Fork me on GitHub
#clj-kondo
<
2023-05-05
>
borkdude12:05:54

Anyone wants to participate in our exciting decision matrix concerning issue https://github.com/clj-kondo/clj-kondo/issues/2063? https://docs.google.com/spreadsheets/d/14bDj85LJwUrImLbt87oz9-_OMfdorU16sT4FqFdBL1s/edit?usp=sharing You can comment in this thread ๐Ÿงต

๐Ÿ‘ 2
woohoo 2
borkdude13:05:10

@U7RJTCH6J Does dewey use :defined-by in its analysis?

phronmophobic17:05:06

The :defined-by key is included in the weekly analysis dumps. I'm not aware of anyone is using the analysis dumps, but there's not really any way for me to know unless someone reached out in some way. Dewey itself only uses :defined-by in https://github.com/phronmophobic/dewey/blob/main/examples/stats/src/stats.clj#L52. It's part of the examples for using dewey data and I use it as part of some statistics I run when I make changes. I look at the stats of dumps from before and after the changes and make sure the results either haven't changed or provide more accurate results.

phronmophobic17:05:17

After reading through the spreadsheet, the only issue I can see is if someone wanted to run some kind of analysis across time to see how codebases were changing, then the fact that :defined-by has a different meaning over time could be annoying. That situational is purely hypothetical and may never come up. Additionally, that's maybe something that should be handled on dewey's end to either rerun all analyses to provide consistent datasets across time or for dewey to provide good documentation about how the data dewey provides changes over time. Just kinda thinking out loud to provide some context.

borkdude17:05:12

option 1 would be to not change the meaning but to introduce another key with the new meaning

borkdude17:05:41

which is I think so far the preferred one

๐Ÿ‘ 2