Proposal: Align on the microsoft bot (fabricbot replacement) policy
- Dominant language
- Bicep
- Stars
- 3.6k
- Forks
- 830
- Avg merge
- 1d 21m
- Merged PRs (30d)
- 79
Description
## Overview
We have a fabric bot policy that is currently being enforced partially due to some inconsistencies:
- The fabric bot rules are in some cases using labels that don't exist in our repo
- The fabric bot rules need to be tuned to meet our workflow, but the workflow is not documented so there are gaps
## Known Gaps:
- Comments from @alex-frankel:
- > is there logic in fabricbot to not auto-close any issue that already has a milestone? so if an issue has "awaiting response", but has a milestone, then it shouldn't be auto-closed.
- > just noticed a small thing - not urgent. Instead of adding "Needs: Attention" can we instead add "Needs: Triage" so it gets into our queue?
- Comments from @StephenWeatherford:
- > @asilverman The bot is closing inactive bugs as "completed". Is it possible to close them as "not planned"? Thx.
> Example: [Existing resources would be easier to spot and read if "existing" came first, not last · Issue #3361 · Azure/bicep (github.com)](https://github.com/Azure/bicep/issues/3361) (which I re-opened and closed as not completed).
- https://github.com/Azure/bicep/issues/2530#issuecomment-1555088706
> @alex-frankel @asilverman What's the process for users to make sure a bug stays on the radar? In this case, it's marked needing feedback from author (you :-), and does not have triage or a milestone. So even if the author answered, it still would never come up for review again, right?
## Calls for action
1. I think it would be useful for us as a team to take some time to watch -> [Watch 'Day of Learning: Get to know FabricBot, with Kim Herzig, Cliff Koh and Beatris Mendez Gandica' | Microsoft Stream (Classic)](https://msit.microsoftstream.com/video/67c8a3ff-0400-96d0-6be5-f1ea9ba86b8f?referrer=https:%2F%2Fdocs.opensource.microsoft.com%2F) since it really describes in detail how we should work with fabricbot
2. I would like to have a Bicep Discussion and fomalize our workflow in a document so that we can then align the fabricbot features to that source of truth
3. I want to identify a list of issues where we saw the fabricbot do the wrong thing so we can analyze on a case by case basis how we iterate on the fabricbot config so that it doesn't repeat the undesired behvior.
Contributor guide
Assessment
This issue has not been assessed yet.