on, they are treated as fatal as !! are.
See the [PMS](https://dev.gentoo.org/~ulm/pms/head/pms.html#x1-830008.2.6) for information on the intended behaviour.The docstring in the Portage module is not accurate. ! blocks are not warned on, they are treated as fatal as !! are.
See the [PMS](https://dev.gentoo.org/~ulm/pms/head/pms.html#x1-830008.2.6) for information on the intended behaviour.https://git.adelielinux.org/adelie/apkkit/-/issues/1Dependency SLOTs are not tracked, leading to impossible resolutions2019-07-11T14:01:42ZA. WilcoxDependency SLOTs are not tracked, leading to impossible resolutionsPackage names always include SLOT if non-zero. However, when reading out RDEPEND, SLOT is not included in the name of the dependency. This means that any package that depends on a slot will have an unsatisfiable dependency. This needs...Package names always include SLOT if non-zero. However, when reading out RDEPEND, SLOT is not included in the name of the dependency. This means that any package that depends on a slot will have an unsatisfiable dependency. This needs to be resolved.https://git.adelielinux.org/adelie/apkkit/-/issues/9Rearchitect filters system2019-07-11T14:01:42ZA. WilcoxRearchitect filters systemGlobal variables are ugly. Split packages (-debug, -doc, etc) should be considered at some point, as well, which could take advantage of split packages as well.Global variables are ugly. Split packages (-debug, -doc, etc) should be considered at some point, as well, which could take advantage of split packages as well.https://git.adelielinux.org/adelie/apkkit/-/issues/6Allow passwords for private keys when signing packages2019-07-11T14:01:42ZA. WilcoxAllow passwords for private keys when signing packages``getpass`` was annoying me so I temporarily disabled it. Ideally, it would be a good idea to somehow detect when keys require passphrases, and only prompt when necessary.``getpass`` was annoying me so I temporarily disabled it. Ideally, it would be a good idea to somehow detect when keys require passphrases, and only prompt when necessary.https://git.adelielinux.org/adelie/apkkit/-/issues/7Allow hash type selection in API2019-07-11T14:01:42ZA. WilcoxAllow hash type selection in APIRight now we only sign packages with SHA-256 (Adélie requirement). As upstream is quite satisfied with the insecure SHA-1 hash and does not plan to merge our SHA-256 code in, we should find a way to handle SHA-1 hashes so that we can in...Right now we only sign packages with SHA-256 (Adélie requirement). As upstream is quite satisfied with the insecure SHA-1 hash and does not plan to merge our SHA-256 code in, we should find a way to handle SHA-1 hashes so that we can interop better with their packages.https://git.adelielinux.org/adelie/apkkit/-/issues/8Mangle data.tar.gz without abuild-tar2019-07-11T14:01:42ZA. WilcoxMangle data.tar.gz without abuild-tarRight now we are relying on upstream abuild-tar to mangle the data.tar.gz file in the way apk-tools expects.
A pure Python implementation instead of Popen is greatly desirable.Right now we are relying on upstream abuild-tar to mangle the data.tar.gz file in the way apk-tools expects.
