you can say that without a spec as well but what does “wrong” mean then? It just means you personally disagree with its behavior.
Nope. Specs can have bugs. Here are the bugs in the C++ spec for example:
https://www.open-std.org/jtc1/sc22/wg21/docs/cwg_toc.html
As I said, specifications are useful and desirable, but the SIL’s dogmatic “no spec = unsafe” is clearly not based in reality.
It’s not because we have tested this program extensively on every C++ compiler, but because the language rules of C++ say so.
Debatable. Saying things in a prose specification doesn’t magically make them happen. Tests and reference models can though.
I also don’t really agree with the SIL requirements that languages need to have rigorous specifications to be safe. Clearly it’s better if they do, but would your rather fly on a rocket controlled by C code or Rust code?
IMO a specification would be really nice to have, but it main purpose is to tick a certification checkbox, which is why the only one that exists was written specifically for that purpose.
PDF writing isn’t too bad IMO, since you don’t need to understand the whole spec. I’ve written a PDF writer for maps from scratch and it was fairly easy and not too much code.
PDF reading though… Yeah I’m happy to leave that to people with more time and use their libraries.
A modern format would be nice, but I don’t think it would be anywhere near nice enough to give up how universal PDF is.
Eh, it practice it works extremely well. I can’t remember a single instance where a PDF document rendered incorrectly.
The format is very old so it’s not surprising it has picked up a few WTFs. I’m happy to keep those hidden below the abstraction.
Nice work. Aren’t wasm proc macros already prototyped in Watt though?
Is there any reason to use this now that Krita exists, sane name and all?
I agree, those are fantastic icons. Very clear.
There’s a “proper” version of this hack called early oom. I haven’t used it though and now that I look at it it seems like it uses the same completely broken “guess which process to kill, who cares if it’s init” system that the normal oom killer uses so your solution sounds better.
Is it so hard to just pause the system and ask the user which app to kill?
Interesting, so that’s sort of customising the image somehow? Does it use an overlay FS or something?
Hmmm I guess this kind of makes sense - most distros push Gnome above KDE (probably because it doesn’t look like this - where’s Tantacrul when you need him?). On the other hand, there’s already Kubuntu…
I’m a bit skeptical about immutable distros too. What if I want to install a package that isn’t already installed and isn’t available as a Flatpak/Snap? Seems like it’s going to run in similar issues to everything else that tries to wade upstream against the bad decisions of the existing Linux packaging zeitgeist, e.g. how Nix has to install everything in one root-owned directory because nobody cares about portable installation.
That’s cool, but in my experience if you get to the OOM killer then 80% of the time it’s too late and your system is basically dead. My laptop hard reboots most of the time when this happens.
Hopefully it works with the early-OOM hacks.
I have 32 GB but it’s not enough. Try opening 8 instances of VSCode, Firefox and Chrome with a few dozen tabs. Unfortunately my laptop doesn’t support 64 GB of RAM.
… in one benchmark.
Yes, but I was talking about Linux in general. I’m pretty sure Gnome at least has commercial backing.
And Linux advocate never say “don’t use Linux; it isn’t a commercial product so it isn’t as good as Windows” do they? They say “you’re an idiot for using Windows; Linux is better”.
GPU reset recovery
Woah catching up to Windows 18 years ago! :D
Tbf I did not think we would ever see this feature. What next, secure login prompts (“press ctrl-alt-del to login in”, which admittedly Windows seems to have dropped)?
This is about Spectre, not about buggy hardware implementations.
Spectre is a fundamental flaw in speculative execution that means it can leak information, so it’s a security vulnerability. Apparently Intel has been imposing draconian requirements on software to work around the issue rather than fixing it in hardware, which is obviously what they should do, but is not at all trivial.
Unless the binary size difference is insane, who would say “oh well we were going to pick the library that wasn’t riddled with security issues but we decided to save 2MB instead, hope that makes you feel better about your $12m cybersecurity fine!”.
Not the kinds of bugs he is talking about. This is about spectre mitigations.
Yes I definitely am. It’s really nice that crates.io is short, and it’s silly to give that up for a miniscule risk of something moderately annoying happening.
Even if the domain goes away we’d just have to all move to a new domain. Annoying but hardly the end of the world. Cargo.io isn’t actually hard-coded in many places. It’s nothing like if github.com stopped existing.
That’s just playing with semantics. They are clearly bugs. They are literally called “defect reports”.
There are many aspects to safety and it’s definitely a good idea to have a spec for a language, but it doesn’t automatically mean safety is impossible without it.
The nice thing about abstraction is that you can talk about software without considering the hardware, more or less. If one says “this software is safe”, it means it’s safe assuming it’s running on working hardware.
It doesn’t always hold up - sometimes the abstraction leaks, e.g. for things like spectre and rowhammer. And there are sometimes performance concerns. But it’s pretty good.