This project forked from kisslinux/repo
KISS Linux - Unofficial Repositories, mirror of https://codeberg.org/kiss-community/repo
Home Page: https://kisslinux.org
License: MIT License
|/ |\ISS LINUX https://kisslinux.github.io ________________________________________________________________________________ Unofficial Main Repository ________________________________________________________________________________ The officially unofficial repository for KISS Linux maintained by kiss-community More information: - https://kisslinux.github.io - https://kisslinux.github.io/package-system - https://kisscommunity.org Enabling SSH signature verification (recommended): ________________________________________________________________________________ $ cd /path/to/repo $ git config gpg.ssh.allowedSignersFile .allowed_signers $ git config merge.verifySignatures true
Instead of just comparing strings, if I understand correctly it's better to just check for the min version.
So check something like is gcc older than 4.8, only then do bootstrap. (that scenario is unlikely on kiss
but you never know)
I can't believe how much electricity and time has been wasted doing this bootstrap nonsense
This will massively speed up compile times as currently it's doing redundant work as shown here:
echawk/kiss-xorg#43
The install of the baseinit package fails if there's any CFLAGS defined. The error log is one line long and only contains the error message mentioned in the title of this issue.
Need to update kiss to use git versions here:
https://github.com/kiss-community/repo-main/blob/master/core/kiss/sources
And maybe update the build also.
Not sure on the best way to handle stuffs like website and wiki but for kiss itself maybe just
git+https://github.com/kiss-community/kiss.git
be the most convenient ?
Trying to build xfsprogs without meson results in an error of meson isn't found
Not sure if anyone else is interested in this change, 10-15 packages from community will need to be tested and a few in main repo
Hello,
Gtk+3 fails to build with the landed remove-atk patch.
http://ix.io/413Y
DESTDIR seems mostly useless to me, but VERSION and other markers have a valid use case shown here in kiss-xorg: https://github.com/ehawkvu/kiss-xorg/tree/master/xorg/libXmeta
It can also be helpful to prevent patch failures if someone is using patch files to add something to sources and build without duplicating the whole package
core:
extra:
wayland:
For the 2.35.2 release of git, there were some internal changes made to mitigate security vulnerabilties.[1]
For us this means that if someone has their kiss repositories owned by another user (let's say root), and they run kiss u the repositories will fail to sync.
For the time being, I have the following in ~/.gitconfig to mitigate this problem:
[safe]
directory = *Otherwise, when trying to access git repositories, you get the following error:
[ethan ~] % git -C /var/db/kiss/repos/repo/core rev-parse --show-superproject-working-tree
fatal: unsafe repository ('/var/db/kiss/repos/repo' is owned by someone else)
To add an exception for this directory, call:
git config --global --add safe.directory /var/db/kiss/repos/repo [1] https://github.blog/2022-04-12-git-security-vulnerability-announced/
[ 2.228958] alsactl[255]: segfault at c ip 00007f3794a72fe7 sp 00007ffe44aaf7a0 error 4 in libasound.so.2.0.0[7f3794a6d000+93000]
[ 2.228964] Code: 00 00 00 5b b8 ef ff ff ff 5d 41 5c 41 5d 41 5e c3 b8 ea ff ff ff eb db 66 66 2e 0f 1f 84 00 00 00 00 00 41 54 55 53 48 89 fd <8b> 47 0c 85
c0 7f 72 8b 47 08 83 f8 03 74 7a 3d 00 04 00 00 75 33
[ 2.229010] udevd[240]: '/usr/bin/alsactl restore 0' [255] terminated by signal 11 (Segmentation fault)
Hello recent alsactl update broke the restore function... Same segfault for every device is experienced on my system hw: 0-2
When building firefox, I encounter
ERROR: Package alsa was not found in the pkg-config search path.
ERROR: Perhaps you should add the directory containing `alsa.pc'
ERROR: to the PKG_CONFIG_PATH environment variable
ERROR: Package 'alsa', required by 'virtual:world', not found
I installed alsa-lib and the build succeeded. Should this be a make dependency of firefox?
# kiss owns /usr/lib/pkgconfig/alsa.pc
alsa-lib
I don't think anything statically links ffmpeg, so static libs waste 100+ MB of space:
2MB /usr/lib/libswscale.a
508KB /usr/lib/libswresample.a
288KB /usr/lib/libpostproc.a
2MB /usr/lib/libavutil.a
19MB /usr/lib/libavformat.a
18MB /usr/lib/libavfilter.a
736KB /usr/lib/libavdevice.a
65MB /usr/lib/libavcodec.a
132MB total (full package, including shared libs)
I hope we will have some news from @dylanaraps soon, but I expect it won't happens before many weeks.
