dbaupp

15 ms·

121.
by dbaupp 6y ago
Ah, thanks for the clarification!
122.
by dbaupp 6y ago
It’s enough of an issue for C at least that a fully formally verified compiler exists: https://en.m.wikipedia.org/wiki/CompCert . That is, there’s proof that its (limited) optimisation passes are correct. There’s other
123.
by dbaupp 6y ago
I think there's another perspective on doctests, that seems more convincing to me: doctest functionality allows testing for the documentation, not a way of writing tests in documentation. In particular, it's great if the documenta
124.
by dbaupp 6y ago
It’s used flippantly at times, but lock-free is a technical term referring to a guarantee of global progress https://en.m.wikipedia.org/wiki/Non-blocking_algorithm (I think it’s also sometimes used as a synonym for non
125.
by dbaupp 6y ago
Very interesting! This seems somewhat similar to the recently published GIFnets[1]. However, I believe GIFnets is training a reusable network to a predict palettes, and pixel assignments, while this post is focusing on optimising the "
126.
by dbaupp 6y ago
Ah, furthermore, looking at a comment below, it seems ATS's support for C libraries is almost identical to Rust's: specify a list of functions and their signatures on the ATS/Rust side. From your adoring descriptions, I had b
127.
by dbaupp 6y ago
It's moving the goal posts and changing the point of the discussion. We can equally well say that that "Rust formally verifies your safe functions", and that this reduces how much code review is required... it's a matter
128.
by dbaupp 6y ago
That helps with the "library availability" sub-point specifically, yes. However, there's a more than that when evaluating a language, as suggested by my comment above. In any case, if one is discussing modern and safe languag
129.
by dbaupp 6y ago
Popularity and momentum translate into (and are proxies for) important things like: library availability, long-term maintenance and support, more edge cases are explored (so less “research”, breaking new ground and bugs when going off the b
130.
by dbaupp 6y ago
It looks like it's used on about 6% of PRs: since the start of this year, 204 PRs[0] match a search for "@rust-timer queue", among 3393 total PRs[1]. [0]: https://github.com/rust-lang/rust/pulls?q=is
131.
by dbaupp 6y ago
> It's the same for any other kind of test; your tests should pass before you merge, not after Rust is very good about testing before merge, not after. It somewhat pioneered doing complete testing before merge (at least, built sig
132.
by dbaupp 6y ago
The grandparent didn’t say anything about why C dependencies will often exist: in almost all cases, it is not because they must exist due to missing functionality in Rust, but because it is easier to use a C library than rewrite/tran
133.
by dbaupp 6y ago
Unsafe generic code likely cares, where it is coercing a &T to a *const T and then relying on pointer stability/consistency. It will also cause difficulties with methods like ‘.capacity()’, because that info is lost with &str.
134.
by dbaupp 6y ago
At the very least, retrieving data from a &String will involve two pointer lookups (one for the &, one for the internal pointer in the String) whereas plain String and &str involve just one. In addition, since the String object
135.
by dbaupp 6y ago
&String cannot, by itself, handle a substring because it has to point to an (immutable) whole String struct in memory. Any extra syntax or types to package up subslice boundaries would almost certainly end up as something similar to &am
136.
by dbaupp 6y ago
Calling just “memory safety” correctness is a low bar and misses all sorts of other correctnesses (type safety, progress, matching-the-spec, etc.). However, it is a fundamental property: without memory safety, no other correctness can be gu
137.
The customer is always right: user observation is an engineering superpower(medium.com)
2 points by dbaupp 6y ago|0 comments
138.
Do I know you? Flexible un- & semi-supervised graph models with DeepGraphInfomax(medium.com)
3 points by dbaupp 6y ago|0 comments
139.
Show HN: StellarGraph – Python+TensorFlow Library for Machine Learning on Graphs(github.com)
1 points by dbaupp 6y ago|0 comments
140.
Faster ML on larger graphs: how NumPy and Pandas slashed memory and time(medium.com)
4 points by dbaupp 6y ago|0 comments
141.
Git worktrees and pyenv: developing Python libraries faster(huonw.github.io)
3 points by dbaupp 6y ago|0 comments
142.
GitPlus: Lab extension to create GitHub commits and pull requests(github.com)
1 points by dbaupp 6y ago|0 comments
143.
Better notebooks through CI: automatically testing documentation for graph ML(medium.com)
2 points by dbaupp 7y ago|0 comments
144.
Practical Challenges for Ethical AI (2019)(gradientinstitute.org)
2 points by dbaupp 7y ago|0 comments
145.
Graph Neural Network model calibration for trusted predictions(medium.com)
1 points by dbaupp 7y ago|0 comments
146.
by dbaupp 7y ago
Without the exclusivity, it would be easy for a reference to be invalidated; the classic example is let nut v: Vec<T> = ...; let r: &T = &v[0]; v.push(...); // A foo(r); // B The referenc
147.
Intro to Rustc's Self Profiler(blog.rust-lang.org)
3 points by dbaupp 7y ago|0 comments
148.
by dbaupp 7y ago
The benchmarked spin-locks are using it, via https://doc.rust-lang.org/std/sync/atomic/fn.spin_loop_hint.... Implementation: https://doc.rust-lang.org/src/core/hint.rs.html#64-93
149.
by dbaupp 7y ago
How does this differ to https://github.com/hadolint/hadolint ?
150.
by dbaupp 7y ago
Why does the Bernoulli formula take exponentially long in the exponent? Python’s native form (using either an integer or float exponent) and even an arbitrary percision computation to 1000 decimal digits using mpmath (also with either an in

More ›