vmchale

43 ms·

451.
by vmchale 7y ago
> audit the DAG of past commits That's a branch-based way to look at it.
452.
by vmchale 7y ago
Nope! But I need reliable version control, so I will continue using darcs until I hear Pijul is more reliable.
453.
by vmchale 7y ago
Compared to git, it relies on patches rather than a branch approach. I think it's more intuitive. Less opportunities to do silly things and some things that are punishing in git are not with patch-based version control.
454.
by vmchale 7y ago
> the tooling around it is atrocious. Compared to something like Bitbucket I've yet to find anything comparable for repository hosting. I use it for projects where some of github's features are frivolous, but it would be a bit
455.
by vmchale 7y ago
Darcs has a some pathological algorithms. IME, darcs has never lost a repo of mine and Pijul has. So I'm not clamoring to use pijul instead - I don't use darcs in a way that runs into the pathological cases anyway.
456.
by vmchale 7y ago
Last I tried it (a couple years ago), repositories got corrupted. I'd like something to replace darcs, but so far I haven't found it.
457.
by vmchale 7y ago
infinite minimum wage = zero suicide
458.
by vmchale 7y ago
That's true. Here in the United States scientists were sued for telling the truth, which caused immense headache.
459.
by vmchale 7y ago
You can use Futhark; it compiles to PyOpenCL and then can be called by python fairly fluently. Accelerate is a nice example of an array library with a JIT backend.
460.
by vmchale 7y ago
> It seems like further advances in this space either require giving the compiler a better understanding of what is "actually" shared In the context of arrays, it helps when programmers use higher-order combinators/functio
461.
by vmchale 7y ago
> I would dearly love to see a solution that allows me to code in a functional style without sacrificing performance. Uniqueness types would make mutation easier to track: you'd still need to write an imperative style, but less to w
462.
by vmchale 7y ago
I think Futhark approaches this with uniqueness types. A bit strange (maybe that's just me?) but it works well with arrays. Also it compiles to Python nicely.
463.
by vmchale 7y ago
> I guess if all you care about is the market going up, you're better off if people know about less risks. It's even more crass: the interests of a few oil/coal companies allow us to paper over the very real dangers of fos
464.
by vmchale 7y ago
Scientists have warned about this for decades, but they were smeared or even harassed by bad-faith actors. I wish that those responsible would be punished or at least demoted.
465.
by vmchale 7y ago
>I'll never understand how Python became the more widely appreciated language of the two NumPy, maybe? Lots of resources make it easy to get small impressive projects off the ground quickly.
466.
by vmchale 7y ago
> Every time I read Ruby code it almost puts a smile on my face, it seems such a joy to write and think in Compared to Haskell, its performance is not good. Same with J for that matter.
467.
by vmchale 7y ago
ML is changing the world.
468.
by vmchale 7y ago
I think the gcc backend is quite attractive! rustc still has some cases it doesn't catch the optimizations gcc does.
469.
by vmchale 7y ago
> Rust came to be to address the issue with security and there is not really an alternative. Yes people use it for cli tools or backend services but that’s just because they are riding the hype train. I like Rust for CLI tools because it
470.
by vmchale 7y ago
Thanks!
471.
by vmchale 7y ago
don't land on those runways!
472.
by vmchale 7y ago
Plus affine types have gone mainstream. Good stuff, but I'd like memory safety!
473.
by vmchale 7y ago
> value performance over total memory safety Is there any evidence this is a tradeoff? What of linear types?
474.
by vmchale 7y ago
That's still not the same as affine/linear types, which other languages offer. Also, testing helps detect bugs, but cannot guarantee security.
475.
by vmchale 7y ago
I'm not surprised this happened, but it is bitter. Glad to see the failure is being pointed out to the public.
476.
by vmchale 7y ago
It looks like it's not memory safe? What does this offer over existing options?
477.
by vmchale 7y ago
that's why I use chrome
478.
by vmchale 7y ago
But why is Rust a problem there? Surely it is not fundamentally incompatible with 32-bit architectures?
479.
by vmchale 7y ago
In my experience, there have been times rustc underperformed relative to both gcc and clang. In the past performance infelicities have been fixed, but there are some that remain.
480.
by vmchale 7y ago
rustc still has a number of cases where it performs poorly compared to gcc and even clang

More ›