Any Rust engineers want to critique this PR <https...
# random
o
Any Rust engineers want to critique this PR https://github.com/seanome/kmerseek/pull/20? I’ve admittedly (ashamedly?) been “vibe-coding” with Cursor to get the Rust work done because I’m still building up my Rust skills, so I’d appreciate pointers to where the AI has led me astray. Thank you so much!
m
@Matthias Zepper don't you have some rust skills?
m
Some, but I am not familiar with the tool and certainly do not know about good libraries or advantageous patterns for a certain task from the back of my head. Additionally, the PR description is quite brief, so it is difficult for me to understand what the bottleneck was, that it tries to mitigate. The advantage of Rust is, that there is no
None
and the compiler will stop you from making simple mistakes like not considering cases in `if`/`elif`/`else` switches. So during a review, one doesn’t need to bother about those, but this means there are very few low-hanging fruits. It is more about unnecessary copies when you can borrow a value etc. to optimize speed and memory. Maybe Olga can trade a PR review with @Clay McLeod and the people from St. Jude? https://github.com/stjude-rust-labs
c
I don't mind taking a look if I can get some background like Matthias mentioned earlier.
🙏 2
And also maybe an idea of the level of review you want; I can do anything from "here's some high level pointers based on what I see" all the way to how I would review my own engineers code (at times, pedantic).
o
Hi @Clay McLeod, thank you so much for offering! I beefed up the description of the PR to go more into depth of the goal of the project, what this PR specifically does, and my Rust specific review requests: https://github.com/seanome/kmerseek/pull/20. Let me know if you have any questions!
c
Great! As luck would have it, I actually just got to your talk in the NextFlow summit playlist this morning. Haven't finished it yet, but definitely get the broad idea. I'll give it a look!
❤️ 1