A pure Python implementation instead of Popen is greatly desirable.https://git.adelielinux.org/adelie/packages/-/issues/15user/libwacom: install has missed the dependency on libgudev2022-02-02T17:33:59ZEmilyuser/libwacom: install has missed the dependency on libgudev| | |
| --- | --- |
| Bugzilla ID | 15 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-08-13 03:54:51 -0500 |
| Modified | 2017-11-24 22:18:14 -0600 |
| Status | RESOLVED WONTFIX |
| Version | ...| | |
| --- | --- |
| Bugzilla ID | 15 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-08-13 03:54:51 -0500 |
| Modified | 2017-11-24 22:18:14 -0600 |
| Status | RESOLVED WONTFIX |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / Intel x86 (64-bit) |
| Importance | Normal / major |
## Description
dev-libs/libwacom::gentoo correctly specifies in RDEPEND:
virtual/libgudev:=
In APKINDEX:
C:Q1NOFzjTa4jBwKzk6+tJYpqS0rMVA=
P:libwacom
V:0.18
A:x86_64
S:71522
I:716800
T:Library for identifying Wacom tablets and their model-specific features
U:http://linuxwacom.sourceforge.net/
L:
D:glib2 v:libgudev
p:so:libwacom.so.2
On a real computer where libwacom has been installed, to satisfy libinput, to satisfy kwin5:
Error relocating /usr/bin/kwin_x11: Cannot load library libgudev-1.0.so.0: No such file or directory (needed by /usr/lib/libwacom.so.2)
Installing libgudev manually fixes the issue and allows KWin to start up.---https://git.adelielinux.org/adelie/packages/-/issues/19user/sddm: graphical login screen manager2022-02-02T17:33:16ZEmilyuser/sddm: graphical login screen manager| | |
| --- | --- |
| Bugzilla ID | 19 |
| Reporter | Tambra |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-08-27 04:04:56 -0500 |
| Modified | 2018-03-10 15:45:43 -0600 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| ...| | |
| --- | --- |
| Bugzilla ID | 19 |
| Reporter | Tambra |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-08-27 04:04:56 -0500 |
| Modified | 2018-03-10 15:45:43 -0600 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / All |
| Importance | Normal / normal |
| URL | https://github.com/sddm/sddm/pull/923 |
## Description
It needs to have a graphical user login screen rather than logging in on a terminal.
I would like to have the login screen similar to the Mac OS X or Windows 7 login screens.---https://git.adelielinux.org/adelie/packages/-/issues/22user/shared-mime-info: gdk-pixbuf fails due to missing triggers2022-02-02T17:32:45ZEmilyuser/shared-mime-info: gdk-pixbuf fails due to missing triggers| | |
| --- | --- |
| Bugzilla ID | 22 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-10-29 20:55:58 -0500 |
| Modified | 2017-08-23 02:15:45 -0500 |
| Status | RESOLVED FIXED |
| Version | 1....| | |
| --- | --- |
| Bugzilla ID | 22 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-10-29 20:55:58 -0500 |
| Modified | 2017-08-23 02:15:45 -0500 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / All |
| Importance | Normal / normal |
| URL | https://roidelapluie.be/wiki/couldnt-recognize-the-image-file-format-for-file-foopng.html |
## Description
Symptoms: things like XSANE and Firefox fail to start or display images: "Couldn't recognize the image file format"
gdk-pixbuf: trigger for /usr/lib/gdk-pixbuf-2.0 needs to run:
gdk-pixbuf-query-loaders > /usr/lib64/gdk-pixbuf-2.0/2.10.0/loaders.cache
shared-mime-info: trigger for /usr/share/mime needs to run:
update-mime-database /usr/share/mime1.0-BETA1https://git.adelielinux.org/adelie/packages/-/issues/23user/evince: (possibly other GNOME desktop) won't start due to missing glib t...2022-02-02T17:32:31ZEmilyuser/evince: (possibly other GNOME desktop) won't start due to missing glib trigger| | |
| --- | --- |
| Bugzilla ID | 23 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-10-29 21:09:33 -0500 |
| Modified | 2017-08-23 02:16:50 -0500 |
| Status | RESOLVED FIXED |
| Version | 1....| | |
| --- | --- |
| Bugzilla ID | 23 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2016-10-29 21:09:33 -0500 |
| Modified | 2017-08-23 02:16:50 -0500 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / All |
| Importance | Normal / normal |
## Description
glib2 trigger on /usr/share/glib-2.0/schemas should run:
glib-compile-schemas --allow-any-name /usr/share/glib-2.0/schemas1.0-BETA1https://git.adelielinux.org/adelie/packages/-/issues/26user/vlc: need trigger for plugin cache2022-02-02T17:31:01ZEmilyuser/vlc: need trigger for plugin cache| | |
| --- | --- |
| Bugzilla ID | 26 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-01-23 16:38:46 -0600 |
| Modified | 2017-11-24 22:40:18 -0600 |
| Status | RESOLVED FIXED |
| Version | 1....| | |
| --- | --- |
| Bugzilla ID | 26 |
| Reporter | A. Wilcox (awilfox) |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-01-23 16:38:46 -0600 |
| Modified | 2017-11-24 22:40:18 -0600 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / All |
| Importance | Normal / normal |
## Description
* Running /usr/lib/vlc/vlc-cache-gen on /usr/lib/vlc/plugins/---https://git.adelielinux.org/adelie/packages/-/issues/27system/gcc: problems getting GCC to work2022-02-02T17:30:38ZEmilysystem/gcc: problems getting GCC to work| | |
| --- | --- |
| Bugzilla ID | 27 |
| Reporter | Samuel Holland |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-01 13:59:11 -0600 |
| Modified | 2017-11-24 22:44:59 -0600 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALP...| | |
| --- | --- |
| Bugzilla ID | 27 |
| Reporter | Samuel Holland |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-01 13:59:11 -0600 |
| Modified | 2017-11-24 22:44:59 -0600 |
| Status | RESOLVED FIXED |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / All |
| Importance | Normal / normal |
## Description
I installed gcc5.4.0, hoping to get a working compiler. Here are the issues I faced:
* No GCC profile is set by default, meaning there was no cc or gcc symlink. This can be set with gcc-config.
