I don't think it's quite a titanic enough endeavour to put slow in quotes. It's been in development for 16 years and only got a stable support for screenshots a few months ago. Does drag and drop work yet?

IMO at this point it is reasonable to say that the idea of having a shared protocol and then making every desktop environment implement the entire display server was not a good one. The Linux community does not have enough manpower to make that work well.

That's just assembling a forge from pieces...

do people really use Lisp family for AI dev tho?

Not since the 70s. They almost all use Python.

FP has absolutely nothing to do with modern AI.

That's a fair point. I don't think that's the case here because he talks about all the bad ways he prefers to receive contributions (email, patch files, git bundle etc.).

Pretty dumb not to use a forge. Adds a huge barrier to contribution for little benefit. None of the reasons he gives make sense.

Maybe a good option for projects that you don't want anyone else to contribute to, but then why make them open source in the first place?

Not using GitHub because it's proprietary is an especially illogical stance. Virtually all websites are proprietary.

You can't use that to assert that your view about not having something is correct.

IMO a bug tracker and PR review system are essentially and cannot be taken away. It would seem like most of the world agrees with me.

SiFive P670

From what I can tell this might be almost as fast as a RPi 5 (single core). Which is almost as fast as my 12 year old i5-2500K. I guess we'll find out when it is available.

I definitely think we'll get an M1/Zen class RISC-V CPU eventually but I doubt this is it.

They're a kind of backup...

Yeah good learning experience. We've all created projects and then found they already exist.

I guess you could say "the name"!

In fairness the first and second results on Google point to the crc32 tool...

https://askubuntu.com/a/303666

 undefined

    
$ sudo apt-get install libarchive-zip-perl
$ crc32 my_file

  

Again not a great package name and it does require Perl, but in Linux at least that's a less painful dependency than Python.

Two things:

1. Desktop requires mature CPUs (large out-of-order designs with high IPC) and there just aren't really any of those yet. They're starting to arrive (e.g. XiangShan which is even open source!) but as far as I know there isn't a single chip available to buy that's faster than a Raspberry Pi 4.
2. Microcontrollers can get away with only the basic instruction set (add, multiply, load, store etc.) but for high performance you need a ton of extensions that are considered standard. x86 and ARM have had decades to build them up but in RISC-V a lot of them are only recently ratified (e.g. Vector) or still in the process of being defined.

I would say we might see cheap Android phones with RISC-V CPUs in maybe 5 years. Though there's an additional difficulty there in that you need to emulate ARM for games, and I don't think anyone is working on that.

It's definitely improving. I thought the same as you but I looked through my recent ChatGPT prompts and it's actually decent now, at least at simple/throwaway tasks. It doesn't stand a chance at the niche domains of my actual job.

Yeah I agree. It's often easier to start from something that's wrong than a blank page.

Sounds dumb but maybe they are trying to say they don't like languages where everything is implicitly a references (Java, JavaScript, Python, etc.) and they prefer languages where references are explicit and objects can be easily copied (C++, Rust, Go).

In which case I totally agree. Making everything a reference is a lazy hack to get decent performance at the cost of unintuitive semantics (pretty good interview questions!) and making actual copying unreasonably difficult.

Until relatively recently the recommended way to deep copy an object in JavaScript was to serialise it to JSON and back.

What's the advantage of compiling to Rust here? Maybe it would be faster if they just skipped straight to LLVM.

Yeah I think it's pretty clear what's going on at this point.

I feel like the best option at the moment is egui. It's native. Works on the web too. Very easy to get up and running. The things I don't like about it:

• I personally think the default style could do with improvement. Mainly it's way too cramped. There's a happy middle ground between no padding and bootstrap. I mean Win32/Qt/etc. got this basically right.
• Immediate mode. Yeah it's easier, especially with Rust, but ... it's surely not how it's supposed to work.
• The low level drawing API (like if you're making custom widgets) is surprisingly amateur. Not something I'd want to target if I'm spending a lot of time e.g. writing a custom map widget or git graph or something.

I also tried Slint. Like the author I think the license is pretty reasonable. But it is pretty involved to set up a project and since it compiles everything from source it can take a very long time for a clean build of hello world. It's like if you were using Qt but instead of a binary package the sources are just included in your app.

Also I have bad experiences from QML (Javascript 🤮, weird scoping rules, etc.) but hopefully they learnt from their experience.

Looking forward to the 2030 edition anyway!

Windows is fine. Get off your high horse.

Yeah unfortunately I have to use Linux for work. I have considered WSL but... I dunno even with its many bugs I think WSL is probably worse. I have no idea how you get X apps working under it for example.

I don't see why that would cause lock ups? I'm pretty sure it's just a driver bug. Didn't used to do it but I upgraded the kernel recently and then it started.

Interesting thread anyway - do you know if they ever fixed the defaults?