Linux: is the machine running fedora or just the user

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > Linux: is the machine running fedora or just the user

«‹›4 »

shackleford: Sep 4, 2006

Captain Foo posted:

can someone square �linux was nearly production ready at 0.96� with �linux was a nearly unusable toy until 2.6�

linux 1.2.13 was a production ready toy for hobbyists

# ¿ Oct 7, 2023 22:01

Adbot: ADBOT LOVES YOU

# ¿ May 4, 2024 22:19

shackleford: Sep 4, 2006

Share Bear posted:

installed debian on an ancient core duo with ancient amd card and it didnt put out graphics via displayport so i hooked it up via hdmi, which is fine but doesnt do 4k? seemingly?

i probably gotta get newer drivers i assume for this

i honestly hate like throwing out perfectly usable computer poo poo, so i'm using this old rear end computer for like tiny dev poo poo and tiny linux based streams

how old is ancient

according to https://tomverbeure.github.io/video_timings_calculator, 4K @ 60 Hz requires a bit more than 11 Gb/s which needs HBR2 mode for DisplayPort (introduced in DisplayPort 1.2+) or 600 MHz TMDS (HDMI 2.0+).

the bandwidth required for 4K @ 30 Hz should be doable with HBR mode (DP 1.0+) or HDMI 1.3+.

on linux the video mode setting stuff is in the kernel driver and the out-of-tree drivers from AMD have probably dropped support for your card so there probably isn't any driver to update if you're on whatever kernel is in debian stable.

maybe check the contents of /sys/class/drm/card0-DP-1/modes or whatever to see what video modes the kernel thinks it can set

# ¿ Oct 11, 2023 20:40

shackleford: Sep 4, 2006

pseudorandom name posted:

no, HDMI 2.1 and newer are secret NDA'd specs. its a completely different higher bandwidth signalling protocol

you can get some low framerate low color depth 4k resolutions with HDMI 2.0 using lossy compression but higher res/framerate/color/lossless can't fit in the bandwidth

there's a few charts in the wikipedia article that show what combinations are available in different HDMI versions

i believe HDMI 2.0 using 600 MHz TMDS signalling has enough bandwidth to do 4K @ 60 Hz without Display Stream Compression. i guess your statement depends on whether you consider 60 Hz low frame rate or not

is it just HDMI FRL signalling that's part of the paywall or is it all of the features in HDMI 2.1 that aren't in HDMI <= 1.4? like presumably there isn't a whole lot of secrets in "run the TMDS clock at 600 MHz instead of 340 MHz" vs. a completely new signalling mechanism

lol what a mess https://tftcentral.co.uk/articles/when-hdmi-2-1-isnt-hdmi-2-1

HDMI Licensing Administrator posted:

1. HDMI 2.0 no longer exists, and devices should not claim compliance to v2.0 as it is not referenced any more

2. The features of HDMI 2.0 are now a sub-set of 2.1

3. All the new capabilities and features associated with HDMI 2.1 are optional (this includes FRL, the higher bandwidths, VRR, ALLM and everything else)

4. If a device claims compliance to 2.1 then they need to also state which features the device supports so there is �no confusion�

# ¿ Oct 12, 2023 01:50

shackleford: Sep 4, 2006

I'd just like to interject for a moment. What you're referring to as Linux, is in fact, Windows/Systemd/Linux, or as I've recently taken to calling it, WSL

# ¿ Oct 12, 2023 20:56

shackleford: Sep 4, 2006

the systemd configuration syntax is based on microsoft .ini files which iirc is one of the reasons the devuan and "veteran unix admins" sickos hate systemd, because they also hate microsoft of course

# ¿ Oct 12, 2023 22:04

shackleford: Sep 4, 2006

BattleMaster posted:

oh no you can't use equal signs, microsoft uses those!!!

exactly, you've got spaces (resolv.conf), colons (passwd/shadow), whatever sick poo poo is going on in sudoers, etc. etc. the specific syntax used by systemd (https://www.freedesktop.org/software/systemd/man/systemd.syntax.html) isn't really used by anything else on a linux system except like XDG stuff and stuff that's intentionally chosen to model itself on systemd's syntax. a lot of this 90's stuff on a linux system is based on 80's unix stuff not 90's microsoft stuff

like all these shell fragments in /etc/default/ aren't the same as .ini files because they use equal signs when assigning to a variable

# ¿ Oct 12, 2023 22:18

shackleford: Sep 4, 2006

maybe .ini files are more common in stuff implemented in python since they have an .ini style parser in the standard library idk

