Thing that kills me about "just rewrite in Rust" energy is that yes, if you translate the O(n^2) algorithm to Rust it is indeed faster, but you still have an O(n^2) algorithm. Maybe think about the problem more abstractly first?

11
Share
Share on Mastodon
Share on Twitter
Share on Facebook
Share on Linkedin

@tenderlove Aaron, how dare you? πŸ˜„

0
1y
awwaiid (Brock Wilcox)

@tenderlove maybe YOU should think something something hugs

0
1y
deSilva

@tenderlove but think about all the ruby memory safety bugs you could fix by riir 😎

1
1y
MaybeMyMonkeys

@tenderlove last I checked, managers didn’t want you to think

0
1y
Daryl Manning

@tenderlove Meh. Why not use Crystal? 😊

0
1y
Petr πŸ‡ͺπŸ‡ΊπŸ³οΈβ€πŸŒˆπŸ‡ΊπŸ‡¦πŸ‡¨πŸ‡ΏπŸ‡³πŸ‡±

@tenderlove This kind of thing is surprisingly common even for senior developers. Or they just get bored and want to try something fun. πŸ€·β€β™‚οΈ

0
1y
Paddy O'Brien

@tenderlove what? You mean to tell me that the speed of an operation is determined by its inputs? 😱

0
1y
Thomas Enebo βœ…

@tenderlove Do people use speed as the reason for that statement? I am not sure I have ever seen that. Mostly safety and secondly higher level language features.

0
1y
MichaΕ‚ ZajΔ…c

@tenderlove BuT iTs GoInG To bE FaStEr

0
1y
Mohit Sindhwani

@tenderlove

Back, in my research centre, we often used to say that the best way to make things faster is first to change the algorithm itself if we can. That can have a many orders of magnitude difference. Everything else could also help, if you couldn't change the algorithm.

0
1y
Replies