Add some guidance on rollups and priorities. #402
Conversation
|
cc @Manishearth |
src/compiler/reviews.md
Outdated
| - Your PR is not landing possibly-breaking or behavior altering changes. | ||
| - Feature stabilization without other changes is likely fine to | ||
| rollup, though. | ||
| - `rollup=maybe`: Use this if you don't have a high confidence that it won't |
There was a problem hiding this comment.
Note. this is the default if you have not mentioned antyhing
src/compiler/reviews.md
Outdated
| specifically, as it would be hard to identify as the cause from a | ||
| rollup). | ||
| - Has a high chance of failure. | ||
| - A high-priority issue that needs to land ASAP. |
There was a problem hiding this comment.
This should just be p=5 or something
| - Your PR is not landing possibly-breaking or behavior altering changes. | ||
| - Feature stabilization without other changes is likely fine to | ||
| rollup, though. | ||
| - `rollup=maybe`: This is the **default** if you do not specify a rollup |
There was a problem hiding this comment.
worth somehow clarifying that this is milder than iffy
src/libs/maintaining-std.md
Outdated
| @@ -248,3 +248,4 @@ Where `unsafe` and `const` is involved, e.g., for operations which are "unconst" | |||
| [RFC 1023]: https://rust-lang.github.io/rfcs/1023-rebalancing-coherence.html | |||
| [RFC 1105]: https://rust-lang.github.io/rfcs/1105-api-evolution.html | |||
| [Everyone Poops]: http://cglab.ca/~abeinges/blah/everyone-poops | |||
| [rollup guidelines]: ../compiler/reviews.md | |||
There was a problem hiding this comment.
| [rollup guidelines]: ../compiler/reviews.md | |
| [rollup guidelines]: ../compiler/reviews.md#rollup |
src/compiler/reviews.md
Outdated
| - `rollup=maybe`: This is the **default** if you do not specify a rollup | ||
| status. Use this if you don't have much confidence that it won't break | ||
| tests. This can be used if you aren't sure if it should be one of the other | ||
| categories. |
There was a problem hiding this comment.
I assume rollupers likely add rollup=maybe PRs to rollup. So, I think it'd be nicer if we mentioned it and encouraged members to use iffy rather than maybe if there's any concerns?
There was a problem hiding this comment.
Hm, I don't really know. Like, I imagine there are quite a few PRs that do not fit the category of "iffy" (that is, "risky"), but the reviewer doesn't have enough confidence to mark it "always". Or, what is probably more common, is "I don't know what it should be", which I think is fine since I don't think we can expect everyone to be deeply familiar with the process and hazards of rollups.
@Manishearth any opinions?
There was a problem hiding this comment.
I basically think folks should never be explicitly marking things as rollup=maybe, it's the default, the reason you do it is to unmark something else from a diferent rollup level.
|
Is there any other feedback anyone wants to give? Maybe @Mark-Simulacrum ? |
src/compiler/reviews.md
Outdated
| implications. Example scenarios: | ||
| - Changes are limited to documentation, comments, etc. that is highly | ||
| unlikely to fail a build. | ||
| - Changes are pure refactoring and cannot have performance implications. |
There was a problem hiding this comment.
Hm I might remove the "pure refactoring" label here, it seems not obviously true that can't have performance implications.
|
@ehuss It seems this now has conflicts. Can you rebase this and then I'll merge it in? |
ehuss
force-pushed
the
rollup-priorities
branch
from
ef18414 to
ff27549
Compare
|
Rebased, thanks! |
This adds some guidance on setting rollup status and priority for the rust-lang/rust repo.
I waffled where to place this. I'm fine with moving it wherever. The rollups page was a high candidate, or a new page in the release or infra sections? No one group really "owns" the rust-lang/rust repo, so it is not really clear where it belongs.
I fully expect there to be some changes here, this is just a rough sketch that I wanted to start some discussion. This is based on some discussion at: