01:31 <farkuhar> braewoods: no worries, the symlink bug has been around a long time, it can wait another decade if users are okay with the occassional 'pkgadd -u -f'. I was just asking in case there was a symlink-friendly implementation of db_find_conflicts (pkgutil.cc starting at line 275) somewhere in your codebase.

01:40 <braewoods> I would have to dig in depth to find out what the issue is.

02:09 <farkuhar> As a starting point, I might recompile pkgadd without CXXFLAGS += -DNDEBUG, and then run pkgadd -u on a package with known symlink issues (like linux-firmware). That would reveal which phase failed to remove the symlink from the "conflicting files" list.

02:50 <farkuhar> The pkgadd debugging output was less helpful than I had hoped. I only learned that after pruning the "conflicting files" list, the symlink lib/firmware/qcom/LENOVO/21BX was not recognized as belonging to linux-firmware. But 'pkginfo -o' does report linux-firmware (20220815) as the owner of that path, it's just a *directory* (not a symlink) in the old version.

12:40 <cruxbot> [contrib.git/3.6]: firefox: 104.0 -> 104.0.2

12:40 <cruxbot> [contrib.git/3.6]: [notify] thunderbird: 91.10.0 -> 102.2.2 new dependency: wasi-libc++

12:40 <cruxbot> [contrib.git/3.6]: cuda: 11.7.0 -> 11.7.1

12:40 <cruxbot> [contrib.git/3.6]: libaom: initial import

12:40 <cruxbot> [contrib.git/3.6]: wasi-libc++: 13.0.1 -> 14.0.6

12:40 <cruxbot> [contrib.git/3.6]: wasi-libc: 232.3eb02a7 -> 0.20220525

12:40 <cruxbot> [contrib.git/3.6]: multimc: 0.6.13 -> 0.6.16

12:40 <cruxbot> [contrib.git/3.6]: firefox: 102.0 -> 104.0

12:40 <cruxbot> [contrib.git/3.6]: cuda: 11.6.2 -> 11.7.0

13:44 <SiFuh_> `I always wonder if farkuhar's handle should end in the letter 'd'

14:01 <cruxbot> [xorg.git/3.6]: [notify] libglvnd: Update the GL headers and XML files from the Khronos repository, remove legacy header that is not referenced in the Khronos headers anymore

14:01 <cruxbot> [xorg.git/3.6]: [notify] xorg-libx11: disable a new option that makes X11 threadsafe as this breaks a ton of things, move a legacy header (XKBgeom.h)out of xorg-xorgproto to xorg-libx11

14:01 <cruxbot> [xorg.git/3.6]: [notify] xorg-libxvmc: vldXvMC.h is now in xorg-libxvmc where it belongs

14:01 <cruxbot> [xorg.git/3.6]: [notify] xorg-xorgproto: move two header files to there respective ports, update this before xorg-libxvmc and xorg-libx11 to avoid file conflicts

14:03 <stenur> romster is fit enough for commits!

14:11 <SiFuh_> I hope so

15:18 <stenur> can anyone use mpg123 with "-o alsa" (default here, too)? it only works with "-t oss"

15:19 <stenur> i get: main: [src/mpg123.c:play_frame():850] error: Deep trouble! Cannot flush to my output anymore!

15:19 <stenur> -o sdl also works fine

15:21 <SiFuh_> stenur: mpg123 -o alsa Smokin\ Indoors\ -\ Sunny\ Ledfurd.mp3 works fine

15:22 <stenur> on 3.7? mysterious.

15:22 <SiFuh_> 3.6 sorry

15:22 <stenur> ..yes.. but mysterious

15:22 <stenur> normally all ogg here, or faad -> sox, all via ALSA, all ok

15:23 <stenur> playing that .mp2 via sox also ok, but mpg123 only via oss not alsa

15:23 <stenur> Thanks SiFuh

15:23 <stenur> Thanks SiFuh_

15:25 <stenur> Ah! what works with alsa is "src/mpg123 -o alsa -a xmix"!

15:26 <stenur> why does it not default to the default device? hm

15:27 <stenur> "-o alsa -o sysdefault" also works

15:27 <stenur> poooh what a mess!

15:28 <stenur> and better use sysdefault otherwise it says "Warning: You triggered the NtoM drop-sample resampler inside libmpg123.

15:28 <stenur> Warning: You could trade CPU for quality by forcing a supported output rate.

15:29 <stenur> though xmix is the sysdefault hmhm; well at least via asound.conf, but mpg123 uses libasound aka alsa-lib no?

15:32 <stenur> and what if not; well yes, via /usr/lib/mpg123/output_alsa.so; that somehow does not care

15:34 <stenur> maybe a bug in mpg123, they do "ao->device ? ao->device : "default"" but want sysdefault in src/libout123/modules/alsa.c

15:35 <stenur> well it is a nice template shall i ever program ALSA, but somehow they do it wrong so asound.conf is not parsed aka used

15:36 <stenur> whatever; i do not even know why i have mpg123 installed, initially

15:37 <stenur> i am only trying to convert an mp2 to ogg

15:38 <stenur> sox can mp3 and mp2 (thus as i learned)

15:38 <farkuhar> lossy format to lossy format? do that enough times and the audio will be seriously degraded.

15:38 <stenur> (..but want to do that regulary on the server now, and therefore trying to use mpg123 for the task..)

15:39 <stenur> it is a MP2 radio stream that is 56kbps 22.05 kHz initially

15:39 <stenur> but 7 hours are ~170MB, whereas in MP4 that is

15:40 <stenur> ~110MB (iirc), but usually i am OGG-only since at least twenty years or so, so that

16:33 <stenur> i opened two bugs at mpg123 let's see; maybe i am doing something obvious wrong. I use sox on the server.

16:35 <SiFuh_> But you are the bug :-P

16:35 <SiFuh_> Haha

17:34 <cruxbot> [contrib.git/3.6]: bmake: 20220909