remexre

38 ms·

451.
by remexre 5y ago
Taking a look at that, I think shapes would still work; [1] has some background material on them, but dynamically changing data is what shapes are for; you check if you should update an object's shape when assigning to it, and can JIT
452.
by remexre 5y ago
> Classes in Python are basically sparkling Dicts and their memory layout can change at any time so basically everything must be boxed. My understanding was that most JavaScript JITs solved this problem well using shapes.
453.
by remexre 5y ago
Scientific computing is actually a use-case I see this working well in; IMO, it's perfectly fine to have terse code that exactly matches up to equations in a longer paper that justifies correctness, etc. (as long as the paper is easy t
454.
by remexre 5y ago
I... usually hear this sentence sarcastically... but here it really seems as if the author isn't using it as such. I dunno, if you're a proponent of Starve The Beast [0], this is a consistent belief I suppose. [0]: https:/&#
455.
by remexre 5y ago
Three halves > 1 full, sounds like a win to me! /s
456.
Common Lisp HyperSpec: Notes about Style for [Comments](lispworks.com)
1 points by remexre 5y ago|0 comments
457.
by remexre 5y ago
The Bluespec Language is another Haskell-like that compiles to hardware, but I think it's a bit more mature, too. It also has a second non-Haskell-like syntax to avoid scaring those used to Verilog.
458.
The Next 700 Module Systems [pdf](github.com)
4 points by remexre 5y ago|0 comments
459.
by remexre 5y ago
Well, #forth on Libera.chat now :)
460.
by remexre 5y ago
> [...] parameter-passing bugs [...] Ah, yeah, incremental development is the panacea here. > [...] Factor (or PostScript) [...] Yeah, agreed; if you're needing to be able to safely execute untrusted code though, I'd rather
461.
by remexre 5y ago
> Is there a codebase you think of as exemplary that I could learn from? My "one codebase I keep rewriting" is [0]; I don't consider the artifact as useful as the process, though. Particularly, > in Forth I have the sam
462.
by remexre 5y ago
> because you're a novice, because you have better ways to spend your Saturday nights than disassembling core dumps My experience is that with a development methodology that emphasizes early-binding, incremental testing, and aggress
463.
by remexre 5y ago
For essentially every form of UB that compilers actually take advantage of, there's a real program optimization benefit. Are there any particular UB cases where you think the benefit isn't worth it, or it should be implementation-
464.
by remexre 5y ago
Typically, the assembly instruction that would do the read in arr[i] can do something like: x = *(y + z); where y and z are both 64-bit integers. If I had int arr[1000]; initialize(&arr); int i = read_int(); i
465.
by remexre 5y ago
I think there's a general push to have dirfd versions of all syscalls, since it makes it easier to write libraries that don't break when the application or another library chdir()s.
466.
by remexre 5y ago
If you care about latency, not performance, I think it's fine? You can do a Cheney semispace collector in which it's safe to allocate in interrupt handlers if you have control over the ABI (and you've got a few registers to s
467.
by remexre 5y ago
I really like Z3 for showing that my weird bitwise hacks (e.g. [0]) are correct -- it's especially useful that I can show not only that a function is correct, but also that it approximates the correct answer to within whatever bounds I
468.
by remexre 5y ago
> What about user-supplied JSON schemas? You can't add types at runtime. Right, well, since they're validators anyway, might as well represent them as a defunctionalized validation function or something. Agreed that this is mor
469.
by remexre 5y ago
In general, JSON Schemas are (wrongly, in my view...) validation-oriented rather than type-oriented (for notions of types that would be familiar to Haskell, Rust, or Common Lisp programmers). I think that schema in particular could be repre
470.
by remexre 5y ago
HVF does build with Nix (with an overlay), so the build experience doesn't suck. [1] for the base, then apply [2] to SLIRP if you don't want to use HVF's network adapter (which needs root). I'm happily running FreeBSD an
471.
by remexre 6y ago
GraalVM (and SubstrateVM) is actually fixing a lot of these; I had a short-lived process that I AOT compiled with SubstrateVM, and I spent a lot of time checking I had done it right and not somehow accidentally compiled in my input data as
472.
by remexre 6y ago
Heck, I prefer surimi to crab!
473.
by remexre 6y ago
Which kind of sucks for everyone else at UMN, including people who are submitting actual security fixes...
474.
by remexre 6y ago
WRT floating-point, worth noting for those who haven't read the patches: - Kernels generally don't want to use floating-point, because saving and restoring the floating-point registers is fairly expensive. - Without some pretty ag
475.
by remexre 6y ago
Yeah, I have https://lkml.org/lkml/2004/12/20/255 bookmarked as an example of great maintainership
476.
by remexre 6y ago
Some technologies (e.g. Secure Scuttlebutt) have the ability for subgroups to exist without aggressively recommending content to every user, as an inherent property of their design; I think "should Peertube adopt some aspects of these
477.
by remexre 6y ago
Yes, it's automated.
478.
by remexre 6y ago
Can't comment on sibling, but https://www.skypack.dev/ lets you import many npm packages; I haven't tried the AWS library, though.
479.
by remexre 6y ago
Well, Deno's not really a framework, it's a runtime (like node.js). If you're writing Typescript, it makes that way easier. If you're doing lots of IO, it makes that easier (promises). If you want sandboxing, it has that
480.
by remexre 6y ago
> Today at GTC 2021 NVIDIA announces its first CPU Wait, Nvidia's been making ARM CPUs for years now; most memorably Project Denver.

More ›