Jacob Pratt

@jhpratt
1.5K Followers
73 Following
94 Posts

Contributor to Rust's compiler and standard library. Maintainer of Rust crates including time, deranged, powerfmt, num_threads, and num-conv. Third degree black belt in American Kenpo 🥋

jhpratt on most platforms

Sponsor me!https://github.com/sponsors/jhpratt
Websitehttps://jhpratt.dev
GitHubhttps://github.com/jhpratt
Discordhttps://discordapp.com/users/jhpratt
Show threadJacob Pratt1d ago

@to I unironically included a link to the USPS where they can look up my ZIP code and see the three towns it includes. Admittedly I am surprised it's only three; I thought it crossed more town boundaries (albeit slightly).

As I'm sure you know, it normally doesn't matter, but when dealing with financials it's important to be correct!

Jacob Pratt1d ago

GitHub Sponsors is telling me to confirm my identity, or more specifically my address. Only one problem: they assume that there is only a single city/town name for a ZIP code…

No one working on this at GitHub has ever lived in a rural area? My town hasn't had a post office in decades; we use the neighboring town's post office and ZIP code, but that doesn't mean my town doesn't exist!

Jacob Pratt2d ago

Good news/bad news. I found a bug in the time crate, but it's apparently obscure enough that no one has reported it in the 3+ years it has existed. And it's has *never* worked since being introduced, so I am not constrained by semver in my fixing it.

Oh, and I was careful enough to explicitly mark that path as a bug, so the error is unambiguous that it's unintentional!

Show threadJacob Pratt6d ago
@cetra3 I've not looked at this issue specifically beyond knowing it's bevy. But yes, crater is Linux-only.
Show threadJacob Pratt6d ago
@cetra3 What you you mean by "extend the Rust release process"? Adding in more checks? Unless things have changed, there is a crater run before every release. That checks every crate on crates.io and on GitHub (with some exceptions due to flakiness) to ensure the lack of regressions. However, a human has to review the output, and that's where things can be missed or dismissed as acceptable breakage.
Show threadJacob Pratt6d ago
@cetra3 @zyx because tests don't cover literally every scenario? Mistakes happen.
Show threadJacob PrattMar 25
@argv_minus_one @unsafe Yeah, it's a proc macro
Show threadJacob PrattMar 25
@hjvt Knowing the individual personally, I'm taking him at his word.
Jacob PrattMar 25

People got extremely angry over a single blog post having a first draft written by an LLM, despite it being rewritten entirely before publishing. There is at least one accepted RFC that started the same way. As best as I know, that was not public knowledge until now.

I'm deliberately not disclosing which RFC was initially drafted by an LLM, as it's not relevant. Just know that it's far more pervasive than you realize. It's no reason to attack someone, personally or professionally.

Show threadJacob PrattMar 25
@unsafe Parsing format descriptions! I used to do it in a single pass, straight into the final vec, but changed it to a mini-compiler a couple years ago. It made maintenance _much_ easier and allowed me to add new components quite easily. Now I'm adding some optimizations to the macro (but not runtime parser) to eliminate some not-uncommon situations that are a tad slower than necessary.