# ¿ Oct 12, 2023 22:20

shackleford: Sep 4, 2006

i haven't checked but i think the devuan folks hate rust too because the compiler doesn't support m68k or alpha or something, but also because they're far-right fascists and rust is woke

# ¿ Oct 12, 2023 22:33

shackleford: Sep 4, 2006

lmao what is going on with google's semantic searching. this is on the first page of results for me

# ¿ Oct 12, 2023 22:41

shackleford: Sep 4, 2006

Captain Foo posted:

�

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1041731

# ¿ Oct 16, 2023 02:37

shackleford: Sep 4, 2006

copying shell commands out of a vendor's PDF documentation but having to fix up all the line breaks, ligatures, quotes, and dashes to get it to run

# ¿ Oct 18, 2023 00:09

shackleford: Sep 4, 2006

mycophobia posted:

i will continue using whatever filesystem debian stable comes with

# ¿ Nov 1, 2023 20:48

shackleford: Sep 4, 2006

the journalctl command, with no arguments, prints out log entries in traditional syslog-ish format to stdout in chronological order. even knowing nothing else about systemd or the journal, piping journalctl to grep is already superior to groveling around in /var/log

# ¿ Nov 2, 2023 07:45

shackleford: Sep 4, 2006

well, hold on

https://sqlite.org/codeofethics.html

# ¿ Nov 3, 2023 23:05

shackleford: Sep 4, 2006

FlapYoJacks posted:

lol at anyone who uses OpenZFS:

OpenZFS Is Still Battling A Data Corruption Issue

lol the hubris. pop the corks we're done here

# ¿ Nov 28, 2023 22:57

shackleford: Sep 4, 2006

so i could not figure out why zoom version 5.15.x up to around 5.15.7 or so would run in native wayland mode and work fine on my wayland compositor, while later versions would insist on running in xwayland mode and gently caress up the UI and be completely unusable. i thought i was missing some environment variable or something when invoking the zoom binary.

on a hunch i traced the getenv/setenv library calls when running the /opt/zoom/ZoomLauncher binary from newer versions and this bastard is overwriting the QT_QPA_PLATFORM environment variable to "xcb" before invoking the main zoom binary. arrrrgh i'm so loving mad

# ¿ Dec 2, 2023 07:33

shackleford: Sep 4, 2006

Soricidus posted:

actually I use regular debian, op

# ¿ Dec 2, 2023 19:55

shackleford: Sep 4, 2006

shitface posted:

I never really thought about this but it as at least a very similar brand of brain worms where anything, anywhere that changes is being 'forced on me aaaagh"

the devuan weirdos who hack the systemd dependencies out of their debian fork call themselves "Veteran Unix Admins", they are absolutely hearkening back to a nostalgic and mythological past (where linux systems could be reliably booted with a pile of buggy shell scripts) and believe the changes today are being forced on them by outsiders

the debian project periodically has a developer vote about systemd (e.g. https://www.debian.org/vote/2019/vote_002_results.png) that the systemd detractors lose and some of them rage quit the project, because an issue they cared about was decided by a democratic vote and they lost and they couldn't handle that

# ¿ Dec 4, 2023 07:23

shackleford: Sep 4, 2006

Hed posted:

lol why is that vote result a digraph? it looks like there was a multiple choice A-H?

Each voter's ballot ranks the options being voted on. Not all options need be ranked. Ranked options are considered preferred to all unranked options. Voters may rank options equally. Unranked options are considered to be ranked equally with one another. Details of how ballots may be filled out will be included in the Call For Votes.
If the ballot has a quorum requirement R any options other than the default option which do not receive at least R votes ranking that option above the default option are dropped from consideration.
Any (non-default) option which does not defeat the default option by its required majority ratio is dropped from consideration.
1. Given two options A and B, V(A,B) is the number of voters who prefer option A over option B.
2. An option A defeats the default option D by a majority ratio N, if V(A,D) is greater or equal to N * V(D,A) and V(A,D) is strictly greater than V(D,A).
3. If a supermajority of S:1 is required for A, its majority ratio is S; otherwise, its majority ratio is 1.
From the list of undropped options, we generate a list of pairwise defeats.
1. An option A defeats an option B, if V(A,B) is strictly greater than V(B,A).
From the list of [undropped] pairwise defeats, we generate a set of transitive defeats.
1. An option A transitively defeats an option C if A defeats C or if there is some other option B where A defeats B AND B transitively defeats C.
We construct the Schwartz set from the set of transitive defeats.
1. An option A is in the Schwartz set if for all options B, either A transitively defeats B, or B does not transitively defeat A.
If there are defeats between options in the Schwartz set, we drop the weakest such defeats from the list of pairwise defeats, and return to step 5.
1. A defeat (A,X) is weaker than a defeat (B,Y) if V(A,X) is less than V(B,Y). Also, (A,X) is weaker than (B,Y) if V(A,X) is equal to V(B,Y) and V(X,A) is greater than V(Y,B).
2. A weakest defeat is a defeat that has no other defeat weaker than it. There may be more than one such defeat.
If there are no defeats within the Schwartz set, then the winner is chosen from the options in the Schwartz set. If there is only one such option, it is the winner. If there are multiple options, the elector with the casting vote chooses which of those options wins.

