Sergei Zharinov
Sergei Zharinov
> > > Should be possible, I'm not interested in implementing that though. > > > > > > But are you interested in having it working? I may take...
I have no WIP, but the outline of what needs to be done is something like that: Extract level: 1. Extract `scalaVersion` (done), `sbt` version and `scalajs`/`scala-native` versions 2. Provide...
I don't think it's directly related to `prCache`. What I suppose is that we should stop using hash comparison for deciding whether cache is modified or not: ``` > JSON.stringify({...
But storing second copy of the cache is memory-consuming, so instead we have to figure out how to `JSON.stringify` with all keys being sorted
Okay, then my steps to reproduce: 1. Have a repository with PR created by Renovate 2. Add couple spaces to the `renovate.json` 3. Run Renovate again and see it does...
Well, now I get it. This was intentional from my side, since I think it's not worth the struggle to implement additional cache modifications on create/update/delete PR as it doesn't...
I think it's not our main performance bottleneck.
But I like how it led us to improve repo cache correctness in another aspect :)
It uses repo cache and fetches/synchronizes only freshly updated PRs.
Okay, my premise was wrong, I'll fix this problem