gf000

43 ms·

721.
by gf000 10mo ago
I think it's more of a user space issue, that the UI doesn't degrade nicely. The kernel just defaults to a more server-oriented approach.
722.
by gf000 10mo ago
The feature itself works. There are just hardware that is buggy and don't support it properly. That's a vastly different statement.
723.
by gf000 10mo ago
> https://web.eecs.umich.edu/~weimerw/2008-415/reading/bacon-g... They are different approaches for the same thing: automatic memory management. (Which is itself a not trivial to define concept) One tracks
724.
by gf000 10mo ago
> Reference counting gives you eager destruction. GC cannot. Tracing GC can't. Reference counting, which is by definition a GC can. It's like insects vs bugs. And destructors are a specific language feature. No one says that
725.
by gf000 10mo ago
It absolutely is (and as per another thread under another post, both pjmlp and me are notorious for correcting people on this specific point)
726.
by gf000 10mo ago
I agree with your first point, and I didn't say anything contrary to your second. My point is about crossing the misunderstanding between the "two camps".
727.
by gf000 10mo ago
> The "lingua franca" for language bindings is the C ABI which every other OS's platform libraries (Win32/Cocoa/GTK) support. Which is more like half of an API, it doesn't specify lifetimes so basically anyt
728.
by gf000 10mo ago
More like "mostly known from stating absolutely ridiculous claims", though I heard they went back on most of them and now are more realistic - but also much less interesting.
729.
by gf000 10mo ago
While I am usually the one that also goes in and correct people incorrectly calling RC not a GC, the important distinction here is that Rust (and C++) has the necessary language constructs to be able to implement ref counting entirely as a
730.
by gf000 10mo ago
Low and high level are not well-defined concepts. One, objective definition is simply that everything that is not an assembly is a high-level language - but that is quite a useless def. The other is about how "deeply" you can cont
731.
by gf000 10mo ago
Value types would solve that issue flawlessly.
732.
by gf000 10mo ago
> Simple package based distros like Arch basically just extract archives. Very few packages trigger post-install steps which usually just (re)generate something like initrd. Sure it's simple, just like 80% solving the problem is usu
733.
by gf000 10mo ago
Is a wasm sandbox, or the browser not a sandbox, independent of the kernel?
734.
by gf000 10mo ago
I was talking about a type system with affine types, as per the topic was Rust specifically. I compared it to a statically typed language with a GC - where the runtime takes care of a property that Rust has to do statically, requiring more
735.
by gf000 10mo ago
We don't actually have empirical evidence on the topic, surprisingly. It's just people's hunches.
736.
by gf000 10mo ago
Constraints Liberate, Liberties Constrain. (I also recommend watching the presentation with the same title)
737.
by gf000 10mo ago
> but one reaps the benefits later thanks to more confidence when refactoring or adding code. To be honest, I believe it makes refactoring/maintenance take longer. Sure, safer, but this is not a one-time only price. E.g. you decide
738.
by gf000 10mo ago
That's only somewhat true if we are talking about the same sandbox nested (which would be quite dumb to do). Escaping two different sandboxes are multiple times as hard, and a sane sandbox is not trivially picked, see web browsers and
739.
by gf000 11mo ago
Because no one cares about 3% of desktop users, when much juicier and larger targets also exist.
740.
by gf000 11mo ago
> My browser got slower Is it the browser, or the websites getting more and more resource-intensive as hardware (and also browser optimizations) got better and more powerful?
741.
by gf000 11mo ago
That's pretty far from the truth though. It is absolutely inferior to a database-like binary format for querying, sorting, searching etc. It's a good tool for certain jobs.
742.
by gf000 11mo ago
Runtime checks can only validate code paths taken, though. Also, C sanitizers are quite good as well nowadays.
743.
by gf000 11mo ago
Runtime memory issues probably have a significantly higher total cost. But of course that doesn't negate this cost, but building can always be done on another machine as a way to circumvent the problem.
744.
by gf000 11mo ago
> that C can express that Rust cannot The reverse is probably more true, though. Rust has native SIMD support for example, while in standard C there is no way to express that. 'C is not a low-level language' is a great blog pos
745.
by gf000 11mo ago
I do agree that each company's influence in case of the kernel is much lower, than Google's relevance in Android, but there are other big-ish players in the space as well, like Samsung.
746.
by gf000 11mo ago
It does on Intel, AMD and a bunch of other huge corps though
747.
by gf000 11mo ago
Making it an expression would indeed by nice - I do like that in Kotlin, though it's still not particularly short there (but given that there are not really checked throws there, it's easy to just have a lambda for that). I believ
748.
by gf000 11mo ago
Well, it's far from trivial to introduce such a change if you don't want to throw away 3 decades of code/assumptions. Of course null safety would be very welcome, I'm with you on this. As for unchecked exceptions, that m
749.
by gf000 11mo ago
Compared to what? Because compared to go which has not one but 2 nulls, and an even more anemic type system it is surely much better. > the type system definitely doesn’t help you write correct programs. It surely helps significantly. Yo
750.
by gf000 11mo ago
Then just return a value representing what you want, instead of breaking a convention and hacking something and hoping that at use site someone else has read the comment. Also, just let the use site pass in (out variable, pointer, mutable

More ›