What do you think about adding some repo's to repology ?
I just realized I won't get any notification for new releases of softwares as soon as a software is outdated on https://github.com/kisslinux/community. It's the reason why I created a PR with an outdated software (PR kiss-community/community#15). So I have to think to a solution to get notification of new releases without the notification from repology for packages on https://github.com/kisslinux/community.
Since the last update, libva tries to pull in wayland which is not in the repo and as I understand the guidestones must not be in the official repo.
Hello @git-bruh and @dilyn-corner I've got a rather philosophical issue here.
But first, i'd like to point out that on github the kiss package manager repository does not have a issues page like this repo-main
repository which is why I am writing it here. And yes I am too lazy to go onto IRC, so deal with it.
I have some repos paths being set to appended to KISS_PATH here.
https://github.com/kyx0r/klec/blob/685c860892ae81c3e009576ce36bb8c633d2e9b5/dotfiles/.profile#L6
As you can see, I am forced to check for every directory because if I don't when I run kiss update I get error message like this one:
/usr/bin/kiss: cd: line 1307: can't cd to /root/kiss/gcommunity/community: No such file or directory
The problem is that I use my configs on many different machines and not every machine has the same repos and requirements. And this error is not handled by the package manager which causes the update to fail. The problem can be solved by simply adding a check inside kiss on the return error code of the cd command.
I hope it's not too big of an ask to add a check + continue in kiss so that I don't have to keep adding the monstrocity of if statements in my .profile . I might as well be beneficial to you also. But again you have to sacrifice 2 lines of code for that, which may or may not surpass the said minimalism standard.
Add a provides system that allows a single package to act as a provider for multiple packages
Eg, zig needs llvm built with all targets so
zig would depend on llvm-fat
llvm-fat would be a provider of llvm as well as llvm-fat
Two packages that provide atleast one common package should be rejected during installation instead of making alternatives
Hi finally I was able to get caught up on all the kiss news this month, I was trying to subscribe myself to the list but I get no confimation email from the server. I emailed [email protected] directly and [email protected] with the keywords.
When trying to install wlroots I get an error saying I do not have a gnu c++ compiler. How do I fix this?
build fails with
13:33.14 xpcom/build/Services.cpp.stub
13:33.60 xpcom/components
13:35.00 /tmp/32170/build/firefox/toolkit/xre/nsEmbedFunctions.cpp: In function 'nsresult XRE_InitChildProcess(int, char**, const XREChildData*)':
13:35.00 /tmp/32170/build/firefox/toolkit/xre/nsEmbedFunctions.cpp:530:3: error: 'XInitThreads' was not declared in this scope
13:35.00 530 | XInitThreads();
13:35.00 | ^~~~~~~~~~~~
13:35.23 xpcom/ds
13:35.93 make[4]: *** [/tmp/32170/build/firefox/config/rules.mk:676: nsEmbedFunctions.o] Error 1
13:35.93 make[3]: *** [/tmp/32170/build/firefox/config/recurse.mk:72: toolkit/xre/target-objects] Error 2
13:35.93 make[3]: *** Waiting for unfinished jobs....
13:51.67 Compiling webrender v0.61.0 (/tmp/32170/build/firefox/gfx/wr/webrender)
14:00.01 Compiling webrender_bindings v0.1.0 (/tmp/32170/build/firefox/gfx/webrender_bindings)
16:11.69 Compiling gkrust v0.1.0 (/tmp/32170/build/firefox/toolkit/library/rust)
19:58.48 Finished release [optimized] target(s) in 19m 39s
19:58.98 make[2]: *** [/tmp/32170/build/firefox/config/recurse.mk:34: compile] Error 2
19:58.98 make[1]: *** [/tmp/32170/build/firefox/config/rules.mk:355: default] Error 2
19:58.98 make: *** [client.mk:89: build] Error 2
19:58.98 0 compiler warnings present.
X11/Xlib.h and -lX11 related
I have wrote about it on valgrind mailing list:
https://sourceforge.net/p/valgrind/mailman/message/37269403/
Also can I add valgrind into community repo, I can maintain it. But first we need to solve this weirdness.
2021-02-17 -- At this point, it is safe to say that Dylan Araps has left the project fairly permanently. We have heard no public message from him in almost 3 months, the domain has expired, and the SSL certificate on git.k1ss.org has expired. While kiss-community has maintained packages, the current mindset of "community maintenance until Dylan returns" seems to be coming less and less realistic every day. Please, make your voice heard and fill out this anonymous, 2 question form. It would be prudent to reread the guidestones of KISS (https://kiss.armaanb.net/guidestones) before continuing.
