Just because things are useful doesn't mean they are magically sound.
Ralf in https://github.com/rust-lang/rust/issues/132442#issuecomment-2636065726
Just because things are useful doesn't mean they are magically sound.
Ralf in https://github.com/rust-lang/rust/issues/132442#issuecomment-2636065726
Comparing compile times to C is sort of like comparing jumping off a bridge with a parachute vs. jumping off a bridge. In the latter case, you'll get to the bottom a lot faster, but not very safely.
Rust isn't a "silver bullet" that will solve all of our problems, but it
sure will help in a huge number of places, so for new stuff going
forward, why wouldn't we want that?
Greg Kroah-Hartman, senior Linux kernel developer in the (Linux mailing list)
Treat anything starting with
cargo
as if it iscargo run
. This applies even to commands that do not build anything, such ascargo clean
, and third-party plugins, such ascargo audit
.
Probably a terrible idea, but I enjoy throwing ideas at the wall, and seeing how sharp their broken fragments are.
Did it work? Itâs Rust, so it worked on the first try!
James Calligeros on behalf of Asahi Linux
The perfect reply to all those who complain Rust is too hard:
Rust is not actually the problem. The problem is you.
"Rust Without Tears: A Guide to Avoiding Common Traps" by Krzysztof Grajek https://www.youtube.com/watch?v=Sv6Bswfjnzo
Tongue in cheek of course.
IEEE 754 floating point, proudly providing counterexamples since 1985!
From Highlight Differences In `assert_eq!()` - #4 by jdahlstrom - language design - Rust Internals, by Johannes Dahlström.
Unethical thought of the week: if someone could ensure @compiler-errors get sick more regularly, rustc would run 2x as fast by the end of the year.
But then type system progress would be 0.5x
Hey, that sounds a bit mean.
This week @compiler-errors fixed a high priority unsoundness bug that I reported a week ago Instructions missing from naked_asm blocks. · Issue #139407 · rust-lang/rust · GitHub and no doubt did a bunch of other useful things as well.
If you follow the context PRs from Reddit, they are performance enhancements[1] that compiler-errors did while being sick (according to the PR descriptions). I agree though, this is not at all apparent from the quote alone.
With huge improvements to the compile time in some cases, e.g. 35% speedup for nalgebra
. â©ïž
Yeah, I tried to find some context. Failed.
That's mean. Please don't be mean.
This one caught me off guard and made me chuckle this morning while reading a thorough and thus lengthy post where boringcactus compares a lot of Rust GUI libraries, by writing a minimal Windows app in each:
The common formulation of Postelâs law is âbe conservative in what you emit and liberal in what you acceptâ, and itâd be more recognizable if I used those terms instead of âpickyâ and âsloppyâ. I havenât done that, though, because you should be conservative never and liberal very rarely. Be radically leftist in everything, even technical blog posts that arenât intrinsically political.
I can't make out any sense of that quote. What are they trying to say.
Am I the only one that thinks Postel's law is nuts? I mean being "liberal" in what you accept" says to me that you should try and accommodate all kind of crap as input. Isn't that why systems have been so insecure for decades?
I see the quote as a double meaning joke where "liberal" and "conservative" are interpreted as political views, rather than the adjectives describing a program's IO behaviour. In combination with a bold political statement at the end and the fact that the quote comes from an otherwise technical (and rather long) blog post adds an element of surprise which I find refreshing when reading such material.
And yes, Postel's law is, at least as it's often interpreted, an awful law that makes any ecosystem into a horrific mess, including an ecosystem of one.
A common exception is human-entered data (so long as the "sloppiness" is well defined and documented!), and dealing with an existing ecosystem that's already a mess (nobody wants to use the zip parser that "won't read my files")
Almost all binary formats are best served by having a single canonical form for any set of data, and even better served by having a reference test suite that ensures bit compatibility among implementations.
I did spot the double meaning there. Then I got lost.
From a silly Go vs Rust conversation @steveklabnik.com on Bluesky
I donât think about rust either. Thatâs a compilerâs job