Even Linux Kernel Has Terrible Code Like This!

When we think of Linux we think it's this super well designed project where everybody knows what they're doing but in reality it seems that it may be a little bit more haphazard with hacky code like this existing.

==========Support The Channel==========

==========Resources==========

=========Video Platforms==========

==========Social Media==========

==========Credits==========
🎨 Channel Art:
Profile Picture:

#Linux #LinuxKernel #OpenSource #FOSS #LinuxDesktop

🎵 Ending music

DISCLOSURE: Wherever possible I use referral links, which means if you click one of the links in this video or description and make a purchase I may receive a small commission or other compensation.

[adamsavard]

Everything in the world runs on hacks and terrible solutions. I'm not saying that's a good thing, but anyone that's surprised by dumb things happening in code isn't a software developer.

[annie]

hacky code is what makes the majority of bigger software run xD At least from my experience in the industry

[thingsiplay]

"The Internet Will Be The Internet"

2022, Brodie on the internet

[thepaulcraft]

I am building a voxel "game" with rust and ogl for school and it feels like onr big hack(that isn't working at all). Great video and "nice too see" that I am not the only person writing shitty code

[notuxnobux]

All graphics drivers do this to apply workarounds for games. Proton (and dxvk) does this too.

[timewave]

I once integrated DOSBox into the build system of ATE software I maintained, because it required 16 bit executables to build some firmware, but needed to build on 64 bit workstations. Some HCL contractors tasked with porting my change to a forked version of the software not only didn't understand what I had done or why—they didn't understand how anyone could have thought to do something like that. Prior to introducing that hack, the same build system broke on some workstations. I traced the cause to the same CMD script I would later move to DOSBox, running out of memory for environment variables, which I was able to mitigate by temporarily unsetting unnecessary envirnoment variables.

[radomane]

The new type of the Mandela effect, Reddit convincing themselves that “if your process starts with X, it breaks on every kernel after 2019”, despite no evidence supporting that claim.

[Acorn_Anomaly]

6:49 I think maybe you misunderstood the Xorg merge request being discussed, or Daniel's response.

Or maybe I've misunderstood something.

They're not considering "changing the name of the process calling it". That merge request is simply hooking up the bits for atomic support. Daniel's information is misremembered from his own patch(saying the name "Xorg" instead of "X"), but he's only pointing out that, regardless of what they do with the atomic code on the X side of things, the kernel simply isn't going to tell them that atomic is available if it's coming from an X process.

Before that, Marten Lankhorst said that this is something that they still have to keep broken in the kernel, and wants to know if it's been fixed, so they can remove the kernel patch.

Emma Anholt replies to him, saying that based on "kernel rules", the kernel is going to have to "remain compatible" with this for "years". In my opinion, basically brushing Marten off, and telling him "the kernel's gotta stay broken for us anyway, stop bothering us."

Daniel is informing Emma that keeping the kernel "compatible" in this case means that the atomic code that they're working on in X will never work, anyway, because the kernel is deliberately telling any X processes that atomic isn't available.

He isn't saying "you have to change the process name to work around this", he's saying "what you're doing isn't going to work without changes on the kernel side."

[brunoais]

9:47: I like the face pallet swap 😆.

Curious: How long did you take between the takes?

[anon_y_mousse]

I can't say that I'm surprised, but it does kind of prove one of the things I've said before, Linus can't check every bit of code to make sure it's perfect, he has to rely on the people he works with to ensure the code isn't bad, and instances like this prove that he really can't rely on them all.

[bruderdasisteinschwerermangel]

computer graphics is proving yet again that it's a complete clusterfuck

[DMSBrian]

God damn now that's a REAL hack right here xD

[Karpfly]

I have done some hacks in university projects not becouse i wanted it in but there was a deadline so i have to glue some parts just to get extra points. other parts where nice and polished for my skill

[DaBoomDude]

Hacky code is my life when i was at uni. If it does the job, its good enough for me.

[KirbyZhang]

No against hacks, but why do people not want to put the hack on the X side?

[RealMazharHussain]

Wow, it's that hot there? You never have the same outro for 2 consecutive videos.

[TuxPeng]

I haven't had a problem with X11 since FGRLX. There are 3 things I use (xclip, xdotool, zoom screensharing) that don't really work in wayland

[Sebanisu]

I feel like all my code is bad and hacky. Though that could be imposter syndrome. :)

[Mitsunee_]

Brodie, did you clean your room? I'm so proud.

[fu]

things will break though they will just break strangely to the point it will not be clear it is the cause and no one will know how to even debug it.