chore(renovate): update semantic commit types in configuration#37658
chore(renovate): update semantic commit types in configuration#37658bircni wants to merge 2 commits into
Conversation
|
depending on #37659 otherwise ci will still fail |
|
This problem is still not resolved: how to handle the "enhancement" label for PRs? |
| "schedule": ["* * * * 1"], // dependency update PRs weekly, vulnerabilityAlerts bypasses this | ||
| "minimumReleaseAge": "5 days", | ||
| "semanticCommits": "enabled", | ||
| "semanticCommitType": "chore", |
There was a problem hiding this comment.
This is already the default, but I guess it's ok to spell it out again explicitely.
I'd keep them at |
Sorry it's unclear to me.
|
|
Just remove it in a future label cleanup. It comes from the default GitHub labels which don't fit in a semantic commit scheme. Same goes for |
What did you mean for that? https://github.com/go-gitea/gitea/blob/main/.changelog.yml#L26 |
|
We've been merging semantic commits since a while now, most PRs are unlabeled because of that. v1.27 release will require a bit of manual categorizing, v1.28 release can be created full off the commit titles alone. |
I am not in the "we" Actually I have objection of "removing enhancement label" The design should be like this in my mind:
But I won't block your decision if most people prefer yours. |
|
In essense this change is almost a noop:
So I think it's actually a slight degradation in functionality and it's better to keep the defaults as-is. |
|
What's your decision to these?
And what's the decision of TOC? |
I suggest the following:
2 nice useful things pop out here:
Comments:
--> Here I would put an enhancement in or other option is to use |
Then I guess you mean "remove the enhancement label" is the answer to #37658 (comment) , right? If so, what are the plan & steps for the removal? |
I'd suggest I open a new branch with the changes then TOC and whoever approves or discusses there |
|
And some real examples, what kinds of these PRs should be?
I think they fit "#37658 (comment)"
Agree |
User-facing behavior improvement. --> fix
UX/text consistency correction. --> fix
Adds externally consumable functionality/API capability. --> feat
Adjusts broken/undesired editor behavior. --> fix
--> purely a UX enhancement --> feat but yes overall an "enhancement" is better - let me try |
|
Well, I am not convinced. The current "label" system works like this:
If every "enhancment" becomes "feature", end users just are not able to find the real new features. |
|
I also think we need an enhancement label in addition to feature. Sometimes a small improvement to the UI or an existing feature fits better under enhancement. It doesn’t mean the previous implementation was wrong, only that the new implementation is an improvement. We could also add a highlight label, so that important features, enhancements even performance changes can be marked and tracked separately. I will always label a PR when possible. It doesn't take more than 1 second. |
|
+1 for keeping enhancement. @wxiaoguang the way you described the current state is also my understanding, and I agree with your conclusion that if every non-bugfix becomes "feature" then it's hard to find the actual new features. |
|
Started here |
For vulnerability alerts it makes sense to keep fix since those are actual security fixes but for others a dependency update is a chore