Note: Options which the voters rank above the default option are options they find acceptable. Options ranked below the default options are options they find unacceptable.

When the vote counting mechanism of the Standard Resolution Procedure is to be used, the text which refers to it must specify who has a casting vote, the quorum, the default option, and any supermajority requirement. The default option must not have any supermajority requirements.

# ¿ Dec 5, 2023 00:43

shackleford: Sep 4, 2006

Woolie Wool posted:

I for one can't believe Linux made it for almost 20 years on what was essentially a glorified version of AUTOEXEC.BAT

yeah looks like about 22 years if you measure from these changelog entries to debian/rhel shipping systemd by default

quote:

2.0 08-Dec-92
- Rewrote the code totally, so started with a new
version number.
- Dropped Minix support, this code now is Linux - specific.
- With TEST switch on, this init & telinit can
run standalone for test purposes.

1.3, 05-Jul-92
- Got a 386, so installed Linux. Added 'soft' reboot
to be default under linux. Fixed some typos.

# ¿ Dec 5, 2023 02:19

shackleford: Sep 4, 2006

modprobe mscdex.ko

# ¿ Dec 5, 2023 02:23

shackleford: Sep 4, 2006

Beeftweeter posted:

yeah its so bad that microsoft implemented it into the nt kernel 30 years ago

(and just gave us a truly useful way of using that like 4 years ago)

as far as i can tell unix support on NT has undergone four distinct phases each of which was mostly unrelated to and supplanted the previous one. the recent stuff is like totally unrelated to the early stuff

there was the original totally useless POSIX subsystem which was just some vintage POSIX.1 APIs and didn't include a shell, purely for decorative FIPS certification purposes

there was the interix acquisition that brought "Services for Unix"

there was WSL1 which was a syscall compatibility layer for linux ELF binaries

then when they got tired of trying to emulate the whole linux kernel/userspace interface in windows they gave up and started running linux binaries in a managed VM, that's WSL2

# ¿ Dec 5, 2023 21:12

shackleford: Sep 4, 2006

Beeftweeter posted:

lol welp. only comment i have here is that i think WSL1 is still being actively developed? it at least was an option not too long ago

ah maybe i've never used it. looks like it only dates back to 2017 so i figure they found out pretty quickly that their approach was just going to be chasing corner cases forever. maybe they can hire some freebsd developers to help 'em out, looks they're still working on linux(4)

# ¿ Dec 5, 2023 21:22

shackleford: Sep 4, 2006

FlapYoJacks posted:

2017 was 6 years ago OP.

lol the CPU in my primary laptop is a skylake chip :corsair:

# ¿ Dec 5, 2023 22:08

shackleford: Sep 4, 2006

BobHoward posted:

no torvalds explosion in what i saw sorry

supposedly he doesn't do that sort of stuff any more after he did drugs or CBT or whatever back in 2018

https://www.zdnet.com/article/linus-torvalds-on-state-of-linux-today-and-how-ai-figures-in-its-future/ posted:

Now for Torvalds, "being there all the time is not a problem because I love doing what I'm doing. I was on vacation a few months ago, and I have my laptop. And if I hadn't had my laptop with me, I would have been so bored. It is what I do. But I realized that's not the life for everybody, especially when you have to put years of your life into this."

It's also something Torvalds has had to learn to be better at as well. "Code is easy to write. You have a right answer, and you have a wrong. People relationships are hard, and being able to work with other developers and maintainers, especially when you have maintainers that work on different things with different goals. They want to push their area in one direction, and another maintainer comes in from another area and wants to pull it in another direction. It can be very stressful."

In 2018, Torvalds decided to pull back from his angry young man stance. He took a break from the Linux kernel to work on his behavior toward other developers. After he got a handle on it, Torvalds returned to the kernel. He's been much more mild-tempered since then. As he mentioned in Tokyo, he won't be "giving some company the finger. I learned my lesson."

