Ted M. Young / is on Nostr

He/Him. Technical coach. All about Making #Java Code More Testable. #TDD #Refactoring #HexagonalArchitecture #JUnit #AssertJ

Website: https://ted.dev
Newsletter: https://mycmt.dev
Twitch Live Coding: https://JitterTed.Stream
YouTube Videos: https://JitterTed.TV

Last Notes

2024-06-15 20:08:35

8.35 mile run this morning and NO PAIN. Hallelujah.
I've been testing out new foot cushioning to deal with apparent sesamoiditis and today was planning on running 7 miles, but felt so good I just kept running.
#Running

2024-06-15 20:08:35

2024-06-02 19:26:17

- reply

Otherwise, I recommend you treat all ideas (mob/ensemble programming, trunk-based development, TDD) as potential experiments that you can run, trying out those techniques.
You can't read your way to doing things better, instead you have to experience ways of doing things and give those experiments an _honest_ try, not "well, we tried it once and didn't like it". How many new things that improved the way you did things did you like on the first try?

2024-06-02 19:26:17

- reply

2024-06-02 19:16:08

If you're asking for papers in published journals as evidence before you try out a technique or process in software development, first ask if your CURRENT way of working has such evidence and support.
#ExperimentOverEvidence

2024-06-02 19:16:08

2024-05-11 23:16:29

I've been reflecting on LLM tools in IDEs, like Sourcegraph's Cody and JetBrains's AI Assistant, and I think that they (and other LLMs targeted for code generation) suffer from the same problem:
The inability to FORGET.
Unless you have complete control over the data used to train the LLMs, you're going to get old solutions for problems that may not exist anymore, because languages, libraries, and frameworks eventually deprecate and remove code, come up with better (or worse, but different) ways to do things, etc.
It's a bit frustrating to get a solution that works, but is the "old" way of doing things (indexed for loop instead of a for-each, or instead of a stream). It's even worse when the solution worked 3 years ago, but no longer does, or worse is a mixture of modern and obsolete code, making it difficult to separate the two (LLMs are entirely based on "mixing", so this happens frequently).
I'm pretty sure the solution for these kinds of tools is to be trained on better (or at least more recent) data.

2024-05-11 23:16:29