08:26 <pitillo> thanks chrcav! here steam is working like a charm again with that glibc-32 patch

09:52 <beerman> i will push that in a bit

09:57 <beerman> i really dislike the way we sideload man pages into opt/btrfs-progs, just to please some people and their weird demands..

10:00 <beerman> i will see to enable python3-sphinx optionally and drop the man pages tar file. if jue or jaeger care enough its easy to revert that change but prebuilding a manpages tar file..? c'mon..

10:14 <jue> beerman: what's so evil with that?

10:15 <beerman> how am i supposed to update that?

10:15 <beerman> its a crux-core tagged port in opt, but it requires me to do a number of extra steps to prepare an update.

10:16 <jue> well, it's a bit manual work, of course

10:16 <beerman> take a look in a bit when i push my changes and see if that works for you as well

10:16 <jue> well, no problem at all, do what you want

10:17 <beerman> you are free to revert it but i won't pregenerate a manual tar to sideload it

10:18 <beerman> especially for stenur who skips no opportunity to flame against me, even if he is wrong ;)

10:18 <beerman> its an upstream decision, really

10:20 <beerman> https://crux.nu/gitweb/?p=ports/opt.git;a=commitdiff;h=392cf1dd110c3ea87b69dc3fddee283c11b083f2

10:20 <beerman> there you go

10:23 <jue> as I said, no problem at all, installing python3-sphinx and all the deps is not a option for me, though

10:23 <beerman> not for me either

10:25 <beerman> but i am not doing those extra hoops romster did, look where he is, burned out from that stuff he did because he wanted to be nice (imo). see rust-bin for example. Nobody cared for rust-bin once romster stopped providing it.

10:26 <jue> tbh, I've stolen the man-pages from arch ;)

10:26 <beerman> lol

10:27 <beerman> it would still require me to fix up my prtup script I use to build things in a container, to pull their package, and repack the manpages. Also makes us dependent on whenever they update their package.

10:29 <beerman> this is a source based distro after all and this way pleases both fractions of people that a) don't want sphinx and can live without man pages b) people that want man pages and can live with sphinx.

10:29 <beerman> those that want things without being able to live with a compromise need to look for a solution to their problem

10:30 <beerman> bark at upstream would be one way to go

10:30 <beerman> but at least now you jaeger and myself can just update btrfs-progs without wasting a lot of time on it

10:35 <beerman> and in the end: feel free to revert it, but the way it was I feel like it falls behind the crux-core ports line, and you would probably need to take care of it alone, as I wouldn't touch it anymore

10:48 <beerman> jue: compromise idea: if we had a script.. let's say pre-maintainter-run, which would generate that man files tar for us and cleans up after itself, i would incorporate that in my script and build it. even though I think thats.. lets say silly.

10:49 <beerman> we already have post maintainer scripts (rust, gtk4, grafana) to clean offending footprint files for example

10:53 <jue> no, let's keep it simple, the man-pages are not really important for me

10:54 <beerman> agreed

10:54 <beerman> btw

10:54 <beerman> i assigned that qemu headless ticket to you, since you own opt/qemu

10:54 <beerman> https://crux.nu/bugs/index.php?do=details&task_id=1924&project=1 no idea if you would want to do that

10:55 <beerman> i don't use qemu at all so i can't speak for or against it

10:55 <jue> yeah, saw it

10:56 <beerman> ok, thanks :)

10:58 <jue> renaming libpixman is a bit problematic cause we have to adjust all dependent, more a thing for our next release at a first thought

10:59 <beerman> true

10:59 <beerman> we wouldn't really need to wait too long for that either

11:00 <jue> yep, that's true

11:00 <beerman> i suppose, at least. 3.7 just took way too long :)

11:01 <beerman> i could see a 3.7.1 where we upate python3 to 3.11 for example

11:01 <beerman> and maybe any other minor changes we would like to see, if any

11:05 <beerman> libsdl2 fails with ccache, currently. [ -e '/usr/bin/ccache' ] && PATH="$(echo ${PATH} | awk -v RS=: -v ORS=: '/ccache/ {next} {print}' | sed 's/:*$//')" this line fixes it, as it fails due to calling itself recursively

11:07 <jue> hmm, works for me, with ccache

11:07 <beerman> mh

11:07 <beerman> did you set it up differently to the wiki page romster did a long time ago?

11:10 <jue> think no, /usr/lib/ccache is the first entry im my $PATH

11:10 <beerman> did you alter that in your profile?

11:11 <beerman> but that wouldn't matter, no idea.. https://dpaste.com/FTEFK6Q7L this is the log on my desktop where it is the same as inside my container

11:11 <beerman> ah, and its distcc and ccache, not ccache alone

11:11 <beerman> my bad ;)

11:12 <jue> ok

11:13 <jue> have to run now, later

11:16 <beerman> laters, would be nice if you could add the line later on

11:17 <beerman> export CCACHE_PREFIX='distcc' thats what I have in my pkgmk.conf needed to reproduce that error

18:14 <jaeger> I don't know how difficult it would be to build man pages for btrfs-progs but for docker I build a new set with every release... at least they make it easy

18:14 <jaeger> If there's a desire to have the man pages I could probably set up a container for building them for btrfs-progs

18:15 <jaeger> I don't have a strong feeling about it. Though I do enjoy the convenience of having a local man page I can always go look it up online if needed

18:20 <jaeger> Unrelated, going to see if I can get a plasma desktop wayland session going with tsaop's kf5 repo

18:29 <beerman> docker would not really be suited for my lxc container and would still add complexity to at least my workflow. it's not the challenge to generate a man-tar but the complexity is what I reject

18:31 <beerman> i would have to make changes to my prtup script, I already catch ports like libreoffice which will trigger a rebuild of harfbuzz and xmlsec before building lo for example. My lo Pkgfile has catches for misconfigured harfbuzz/xmlsec and quits the build early, as configure would already fail to run

18:31 <beerman> plasma wayland session should be no problem, sounds like a fun thing to do :)

18:31 <jaeger> Like I said, not a strong feeling about it. If you don't want to mess with it, that's fine

18:32 <jaeger> I like Plasma *far* better than any of the old KDE stuff... just much smoother and nicer

18:32 <beerman> well, messing with it.. it would require building btrfs-progs with sphinx installed, then package those manfiles and saving the artifact in the port directory, then rebuilding the port without sphinx installed and sideloading the man-tar

18:33 <beerman> it's not impossible but I am not keen in providing that either ;)

18:34 <beerman> yep, agreed, Plasma looks very neat. I only set it up for others, no real experience in day to day use. I tried it on the pinephone but its far to ressource hungry for that :)

18:37 <jaeger> I've used it a few times on arch installs but hadn't messed with it on crux yet

19:40 <stenur> oh there was a bug tracker instance of btrfs-progs. i had forgotten. Thanks again jue@ for adding the manual.

19:41 <stenur> Yes i asked on IRC but left immediately again. I do not know.

19:41 <stenur> (Asked whether they would provide a manual pages ball like git does.)

20:06 <jaeger> ok

21:35 <stenur> I mean, free some logs of that magic CRUX home server (our syslog can auto-rotate, also based on size. it is great), and install a qemu or i-do-not-know-what with all the software needed to create documentation of core and opt tools which stopped producing manuals. Cool!

22:06 <stenur> The fun thing is that i now commit an ipcalc update, where those guys converted a very old UNIX manual to "ronn" syntax, and did it wrong! So now in 1.0.2 you get "updated manual to comply with ronn syntax". Hey RedHat guys, 30bln bucks made you famous! Huhu.