Test PR

[timokau]

This is another PR test.

I could initialize with "needs work" as well, but lets go ahead and get a review.

@marvin-mk2 needs review

[ghost]

Hi! I'm an experimental bot. My goal is to guide this PR through its stages, hopefully ending with a merge.

I have initialized the PR in the needs_review state. This indicates that you consider this PR good to go and makes it easily discoverable by reviewers.

Once a reviewer has looked at this, they can either

• request changes and instruct me to switch the state back (@marvin-mk2 needs work)
• merge the PR if it looks good and they have the appropriate permission
• switch the state to needs_merge (@marvin-mk2 needs merge), which allows reviewers with merge permission to focus their reviews

If anything could be improved, do not hesitate to give feedback.

[timokau]

@marvin-mk2 needs work

[bhipple]

For the approved and needs work stages, why use this instead of the GitHub builtin PR review statuses?

[doronbehar]

I think this is better because it's more verbose - you can see the "event" of the label being changed right near the comments, and you get an email notification for the '@bot needs work` message so it's even clearer what should be the status of the PR even without opening GitHub.

[turion]

Can arbitrary people trigger these messages?

@marvin-mk2 needs merge

[turion]

@bhipple What are the builtin PR review statuses? Can anyone who reviewed or created a PR change them?

[timokau]

For the approved and needs work stages, why use this instead of the GitHub builtin PR review statuses?

Basically because those stages are pretty limited and ambiguous. "approve" can mean anything from "I like the change based on the PR description" to "I have read the diff & done extensive tests". With this bot, people have to be much more explicit what they mean. Its also more extensible and flexible. PR authors can reset the status if a reviewer requests changes and then becomes unresponsive.

We might incorporate the builtin approval process (timokau/marvin-mk2#1), but I'm not entirely sure that is a good idea due to the ambiguity.

Can arbitrary people trigger these messages?

Yes, the intention is to give everybody the tools to manage their own PRs state and give useful reviews. One addition might be to keep the PR author from setting "needs merge", but that's not very high priority.

[timokau]

Probably worth mentioning explicitly: Probably marvin will only respond to comments on PRs with the marvin tag. It adds that tag if it sees @marvin-mk2 needs work or @marvin-mk2 needs review in the first message. All other PRs are unaffected, which is what I meant by "opt in" in the discourse thread.

[timokau]

/status needs_review

[timokau]

/status needs_merge

[marvin-mk2-test]

Sorry, you cannot set your own PR to needs_merge. Please wait for an external review. You may also actively search out a reviewer by pinging relevant people (look at the history of the files you're changing) or posting on discourse or IRC.

[timokau]

/marvin opt-in
/status needs_review

[marvin-mk2-test]

Hi! I'm an experimental bot. My goal is to guide this PR through its stages, hopefully ending with a merge. The stages are

• needs_review, if the author considers this PR ready
• needs_work if the PR in its current form is not ready yet. Maybe the reviewer requested changes, there is an ongoing discussion or you are waiting for upstream feedback.
• needs_merge can be set by reviewers who do not have merge permission but would merge this PR if they could.

Anybody can switch the current state with a comment of the form /state <new_state_here>.

Feedback and contributions to this bot are appreciated.

[timokau]

/marvin opt-in
/status needs_reviewer

[timokau]

Yay URL encoding. I added timestamps to the search in NixOS/nixpkgs#92719, which include "+" delimited UTC offsets. The + is interpreted as a space/parameter separator, so the search fails.

Sorry again symphorien & glittershark. You may still want to unsubscribe from this PR since I will continue to use it for testing.

[pjjw]

Yay URL encoding. I added timestamps to the search in NixOS/nixpkgs#92719, which include "+" delimited UTC offsets. The + is interpreted as a space/parameter separator, so the search fails.

Sorry again symphorien & glittershark. You may still want to unsubscribe from this PR since I will continue to use it for testing.

not sure that pull request is the right one, but would love marvin's attention on it!

[timokau]

Oh yes, I meant timokau/marvin-mk2#79 ;)

Unfortunately fgaz seems tight on time right now, so feel free to set back to needs_reviewer. We're also a bit tight on reviewers though, so if you want to help out and shorten the queue have a look at https://github.com/NixOS/nixpkgs/labels/needs_reviewer.

[marvin-mk2-test]

Reminder: This Pull Request is awaiting review. If you are the assigned reviewer, please have a look. Try to find another reviewer if necessary. If you can't, please say so. If the status is not accurate, please change it. If nothing happens, this PR will be put back in the needs_reviewer queue in one day.

[marvin-mk2-test]

Reminder: This Pull Request is awaiting review. If you are the assigned reviewer, please have a look. Try to find another reviewer if necessary. If you can't, please say so. If the status is not accurate, please change it. If nothing happens, this PR will be put back in the needs_reviewer queue in one day.

[marvin-mk2-test]

@glittershark please review.

[marvin-mk2-test]

@glittershark please review.

[timokau]

Sorry @glittershark. Just testing, didn't mean for the bot to ping you. You might want to mute this issue.

[timokau]

/status awaiting_reviewer

[marvin-mk2-test]

Reminder: This Pull Request is awaiting review. If you are the assigned reviewer, please have a look. Try to find another reviewer if necessary. If you can't, please say so. If the status is not accurate, please change it. If nothing happens, this PR will be put back in the needs_reviewer queue in one day.

[timokau]

Test

[timokau]

/status awaiting_changes

[timokau]

/status awaiting_changes