<ukky>
farkuhar: I would be like an alien in that channel, and my opinion has no value among Crux maintainers/developers.
<ukky>
The size of dash vs bash binaries should be good enough argument for a decision which interpreter to use. Another argument that dash targets processing shell scripts and bash is targeting interactive shells.
ppetrov^ has joined #crux-social
zorz has joined #crux-social
zorz has quit [Quit: leaving]
<farkuhar>
ukky: If somebody changed the sign in #crux-devel to read "aliens welcome", would you feel less reluctant to join it and offer your thoughts?
<farkuhar>
ukky: When you get to your work PC today, see if the command history can reveal the exact sequence of commands that might have affected the /sbin/init entry in your pkg db. I'm curious how you managed to clobber a file supposedly owned by sbin-init-runit.
<SiFuh-Yamataga>
Some Bars don't accept me because I am foreigner
<SiFuh-Yamataga>
So what I have been doing is find a bar near that bar and we have a good time. I make sure they know the bar that rejected me. Just to piss that bar off
SiFuh-Yamataga has quit [Quit: Client closed]
SiFuh-Yamataga has joined #crux-social
ppetrov^ has quit [Quit: Leaving]
SiFuh-Yamataga has quit [Quit: Client closed]
<farkuhar>
Heh, opt/cdrkit hasn't seen a version bump in 15 years, and the URL points to a Wikipedia page. As for the last update of this (one-sided) summary of licensing concerns, https://cdrtools.sourceforge.net/private/linux-dist.html , the use of a two-digit year makes it hard to tell whether the date format is big-endian or little-endian.
ppetrov^ has joined #crux-social
<farkuhar>
Bad timing again. I post something related to one of ppetrov^s ports just before he joins the channel.
SiFuh-Yamataga has joined #crux-social
<SiFuh-Yamataga>
Heh
<SiFuh-Yamataga>
Also so many dead people here
<SiFuh-Yamataga>
I go walk alone in the forrest and mountains and always xome across graves
SiFuh-Yamataga has quit [Quit: Ping timeout (120 seconds)]
SiFuh-Yamagata has quit [Quit: Ping timeout (120 seconds)]
SiFuh-Yamataga has joined #crux-social
<farkuhar>
Anyway, it's unsupported to sync your core repo with the 3.8 branch and not also follow the 3.8 branch for the other repos (opt, xorg, contrib). But still, the heavy patching required for opt/cdrkit should be a strong indication that we're keeping abandonware alive.
zorz has joined #crux-social
<SiFuh-Yamataga>
Dead people
<ppetrov^>
farkuhar, about cdrkit?
<ppetrov^>
i tried to make my cdrtools coexist with cdrkit, but it did not work nicely...
<farkuhar>
ppetrov^: My point is, we should reconsider whether it's worth having cdrkit in the opt repo at all. There's no official upstream site; we just keep patching a 15-year-old version every time the toolchain gets updated. As SiFuh might say, very few people use circular spinning plastic to back up their data nowadays.
<ppetrov^>
yep
<ppetrov^>
just ditch it
<ppetrov^>
anything depending n it?
<farkuhar>
Nope
<ppetrov^>
creating an ISO file?
<ppetrov^>
no idea how it's done...
<ppetrov^>
or why not switch to cdrtools, instead?
<ppetrov^>
btw, i am impressed you remember I ever asked about cdrkit vs cdrtools
<ppetrov^>
or for sure it does not need to be installed by default
<farkuhar>
Actually, the iso Makefile does run `genisoimage`, so that will need to be rewritten if we switch to cdrtools. But the cdrtools counterpart `mkisofs` has very similar CLI syntax.
<ppetrov^>
well, if it ain't broke...
ppetrov^ has quit [Quit: Leaving]
SiFuh has quit [Remote host closed the connection]
SiFuh has joined #crux-social
<farkuhar>
ppetrov^: maybe it wasn't broke 15 years ago, when cdrkit 1.1.11 was first released. But if it targeted an unstable language spec, which modern toolchains no longer support, and distro patches aren't getting merged into an upstream repository, then you're on shakier ground to say "it ain't broke".
<farkuhar>
I'm trying to think how r0ni could run `pkgmk -is` in a port dir and still see a signature mismatch error, yet be able to bypass the signature check by editing the line in pkgmk.conf. That behaviour is consistent with a Pkgfile that itself defines IGNORE_SIGNATURE=no, because the parsing order is: 1 - command-line options, 2 - Pkgfile, 3 - pkgmk.conf
<farkuhar>
I didn't get a sense from r0ni's description that the unexpected signature check was happening with a specific port, though. If it's happening for all ports, I tend to concur with "user misconfiguration" as the diagnosis.
<ukky>
farkuhar: Crux at work is recovered by recompiling some ports with my patches, namely dhcpcd, lvm2 and xorg-server.
<ukky>
Removal of sysvinit was not at fault. dhcpcd was compiled with udev support (default in Crux), and that was the reason system had no IP, thus couldn't be ping'ed or ssh'ed into.
<ukky>
The system was bootable. Xorg server did not work either, as by default in Crux, it also required udev to function properly.
<farkuhar>
ukky: Thanks for the details on what could happen when trying to rip out the udev dependency from CRUX ports.
<farkuhar>
In other news, it looks like beerman is open to the possibility of rewriting rc scripts to target dash.
<ukky>
farkuhar: that's actually a great news. I didn't expect it to happen so easy.
<ukky>
farkuhar: btw, there is another issue that raises red flag in Crux. Both Debian and Void do not shut down 'lo' network interface upon shutdown, but Crux does. I have seen explanation somewhere, but do not recollect why they do not shutdown 'lo' interface.
<farkuhar>
As for replacing cdrkit with cdrtools (and changing system/iso/Makefile accordingly), it was fascinating to read up on the history of why the cdrkit fork happened in the first place. Apparently the Debian team included one RMS fanboy (that era's equivalent of uwumeowmeownyaa) who didn't like the license of cdrtools.