It sounds like he doesn’t want update notifications. I’m sure that that’s possible, but I’ve no idea what sort of update notification he’s upset about. Could be from some kind of package manager, Firefox itself, whatever. He might also have a nightly build version installed or something, when he could benefit from extended support version.

I’m sure that whatever platform you’re on lets you not update if you want.

I wouldn’t advise that from a security standpoint, but if you’re set on it…

Secondly, I couldn’t even install qemu if I wanted to because it wasn’t in the apt repositories that shipped with Debian.

Debian has a non-free repo containing non-open-source software that it hasn’t historically enabled by default, but I don’t think that that’d apply to qemu. I’m pretty sure that’s all open-source.

goes looking.

qemu’s been in the Debian repos since…checks sarge, which was released as a stable release in 2005.

And it was in main, not non-free, so it should have been there as an out-of-the-box enabled repo:

https://snapshot.debian.org/archive/debian/20050312T000000Z/pool/main/q/qemu/

QEMU only came out in 2003.

https://en.wikipedia.org/wiki/QEMU

QEMU is free software originally developed by Fabrice Bellard; the first preview release was in 2003.

It looks like it was packaged in Debian unstable since 2004, though I wouldn’t recommend jumping right on unstable to a new user.

$ apt changelog qemu-system 2>/dev/null|tail -n 15

 -- Paul Russell <prussell@debian.org>  Mon, 15 Mar 2004 23:56:25 +0100

qemu (0.5.2-2) unstable; urgency=low

  * Fix build problem so bios.bin etc. can be found. (Closes: #237553)

 -- Paul Russell <prussell@debian.org>  Fri, 12 Mar 2004 05:43:00 +0100

qemu (0.5.2-1) unstable; urgency=low

  * Initial Release. (Closes: #187407)

 -- Paul Russell <prussell@debian.org>  Wed,  3 Mar 2004 02:18:54 +0100
Fetched 314 kB in 0s (1,431 kB/s)
$

True CLI…How do you currently automate screen capture on…Linux?

Under Wayland?

$ grim

For anyone else who, like myself, doesn’t know what a BC-250 is:

https://bc250.info/

AMD BC-250

A PS5 APU’s second life as a budget Linux gaming PC

Originally built for crypto mining, these cut-down PS5 APUs were headed for the scrap heap. Then Linux driver support arrived — and everything changed. Now they’re surprisingly capable mid-range gaming PCs with 16GB GDDR6 that double as local AI inference machines.

https://en.wikipedia.org/wiki/LRC_(file_format)

For time-synchronized lyrics.

It looks like my Scandroid audio files from Bandcamp have embedded non-synchronized lyrics.

$ metaflac --list *08*.flac

[snip]

    comment[12]: UNSYNCEDLYRICS=Beware the shadows of the drones 
Destruction wrapped in pretty silicone 
They’ve taken everyone I’ve known 
And now I walk these empty streets alone 

In my memory the past is fading 
The future has been redesigned 
It’s hard to focus on it when I’m running out of time 

Surrounded by streetlights at midnight 
My destination is unknown 
I walk these empty streets alone 
Digital dreams thrive in the moonlight 
I’m only flesh, circuit and bone 
I walk these empty streets alone 

Beware the faces of the clones 
Deception sent from Neo-Tokyo 
They’ve erased everyone I’ve known 
And still I walk these empty streets alone

I don’t know how widespread that is.

Note that lyrics are themselves copyrighted works, regarding simply sharing them. That is, if you want to create, say, “Lyricsnet, the new Fediverse service” or something like that, you may attract attention from IP owners.

Those who believe we’ve been in an incrementally escalating software crisis since at least 2007.

Web development is now especially notorious for completely disregarding accessibility, user device capabilities, and regulations. Most of the ideas of user-centred design are alien to modern developers.

goes to archive.org’s copy of Yahoo from the golden era of 2006

https://web.archive.org/web/20061212011659/http://www.yahoo.com/

Why miss out?
To see all the new Yahoo! home page has to offer, please upgrade to a more recent browser.