To sum it up, Torvalds said, "It's one of those things where a lot of people seem to think that open source is all about programming, but a lot of it is about communication, too. Maintainers are the ones who translate. I don't necessarily mean language. I mean, the context, the reason for the code. That makes for a tough job. But, if you want to be a maintainer, trust me, there's room at the top."

# ¿ Dec 10, 2023 23:53

shackleford: Sep 4, 2006

Truga posted:

i don't really care about which i use (i'm more used to apt so i use that most of the time), but good lord why are dnf/yum often so slow?

the red hat stuff is python and BerkeleyDB databases, the debian stuff is fast optimized C++ stuff that reads and rewrites text files

# ¿ Dec 11, 2023 20:20

shackleford: Sep 4, 2006

has anyone said uhh hail satan yet

https://lkml.org/lkml/2023/12/11/226

# ¿ Dec 11, 2023 21:59

shackleford: Sep 4, 2006

so apparently nixos requires a fairly large amount of cloud storage and awkward things have to happen if that storage isn't comped

https://discourse.nixos.org/t/the-nixos-foundations-call-to-action-s3-costs-require-community-support/28672

https://discourse.nixos.org/t/nixos-s3-short-term-resolution/29413

looks like that 425 TB is about ~100X the size of the debian archive

https://www.debian.org/mirror/size

# ¿ Dec 12, 2023 19:41

shackleford: Sep 4, 2006

Athas posted:

I don't think Nix actually works through weird dynamic ld glue - usually the library paths into the Nix store are hardcoded in the executables. (Which makes them completely unportable on the own, although you can compute a "closure" of the parts of the store they depend on, which is... essentially done by grepping for hashes in all files. Don't look behind the curtain.)

huh does that mean if you update a library to a new point release (i.e. an update that doesn't break the library's ABI/API) the new library goes into its own unique path and you have to rebuild (or at least relink?) all the binaries that link against the old version?

# ¿ Dec 13, 2023 00:24

shackleford: Sep 4, 2006

euroshopper posted:

wasn't nt written on MIPS or some related risc processor and then entirely abandoned the "universal os" philosophy because no software vendors were willing to rewrite their wintel applications for a barely used version?

they had a workaround for that at least on alpha

https://en.wikipedia.org/wiki/FX!32

the workaround was emulation and binary translation

# ¿ Dec 13, 2023 05:54

shackleford: Sep 4, 2006

seemed to work out ok for AMD and radeon.ko / amdgpu.ko

i considered getting one of the low end intel arc cards (one of the models without an auxiliary PCIE power connector) for my posting station but apparently if you have dual 4K monitors the idle power draw is like 40+ watts, and supposedly it's a bug in the hardware that they can't fix in the driver

(comparable low end radeons are like 5-8 watts idle)

# ¿ Dec 16, 2023 18:58

shackleford: Sep 4, 2006

specifically nouveau + NVK

https://www.collabora.com/news-and-blog/news-and-events/introducing-nvk.html

https://airlied.blogspot.com/2023/08/nvk-kernel-changes-needed.html

# ¿ Dec 17, 2023 16:47

shackleford: Sep 4, 2006

https://lists.debian.org/debian-devel/2023/12/msg00127.html

LibrePGP? what is that man doing to his PGP implementation

# ¿ Dec 21, 2023 19:03

shackleford: Sep 4, 2006

quote:

There is a lot of momentum in our organization. Take timekeeping.
There is local time, UTC, TAI, GPS time, Julian time, modified Julian
time, nutation sidereal time, mean sideral time, Greenwich sideral time,
modified local sidereal time, and probably more I can't remember. We
have to keep track of contintal drift because of General Relativity with
regards to timekeeping, and our most demanding applications require
femtosecond accuracies in the clocks (which isn't possible with today's
technology). Not only do we use all of those timekeeping standards in
different places, but we also invented our own timekeeping standard back
in the 1970s that we still use and is one of our main ways of keeping
track of things.

Some groups here here don't trust DHCP. They started out soldering and
wirewraping IP addresses into boards that are inaccessible in the bowels
of the machines, but have progressed to selectable IP addresses
controlled by dip switches inside brass boxes that have up to 206 screws
holding them closed.

The modern data stream uses ethernet, but I'm told not ethernet
collisions. It was deemed better to time every packet from every device
so that no collisions ever happen in the ethernet network.

But containers are something we plan to try soon! This coming calendar
year actually. Using one for RHEL6.5 to support Xilinx is high on my
list.