Note that a BDFL as chosen here may not require the same skills / presence that Dylan had. The BDFL's job would be to maintain the vision of the distro, make executive decisions, manage the repositories, and delegate tasks which they may not be capable of doing themselves. It is also important to note that if Dylan is to return to the project, it is expected that he be amicably offered the role of BDFL once again.
This form will expire on 2021-02-24 at 11:55PM UTC.
This option is deprecated and no longer needed.
As well as https://k1sslinux.org/wiki/dev/replacing-udev outdated too.
Error returned is wget: error getting response: Connection reset by peer, but it works fine if run with the option --no-check-certificate
It's possible that mdev might start after dhcpcd, so dhcpcd stops looking for new interfaces altogether and doesn't output anything after this:
no such user dhcpcd
dhcpcd-9.4.1 starting
dev: loaded udev
DUID 00:01:00:01:2a:93:78:3a:d0:37:45:47:f3:16
no valid interfaces found
no interfaces have a carrier
Restarting dhcpcd makes it work properly till shutdown:
no such user dhcpcd
dhcpcd-9.4.1 starting
dev: loaded udev
DUID 00:01:00:01:2a:93:78:3a:d0:37:45:47:f3:16
no interfaces have a carrier
wlan0: waiting for carrier
wlan0: carrier acquired
wlan0: IAID 45:47:f3:16
wlan0: rebinding lease of 192.168.29.156
wlan0: probing address 192.168.29.156/24
wlan0: leased 192.168.29.156 for 3599 seconds
wlan0: adding route to 192.168.29.0/24
wlan0: adding default route via 192.168.29.227
This dependency can be removed with this patch. It's been submitted upstream.
yash >> ssu kiss-chroot kiss_rootfs > /dev/null
Illegal instruction
I guess the rootfs has been compiled with -march=native or something like that.
It's been 9 month for gcc, and I am tired of compiler bugs at different optimization levels. Maybe it's time to update it because it has probably already got a lot of bugs fixed hopefully.
I have some garbage drivers that need to be compiled as a module, otherwise kernel is not able to find the right firmware. I have a script that loads all the modules like so:
var=$(cat /lib/modules/5.11.0_1/modules.order)
for p in $var
do
modprobe $(basename "$p")
done
The problem is if this is put into /etc/rc.conf this script will also run when doing poweroff, which isn't right. (But it does not cause problems usually) If I put this into /etc/rc.d/mod.boot instead the problem of being run on anything but boot will be solved, however the device manager will start before the kernel modules are loaded, which will cause for example sound to not work, or rather be muted, since device manager did not find any sound cards at boot. Is there any better way to handle this crap?
I tried updating my older kiss linux install on one of my rigs. Currently it has llvm 11.0.0 1 installed. During the build it failed and the kernel reports:
[ 7888.231917] traps: llvm-tblgen[1466] general protection fault ip:7f00ffe945c1 sp:7ffdf44efdd8 error:0 in libc.so[7f00ffe33000+76000]
I am attaching relevant part of the log here:
https://pastebin.com/embed_js/AZTbzztT
Based on the recent git history I see that llvm has been updated, so I need someone to verify if they can reproduce the crash or not.
Despite this rig being pretty old, it still have the latest gcc version gcc 10.2.0 1 so we should be getting the same codegen on x86_64 platform i suppose.
-> Resolving dependencies
-> Building: cmake
-> cmake Checking repository files
-> Checking for pre-built dependencies
-> cmake Downloading sources
-> cmake Found cached source 'cmake-3.20.2.tar.gz'
-> cmake Found local relative source 'patches/cmake-no-execinfo.patch'
-> cmake Checksum mismatch
ERROR Checksum mismatch with: cmake
Let me know if more info required.
The repo on sr.ht is now considered the primary repo, GitHub is a mirror.
@git-bruh latest version is 0.5.1 compared to 0.4.8
In order to build Zig, it is required that LLVM is built with all default targets enabled. Is it possible that the LLVM package build script could be changed to do this?
Hi @git-bruh
I've just tried to update e2fsprogs several times and it's failing every time:
LD e2fsck
/usr/bin/ld: rehash.o: in function `e2fsck_rehash_dir':
rehash.c:(.text+0xebd): undefined reference to `qsort_r'
/usr/bin/ld: rehash.c:(.text+0x135c): undefined reference to `qsort_r'
collect2: error: ld returned 1 exit status
make[2]: *** [Makefile:453: e2fsck] Error 1
make[2]: Leaving directory '/tmp/build-383859/e2fsprogs/e2fsck'
make[1]: *** [Makefile:443: all-progs-recursive] Error 1
make[1]: Leaving directory '/tmp/build-383859/e2fsprogs'
make: *** [Makefile:371: all] Error 2
-> e2fsprogs Build failed
xinit's startx uses non-portable grep flags (-o) on line 109
So proposing we change it from
tty_num=$(echo "$tty" | grep -oE '[0-9]+$')
to
tty_num=$(tty | cut -c9)
Which could be done via
sed -i "s/tty_num=.*/tty_num=\$(tty | cut -c9)/g" startx.cpp
When running busybox powertop (as root), I see powertop: no stats available; run as root or enable the timer_stats module. Looking at powertop.c, it looks like it uses /proc/timer_stats , but looking at man 5 proc (http://manpages.ubuntu.com/manpages/bionic/man5/proc.5.html), it looks like this was removed in linux 4.10.
I personally am not excited about a busybox version of powertop; I would propose disabling CONFIG_POWERTOP in the busybox config.
When I extract rootfs as root, I get https://0x0.st/oSiA.txt. The --no-same-owner flag helps.
Also, when I try to build the package, I get:
fatal: unable to access 'https://github.com/vim/vim/': error setting certificate file: /etc/ssl/certs/ca-certificates.crt
Currently all runit services just spam the tty on boot, and logs aren't stored anywhere. I use the following script to enable logging on my server:
#!/usr/bin/env sh
LOG_BASE="${1:-/var/log/sv}"
# /etc/sv/$sv_name/log
sv_name="$(basename "$(realpath "$PWD/..")")"
logdir="$LOG_BASE/$sv_name"
mkdir -p "$logdir"
exec svlogd -tt "$logdir"Enabled per-package with another script:
#!/usr/bin/env sh
set -eu
SV="$1"
SVBASE="/etc/sv"
RUNBASE="/run/runit"
LOGDIR="$SVBASE/$SV/log"
[ ! "$SV" ] && exit 1
[ -d "$SVBASE/$SV" ] || exit 1
mkdir -p "$LOGDIR"
ln -sf "$SVBASE/logger" "$LOGDIR/run"
ln -sf "$RUNBASE/supervise.$SV-logger" "$LOGDIR/supervise"Should we include something like this in the default busybox package or is it better suited for community?
Unsure if logs are needed for this, but I'm getting a checksum mismatch with the new cbindgen v0.17.0 update.
According to Linux Kernel Documentation:
Kernel headers are backwards compatible, but not forwards compatible. This means that a program built against a C library using older kernel headers should run on a newer kernel (although it may not have access to new features), but a program built against newer kernel headers may not work on an older kernel.
Basically if you have linux-headers 5.17 installed (which is the current version in the repository) and have an LTS 5.15.x kernel (like myself), if you recompile libc and try compiling new programs, they might not work.
As Dylan put it:
The kernel headers in KISS are pinned to an LTS kernel version and are only updated when there are changes of interest in the kernel or headers themselves.
As I'm not aware of a good reason to update the headers to the stable branch, I think It'd be a good idea to keep the linux headers at the latest LTS (currently 5.15.41), or the previous LTS (5.10.117), as they're recent enough to support most new features, but stable enough to not break anyone's system.
I need to use seatd to load .so files is it anywhere in the repos?
checking whether gcc accepts -Wnull-dereference... yes
checking whether gcc accepts -Wimplicit-fallthrough... yes
checking for library containing argp_parse... no
configure: error: in `/root/.cache/kiss/proc/2371/build/libelf':
configure: error: failed to find argp_parse
See `config.log' for more details
-> libelf Build failed
Looks like the no argp_parse option does not work anymore.
core:
extra:
xorg:
It tries to download libx11 from the alpine repos, but an older version, 1.7.3. It has now been updated to 1.7.5: https://pkgs.alpinelinux.org/package/edge/main/x86_64/libx11
On the website and in kiss help install the link is:
$ url=https://github.com/kiss-community/repo/releases/download/2021.3-1
but that isn't right. when I download it and extract every folder is empty and symlinks are broken.
Make sure to update those things when new release, also delete the broken 3-1 release as there is no point in keeping it.
@dilyn-corner
Packages can individually install rules in the mdev.conf syntax to /etc/mdev.conf.d/*, which will be amalgamated and passed to the mdev implementation
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.