palordrolap

This has bell curve meme vibes. I’m just not sure what the middle guy would be saying.

You need to lay at least some blame on Logitech for that one.

They’ve sold drivers to Microsoft, but since no-one writing Linux would give them any money, they wouldn’t provide drivers for their proprietary hardware.

This then lead to early Linux adopters buying non-Logitech devices and not seeing a use-case for rolling a reverse-engineered driver into the kernel.

Logitech still haven’t written their own Linux driver. I wouldn’t be surprised if some of the money from Microsoft is so that they don’t.

Pretty sure Cinnamon panels are designed to fit to the width (or height if they’re attached to the side) of the screen and can’t be altered. (The depth/breadth/“thickness” of the panel can be changed, but that’s not particularly relevant here.)

The same may also be true of panel-like features in XFCE and MATE too, but I can’t easily confirm.

Do. Take a boot USB for a spin. Try a few distros.

I’ve been on Linux (Mint) for years and never had a mouse-wheel not work or any problems with sound (hardware failure notwithstanding). The computer’s been the same all the way through, but it is a bit of a Ship of Theseus at this point. Mint has had no problem with new (and old) parts that I’ve thrown in. Or new mice, as I implied before.

Getting old Windows games to work has been the biggest non-starter, which is pretty much where OPs friend was having trouble too.

Minecraft (Java) runs fine with the standard launcher, but I do get FPS problems if I’ve had an Xorg update. That’s more of a “your graphics card is so old Mint doesn’t really support it any more” problem, which I know how to work around.

I did have problems getting Linux to run on a laptop once, but then it was 1998 and Linux drivers weren’t quite so plug and play. I had no idea what refresh rates my TFT screen needed and neither did Linux, boldly warning that if I set them wrong I could burn out my screen. Since I needed a GUI, I went back to Windows 95.

True. There are various legitimate tools that are only really one step away from malware, so it’s not too hard to imagine going that one step further.

Thinking specifically of the fact that a new process is allowed to change its apparent name, as well as creating secondary process pools, but there are bound to be other, deeper ways.

Be aware that for some removable (or otherwise non-local) media, some systems will create a .Trash-### directory on the media itself in the root directory.

This prevents unnecessary copying of files from the media to a local disk, and only a few media-specific location indicators actually need to be changed for the Trashed file(s).

The ### is generally the user’s ID number as stored in /etc/passwd, and, on Debian derivatives at least, is usually 1000 for the first user, 1001 for the second, etc., but I have heard of some systems that just use .Trash with no suffix, or did so at some point in the past.

Is it still the norm to go to the dev’s office, yank their power cord and when they ask what we’re doing, tell them we’re shipping their machine to the client because it’s the only one that the code runs on?

And can we do that with whatever server ChatGPT-4o is running on?

I’m assuming that this response from 4o isn’t real and was invented for the laugh, but it would be tempting to throw this scenario at it if it decided to give this response.

My guess is a “solution” to the age-old problem of needing to store a secret in a file that the user can download, thus making the entire system insecure.

This “solution” appears to be either that the string itself is so outrageous that the user would not believe that it’s the real secret when it is in fact the real secret, leveraging security through obscurity, or else it’s there in place of the real secret that cannot be revealed under pain of death firing, and therefore is accidentally being used instead of that intended secret… so it’s not secret after all.

Unless they’re doing something incredibly clever to substitute that secret string for the real thing when the time is right and doing it in such a way that the user can’t intercept, someone’s getting fired.

Someone told me every processor used 0xEA

Not sure if this is a riff on the joke or not.

Back in the day I dabbled in 6510 code, and up until today hadn’t even bothered to look at a chart of opcodes for any of its contemporaries. Today I learned that Z80 uses $00 for NOP.

Loth as I am to admit it, that actually makes sense. Maybe more sense than 65xx which acts more like a divide-by-zero has happened.

The rest of the opcode table was full of alien looking mnemonics though, and no undocumented single byte opcodes? Freaky, man.

But the point is that not even Z80 used $EA. If the someone was real they probably meant every 65xx processor.

They need to update that. They jumped to version 24 for 2024. 24.2.3 is the current version.

Someone else points out that Python’s native bool is a subtype of int, so adding a bool to an int (or performing other mixed operations) is not an error, which might then go on to cause a hard-to-catch semantic/mathematical error.

I am assuming that trying to add a NumPy bool_ to an int causes a compilation error at best and a run-time warning, or traceable program crash at worst.

80 characters

Two hours and no-one’s challenged this? People must be asleep.

(This is not that challenge. Only pointing out that someone usually has by now.)

Ones I have used: GNOME Disks’ create and restore image features. Possibly Mint’s mintstick for writing a distro’s .iso out to a USB stick. I am not too sure on that.

I assume old-school dd still works as well, which might be a better option for scripted backups or minimal systems.

named myself

Programmer error: I totally missed a bunch of edge cases there. Cases 32 to 35 for sure

Buuut I’m guessing you didn’t name yourself after the editor.

A wild username reference appears!

There’s an editor called Kate. It’s probably not named after you, but if you’re young enough and the person who named you was into tech, you might be named after it.

So, it’s trying to replace Emacs?

I was counting WebKit and Blink (Chromium) as cousins, both descendents of KHTML, but maybe they’ve diverged enough as to not easily be able to borrow from each other any more and it really is three.

This is pretty big if it has a completely separate browser rendering engine to the two remaining families, even if the feature set is small.

The more current alternatives we have, the better.

Comparing audio cassettes to modern high-density tape storage is pretty much the same comparison as an 8-bit computer with a modern 64-bit server, or, say, a hamster with a human.

Basically the same thing, but the differences are somewhat notable.

512KB? At the risk of going all Four Yorkshiremen, that sounds luxurious.

Floppy disks held 170KB if you were lucky to have a drive. The PET line, like many 8-bit computers, used a cassette tape drive (yes, those things that preceded CDs for holding and playing music). Capacity depended on the length of the tape. And it took ages to load.

The PET was fancy because it had a built-in cassette drive. That’s what you can see to the left of the keyboard in the picture.