Supported browsers include:
Internet Explorer 7 optimized by Yahoo!
Firefox 1.5
Safari 2.0
Opera 8.5

You can see every hop an IP packet takes at IP level with traceroute or mtr.

Do you get many hallucinations?

So, are you going to require matching the framerate between the user’s camera and the played video? Because if not, there won’t be a 1:1 correspondence between the two, and even if they’re at the same framerate, additionally you’re going to need to identify the start frame.

I didn’t poke around too deeply, but if it assigns a frame number to each QR code and encodes it in the code, maybe you can scan each frame of user video and discard duplicates.

If this is going to work reliably, you’re going to need many QR reads to work without failure or it’ll fail. Like, aggregate failure rate is going to be higher than failure rate for a single QR code. I think that I’d again probably look into some kind of FEC — I don’t know if PAR2 can generate chunks small enough to be individually stored for a QR code, but if it can, maybe use one of those chunks for each QR code instead of the output of split. That way, you only need N% of the QR frames to be successfully read.

searches

Hmm. Setting aside minimum chunk size, PAR2 may not support enough chunks for your use case — only 32k. I don’t know how large the files are that you’re trying to encode.

The project is supposed to convert files to QR code videos and back.

I don’t really understand what you’re trying to do.

I don’t think that you’re trying to just encode an URL, from a glance.

If you want to encode a video file as a QR code, you’re going to face the pretty fundamental limitation of information content.

A QR code can store a maximum of (checks) under 3 KiB. Virtually any video file cannot fit in those space constraints.

I am not a coder. I have been staring at a screen the past two weeks trying to “vibe code” for the first time with unpaid chat bots, for a seemingly simple project. But I still haven’t even gotten a single version to the real-world testing stage, keep getting failures on more basic self-tests.

So, chatbots in 2026 are not really a drop-in replacement for a computer programmer. My guess is that while we’ll probably get there, it’s probably going to require something more elaborate than just an plain old LLM. Like, it’s not going to be a minor refinement of what’s currently there, but a pretty fundamentally different system (though that system might potentially incorporate an LLM).

EDIT: Ah, it’s a series of QR codes. I mean, that’s going to be a pretty huge number for any real-world video.

EDIT2: Oh, you want a video of QR codes from a file. I think that a problem that you’re gonna hit there is the transport, because your video is gonna be playing at a constant rate and real-time QR code recognition isn’t perfectly reliable, and you’re probably not going to reliably capture every frame. Maybe someone could make something with FEC so that dropping a certain percentage of frames is viable, like, use PAR2 or something.

Also, assuming that you’re trying to make something that runs automatically, the generated code has stuff that will bail out if used other than if run interactively.

EDIT3: Also, it sounds like you want something portable. The script itself may (or may not be) portable, but it is trying to run external tools that are not going to be guaranteed to be present wherever.

EDIT4: I don’t know why you’d be running dash on denuvan, as it sounds like its default shell is bash, but if /bin/sh is a symlink to dash on that system, what you could try doing is generating a bash-specific version. It…might work if you just switch the shebang line to bash at the top of the file, but bash in sh compatibility mode is not identical to bash in bash mode.

If you want some sort of generic portability debugging tool, I’d add set -x as the second line of the file. That will enable trace mode. Every line that runs will be printed to stderr. Then run it like so:

$ yourscript.sh 2>debug1.log

That’ll output every line being run. Run it on both systems, and you’ll get a log on each. Then you can run:

$ diff -u debug1.log debug2.log|less

That’ll show you where the behavior starts to differ on the two systems.

But…I’m going to be blunt. I think that you’re asking more of current AI chatbots than is reasonable given their current state, and that it’s going to wind up being pretty frustrating if you’re aiming to do it with zero programming.

Magewell Pro Capture card

I’ve been kind of shifting towards use of USB devices over internal cards.

All of the USB devices that I have still can be connected to computers. Ditto for DE-9 serial ports, though I might need a USB adapter.

