Fi 🏳️⚧️the godot project points out in this post that reviewing machine-spewed PRs is demoralizing as fuck and the reviewers' time is wasted, as the LLMs cannot learn nor grow from the review process.
| GitHub | https://github.com/weavejester/ |
Weavejester
@weavejester
- 336 Followers
- 92 Following
- 205 Posts
@iamnmm I don't think I'd currently trust any AI with code changes, even with tests, unless it was a throwaway script with easily verifiable output.
I've been testing out Claude as a debugging aid recently. It's pretty good at catching basic mistakes or typos, and has been able to suggest leads on some thorny race conditions and performance issues.
However, it's also made a number of errors itself - it's not a substitute for thinking through a problem.
(λ. borkdude)New blog post!
Automatic help and completions in Babashka CLI
https://blog.michielborkent.nl/babashka-cli-help-and-completions.html
I've released TeensyP 0.7.0, a non-blocking TCP server for #clojure, available at: https://github.com/weavejester/teensyp
I've been making quite a few changes on TeensyP as part of my work on Capra, a new Ring adapter for HTTP 1.x that's written entirely in Clojure.
Just released TeensyP 0.5.0. TeensyP is a small NIO TCP server library for #clojure. https://github.com/weavejester/teensyp
@plexus My experience is that LLMs are most useful for tasks that are trivially verifiable. I tend to use them for small, one off scripts or for searching through documentation. I haven't found them that useful for debugging, except as a rubber duck.
Had a great time at #clojuredays. Some very interesting talks, an excellent venue, good food and of course great company.
Joacim Jacobsson
@muep It was done that way to maintain backward compatibility with earlier versions of Ring. Form parameters are a difficult compromise because while parameter names are not required to be unique, in practice it's quite rare to use duplicate names, as its usually better to use explicit IDs than rely on implicit ordering for identification.