* gcc-config refused to run because it could not determine the CHOST. I got around this by exporting CHOST on the command line based on the output of `gcc-5.4.0 -v`
* After that, I noticed that gcc failed when it tried to run `as` because binutils also has no default version set.
* binutils-config cannot list possible profiles because CHOST is not set (and portageq is not found when it tries to auto-detect CHOST).
* Once that is fixed, gcc cannot find libgcc. This is because libgcc.a is in the gcc5.4.0-devel package, which gcc5.4.0 does not depend on.
* Also, musl-devel is not pulled in by any of these packages; I'm not sure if that is intended.---https://git.adelielinux.org/adelie/packages/-/issues/28eselect is broken2021-05-12T03:12:35ZEmilyeselect is broken| | |
| --- | --- |
| Bugzilla ID | 28 |
| Reporter | Samuel Holland |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-01 14:04:26 -0600 |
| Modified | 2017-08-23 02:04:02 -0500 |
| Status | RESOLVED WONTFIX |
| Version | 1.0-A...| | |
| --- | --- |
| Bugzilla ID | 28 |
| Reporter | Samuel Holland |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-01 14:04:26 -0600 |
| Modified | 2017-08-23 02:04:02 -0500 |
| Status | RESOLVED WONTFIX |
| Version | 1.0-ALPHA1 |
| Hardware | Adélie Linux / Intel x86 (64-bit) |
| Importance | Normal / normal |
## Description
Any invocation of eselect that loads a module and enumerates its actions (e.g. `eselect vi`) fails with the error message:
/usr/share/eselect/libs/core.bash: line 124: /root/portage/bin/ebuild-helpers/bsd/sed: No such file or directory
This seems to be leftover from building the package, as neither that path nor anything like it is present on my system.---https://git.adelielinux.org/adelie/packages/-/issues/30user/clang: emits broken binaries2022-02-02T17:29:30ZEmilyuser/clang: emits broken binaries| | |
| --- | --- |
| Bugzilla ID | 30 |
| Reporter | Kiyoshi Aman |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-25 12:32:38 -0600 |
| Modified | 2017-11-24 22:46:46 -0600 |
| Status | RESOLVED WONTFIX |
| Version | 1.0-ALP...| | |
| --- | --- |
| Bugzilla ID | 30 |
| Reporter | Kiyoshi Aman |
| Assignee | A. Wilcox (awilfox) |
| Reported | 2017-02-25 12:32:38 -0600 |
| Modified | 2017-11-24 22:46:46 -0600 |
| Status | RESOLVED WONTFIX |
| Version | 1.0-ALPHA2 |
| Hardware | Adélie Linux / Intel x86 (64-bit) |
| Importance | Normal / normal |
## Description
clang on x86_64 presently builds binaries with an incorrect interpreter inserted. This was originally discovered when building a standalone libblocksruntime, as the testsuite provided failed due to the binaries produced failing.---