# ¿ Dec 22, 2023 18:56

shackleford: Sep 4, 2006

Captain Foo posted:

what the christ is this lmao

radio astronomers apparently lol

https://lists.isc.org/pipermail/bind-users/2023-December/108219.html

shackleford fucked around with this message at 08:01 on Dec 23, 2023

# ¿ Dec 23, 2023 07:59

shackleford: Sep 4, 2006

Beeftweeter posted:

i'm trying to get a custom build of ffmpeg going but hunting down the dependencies manually is a little bit more effort than i'm willing to put in

e: i do have it building btw, that's not the issue. it's just not with all of the external libraries i'd like (oddly intel's own loving distro doesn't come with libmfx or libvpl or have the packages available)

hahah i don't know anything about the RPM world but does it have an equivalent of "apt build-dep" because hunting down all the build deps for a typical build of ffmpeg is a lot of packages

code:

$ podman run -it --rm debian:sid
root@755ed92f46b8:/# sed -i -e 's/^Types: deb$/Types: deb deb-src/' /etc/apt/sources.list.d/debian.sources
root@755ed92f46b8:/# sh -c 'apt -qq update && apt -qqy full-upgrade && apt -qqy install build-essential' 1>/dev/null 2>&1
root@755ed92f46b8:/# apt -o APT::Get::Show-Versions=1 build-dep ffmpeg
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
The following NEW packages will be installed:
   autoconf (2.71-3)
   automake (1:1.16.5-1.3)
   [...]
   xtrans-dev (1.4.0-1)
   zlib1g-dev (1:1.3.dfsg-3)
0 upgraded, 544 newly installed, 0 to remove and 0 not upgraded.
Need to get 299 MB of archives.
After this operation, 1257 MB of additional disk space will be used.
Do you want to continue? [Y/n] ^C
root@755ed92f46b8:/#

# ¿ Dec 23, 2023 16:16

shackleford: Sep 4, 2006

Beeftweeter posted:

https://www.clearlinux.org/clear-linux-documentation/guides/clear/swupd.html

i love the juxtaposition of the cute li'l warning message that it prints about violating IT policies when you disable automatic updates and also sometimes you just lose track of what all got installed into /usr/lib/python* �\_(ツ)_/�

# ¿ Dec 23, 2023 17:06

shackleford: Sep 4, 2006

https://lore.kernel.org/lkml/CAKMK7uGK_N1ReMDY_Os0dyRpE9XixuDkT=+WfL8c_=shSoQnzA@mail.gmail.com/

here is a senior intel linux graphics guy expressing hope that the "poo poo driver situation" can be cleaned up somewhat before he retires

# ¿ Dec 23, 2023 17:23

shackleford: Sep 4, 2006

Beeftweeter posted:

either way there's really only a few things i like to compile on my own, and they stem from two basic packages: ffmpeg/libav* and wine (and to be fair, both of those have a shitload of dependencies if you want them). but it's not an unmanageable amount of stuff i think, i used to do both on macos even with brew/macports since those were frequently outdated and by default only did basic-rear end optimizations anyway

but once you do have a lot of the basic stuff (e.g. libjpeg-turbo, libxml2, freetype, etc.) compiled the way you want (again, if you want), you usually don't need to update them all that often, and even then that's only if there's some reason the provided versions aren't adequate. like i might want to optimize the hell out of libx265 or something similar, but i don't really care about, idk, libmp3lame. you know?

so if i were going to do this in the debian ecosystem (and it's looking like i am because i think i found an issue in libavformat's RTP reassembler that i can only work around by patching ffmpeg) i'd pull the ffmpeg source package from unstable, patch the source to my liking, build a new source package, and then feed my new source package into pbuilder which will perform the whole build including build dependency resolution etc. in a temporary, clean chroot environment. this then produces a pile of custom .deb's i can schlep around or a huge build log to troll through if i hosed something up.

where this falls down of course is that the "debian ecosystem" is an incredible multi-decade accretion of slowly evolving tooling that requires a huge investment of time to learn and to keep current on in order to do basic things, so unlike the a/n/g users i must decline to evangelize.

one of the things i really like about the red hat world (but maybe this is just my impression since it's probably been a full decade since i've built an rpm or even looked at a spec file) is that they have just enough centralized control to tamp down on the complexity and profusion of competing workflows and tooling that you can get in a sprawling volunteer project like debian. so if you're looking at some random rpm packaging repo with a spec file and some patches in it they're all pretty similar and you have a pretty good likelihood of being able to understand it.

# ¿ Dec 24, 2023 20:38

Adbot: ADBOT LOVES YOU

# ¿ May 4, 2024 22:19

shackleford: Sep 4, 2006

drat look at those CFLAGS. may as well just run debian unstable

# ¿ Dec 29, 2023 18:24

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > Linux: is the machine running fedora or just the user

«‹›4 »