Yeah, I find it really tricky, because LLMs excel at adding yet another implementation to the side, whereas they will barely ever introduce an abstraction to cover both requirements.

And the humans often don’t really read the code anymore either. They ask the LLM to make a suggestion, then check the diff for plausibility and that’s it. They don’t have to solve it themselves, so they never properly dig into the surrounding context and therefore cannot notice the missing abstraction either.

And in the end, it’s not just that the humans don’t know the code, it’s also that it becomes near-impossible to refactor, because it’s just too much code to fit into a brain at a time.

You can’t anymore realize that one part of the codebase should’ve been solved by an already-existing abstraction.
Instead, you find five places in the codebase that really look like they should’ve been solved with a shared abstraction and then good luck figuring out what the common denominator is and how each of those places should be adjusted to fit this common abstraction, oh and also you didn’t see two more places that really should also be solved that way, but for them your chosen abstraction doesn’t make sense anymore, and then you ask the others and they tell you, oh man, that never made sense, we never really thought about it, because it was so easy to just tell the LLM to build it, and now that feature has to be removed in its entirety, so you just wasted two weeks trying to understand code that never made any sense in the first place.

Ask me how I know. 😑

I always feel like the much bigger challenge is whether anyone can still maintain the codebase afterwards. You throw away so much institutional knowledge by doing a conversion like this…

Yeah, computationally intensive stuff is going to be slow when implemented in Python. You will find lots of data science libraries for Python that do computationally intensive stuff, but they’re then generally implemented in C/C++ or Rust, with only a thin API layer written in Python.

Expecting the runtime/browser to ship a native implementation of virtually everything you might want to do in a turing-complete language, is also really not sustainable, though.

It was decided after Fukushima under Merkel.

Could be that the companies they supplied to filed for damages, due to them not meeting supply targets…

“Oven is going to be a grind, especially the first nine months or so. If work-life balance means a lot of time spent not working, it’s probably not a good fit.”

Oh man, someone here recently shared a blog post on startup culture in software companies, which likened software to an oven, where you’d add a specific button for a customer and that fucks up your whole architecture for years to come etc…

For just a moment, I thought this was a direct quote from that blog post.

I mean, I’d need to fiddle around with it, but presumably this wouldn’t be all too hard to script.

Just do a for-loop over the output of ls. And then in each iteration step, you’d do something like:

mkdir $name
mv $name $name/$name

This will mean, though, that the file extension is part of the folder name, if you care about that.

Well, in hopes that you’re not just trolling, maybe you’ll believe a dictionary that “stable” is ambiguous in this context. Because this is one of the listed meanings:

(computing) Of software: established to be relatively free of bugs, as opposed to a beta version.

https://en.wiktionary.org/wiki/stable#Adjective

It also lists your meaning. I’m not saying that you’re wrong. I’m saying it’s ambiguous, i.e. there’s two meanings that could apply here.
Well, and personally, I do feel like more people will interpret “stable” to mean bug-free here, because Debian is a piece of software to them.

Yeah, but that’s exactly the meme that I’m talking about.

It’s always ambiguous what is meant by stability. And as soon as you complain about Debian actually breaking very easily, folks will readily tell you about the technicality that it just means it doesn’t change very often.

Having btrfs+snapper set up by default sounds good. I wish Debian-based distros in general would finally make a move there. It’s a bit of a meme that folks laud Debian for its stability, but you can easily break it with one wrong command.

And who knows, maybe TuxedoOS adopting it can serve as a proof of concept and get Debian itself to adopt btrfs sooner.

I also like how they claim it’s completely different to KRunner and then, save for OCR and whatever “circle to search” is, these are all features in KRunner.

More of a general tip: https://reuse.software/ can help to get this stuff right.

Pretty sure, it won’t be able to help with your use-case of copy-pasting code snippets, though (it’s not at all trivial to know which parts you copy-pasted), but it gives you a structure to work with and can automatically annotate new source files.

Hmm, I’m not sure you do. It’s a pun. “Just dua lipa” sounds like “just do a leap-a”, like someone with an Italian accent might sound like.

Took me a moment to get it, too, though…

Looks like a typo: !datahoarder@selfhosted.forum

df only shows partitions, whereas du adds up the file sizes in the directory you specify.

So, in particular, if you want to find out what’s taking up so much space, you can repeatedly run du -sh * and cd into the largest directory.

Hmm, don’t think you should need to compile it yourself? They offer downloads for Windows and macOS, and on Linux, the normal Firefox from the distro sources generally already has this compile flag enabled, as far as I know.

I mean, I actually never had a need for unsigned extensions, so never tried it myself, and obviously I wouldn’t know, if every single distro does it the same.

As for wanting to contribute to the statistics, I do understand that desire. But I believe, for the most part, it’s the user agent that counts. I doubt, they would set that to something else in the unbranded builds…

Posteo is mostly just a deployment of RoundCube.

Mind you, I do assume they have deployment scripts and code for handling payments and such, which aren’t open-source.

I also have to say, though, a hosted service using open-source is *nice*, because they will hopeful contribute back to that, but it’s not much beyond that.
You cannot know, whether they actually run the software that they publish in their repository, so auditing the open-source code hardly tells you anything about the trustworthiness of the actual service.

It is possible on Firefox Stable, by using an unbranded build.

Yeah, we’re also still on 24.04. And from what I’ve seen, 26.04 will ship with Plasma 6.6, not 6.7.

Maybe the kubuntu-backports repository can give it earlier than that, but yeah, I don’t even know yet, when we’ll upgrade to 26.04.
Good point, though, about the .01 release. Maybe I’ll start bothering IT when that’s out.