But I’ve seen ISA->PCI/AGP->PCIe obsolete a lot of old hardware that I’ve had sitting around, and that’s just on the PC. That includes my video capture hardware.

Sorry for the delay in getting back to you.

I hope I ran the vkgears test correctly?

Yeah, that’s fine.

Both of those should be using hardware rendering, at least based on my understanding of the text. You have the name of your video card where “llvmpipe” would show up, right above “64bit”, which is what happens on my system when using hardware rendering.

But…for some reason, you’re consuming a ton of CPU time when rendering using OpenGL, despite doing hardware rendering. That’s not what happens on my system. I don’t know what would cause that.

One would want it fixed either way. For Steam, one can force Proton to use OpenGL rather than Vulkan as a Direct3D backend by setting the environment variable PROTON_USE_WINED3D=1, which will cause many Windows apps to use OpenGL…but your problem is the opposite. Vulkan looks fine.

thinks

The only thing that comes to mind would be that there’s an Nvidia mechanism on systems where you have multiple GPUs — this can happen when you have an integrated on-CPU GPU and a discrete GPU on a laptop, say – to render on one and then copy to the other. I don’t know what text would show up as the renderer in that case, and I don’t have Nvidia hardware, much less Nvidia hardware plus an integrated GPU to test. I don’t think that that’s probably what’s going on here, but I don’t know what mangohud reports in that case. I would think that mangohud would be smart enough to actually display the renderer being used, but…maybe it’s not. But if you want to try it, you could give this a shot. I’m taking a stab in the dark rather than really analyzing it:

$ __NV_PRIME_RENDER_OFFLOAD=1 __GLX_VENDOR_LIBRARY_NAME=nvidia MANGOHUD_CONFIG=full mangohud glxgears.x86_64-linux-gnu

If the CPU usage when you run the above command goes from ~20% (as is currently the case for glxgears in your above screenshot) to ~4% (as is currently the case for vkgears), that might be what’s going on. If it is, then I’d try running your game with the __NV_PRIME_RENDER_OFFLOAD=1 __GLX_VENDOR_LIBRARY_NAME=nvidia environment variables set. I wouldn’t bet much money on it working, but I guess it’s not hard to try.

EDIT: If anyone else with an NVidia card wants to run the MANGOHUD_CONFIG=full mangohud glxgears.x86_64-linux-gnu command and report whether their system uses a ton of CPU time on all cores, that’d be a useful data point; I can’t, as I don’t have the hardware. I guess it’s possible that that the CPU usage could be normal — this is going through xwayland, and maybe something there causes that. I don’t want to flag it as something abnormal on OP’s system if it’s not. But it’s not the way my AMD system acts.

I noticed today that the CPU usage jumps to ~98% as soon as I start WoW. Even just in the main menu. I had previously started CS as a test to see how it behaves there. In CS, the CPU load stays around 25% even when playing against bots (so not just in the main menu) and with significantly higher graphics settings. So it seems there might be some kind of problem with the CPU?!

It’s probably not anything physically wrong with the CPU. It just means that the WoW binary is trying to do something in a loop and not getting anywhere (at least not quickly). It’s consistent with falling back to software rendering, which is what I guessed the problem might be, but other things could also cause that.

For Counterstrike, assuming that that’s what you mean by “CS”, I’d guess that the game’s engine probably uses about four threads, and it looks like you have 16 cores.

If you haven’t done the vkgears/glxgears stuff that I mentioned above to make sure that you can render using hardware acceleration using Vulkan and OpenGL, I’d do that. Given that you’ve said that you also get slow rendering performance with some other games, I’d be suspicious that that’s your problem, since that’s the behavior I’d expect to show up if you were falling back to software rendering — very slow rendering performance.

pokes around

It sounds like World of Warcraft switched to a 64-bit binary some time back, so probably no need to check that hardware rendering works for 32-bit binaries — just 64-bit, both Vulkan and OpenGL.

If you’ve already done the vkgears and glxgears stuff that I mentioned above and those work using hardware rendering without issues, then the underlying Linux 3D rendering stuff should be okay. That means that whatever’s at issue is probably in Proton or World of Warcraft. If you’re wanting to troubleshoot further, I’d probably try launching Steam from a virtual terminal (like, probably Konsole in your case) and then launching WoW and seeing if WoW prints anything there.

If nothing comes up there and you just get a black screen…well, let’s see. All of the important state should be on the WoW servers, so you could try uninstalling it and reinstalling WoW and see if that resolves whatever’s the issue. Take some time to download the game, but it’s not human time, so…shrugs it’d probably be the next thing I’d try.

But I wouldn’t recommend doing any of that until I’d done the vkgears/glxgears test I mentioned and confirmed that you’re rendering using your 3D hardware.

I find git to be a rather deep and complex video game with plenty of challenges to overcome as it is without trying to add even more layers on top.

Nothing there looks clearly broken to me. XDG_DATA_DIRS pointing at a directory that doesn’t exist shouldn’t hurt anything. Note that Steam does update itself occasionally when it starts, and I don’t recall what that looks like, so it’s not impossible that it was updating itself.

But Lisp is case-insensitive

looks bemused

I don’t do that much Lisp, mostly use it for emacs, but I’m pretty sure that it’s not.

opens emacs

(setq foo 1)                                                                                                                                                    
(print foo)                                                                                                                                                     

1

OK. So far so good.

(setq foo 1)                                                                                                                                                    
(print FOO)                                                                                                                                                     

Debugger entered--Lisp error: (void-variable FOO)
  (print FOO)
  (progn (print FOO))
  eval((progn (print FOO)) t)
  elisp--eval-last-sexp(nil)
  #f(compiled-function () #<bytecode 0xf6febdfec01a>)()
  eval-last-sexp(nil)
  funcall-interactively(eval-last-sexp nil)
      command-execute(eval-last-sexp)

Elisp sure doesn’t look to be case-insensitive. Maybe he meant some specific variant? Common Lisp?

$ sudo apt install sbcl

Apparently sbcl’s REPL doesn’t support readline.

$ sudo apt install rlwrap
$ rlwrap sbcl

Huh. Looks like with readline, I also get cursor flashing to do paren matching, kinda like emacs can do. I had no idea that readline could do that. Apparently Common Lisp doesn’t do setq either.

more experimentation

* (let ((foo 1)) (print FOO))

1 
1

Huh. So, yeah, I guess that Common Lisp is case-insensitive. That is a bit wild. I guess I do remember vaguely seeing old Lisp stuff with keywords in all-caps.

Is Scheme?

$ sudo apt install guile-3.0

Apparently the guile REPL doesn’t support readline either. God.

$ rlwrap guile

And it looks like “print” is “display” in Scheme-land.

scheme@(guile-user)> (let ((foo 1)) (display foo))
1

Okay, so that’s the syntax. Case-insensitive?

scheme@(guile-user)> (let ((foo 1)) (display FOO))
;;; <stdin>:2:24: warning: possibly unbound variable `FOO'
ice-9/boot-9.scm:1676:22: In procedure raise-exception:
Unbound variable: FOO

Nope.

I kinda feel like there are Lisps that the author could have used if they wanted Lisp and case-sensitivity, if that was the major irritation.

Today, Steam suddenly wouldn’t start at all. Even though the last thing I did yesterday was play CS. I turned on my PC today and Steam isn’t working anymore.

You can run the Steam client from a terminal like Konsole, and it’ll print a bunch of information about what it’s doing.

Can also have it write that information to a logfile, like:

$ steam 2>steamlog.txt

If you’re willing to wait until 2028 when memory prices are expected to drop, and if you’re willing to get new hardware if memory prices drop, I’d give real consideration to waiting until then. There’ll also probably be better hardware and better models then.

If you can constrain yourself to MoE-based LLMs, they’ll generally deal better from a performance standpoint with not entirely fitting in VRAM better than non-MoE LLMs, as experts may not get loaded into VRAM at all.