Page MenuHomeFreeBSD

git-arc: Add a draft mode
ClosedPublic

Authored by markj on Mon, Apr 27, 3:21 PM.
Tags
None
Referenced Files
F155005015: D56664.id176634.diff
Thu, Apr 30, 3:22 PM
F155001516: D56664.id176720.diff
Thu, Apr 30, 2:49 PM
F155001320: D56664.id176660.diff
Thu, Apr 30, 2:47 PM
Unknown Object (File)
Wed, Apr 29, 11:37 AM
Unknown Object (File)
Wed, Apr 29, 9:24 AM
Unknown Object (File)
Wed, Apr 29, 8:05 AM
Subscribers

Details

Summary

Make it possible to create a review without publishing it. This should
be useful when one wants to restrict the visibility of a review, as this
cannot be done via the command line.

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Skipped
Unit
Tests Skipped
Build Status
Buildable 72572
Build 69455: arc lint + arc unit

Event Timeline

markj requested review of this revision.Mon, Apr 27, 3:21 PM
tools/tools/git/git-arc.1
103

Should this be *cannot* be configured prior to publication?

tools/tools/git/git-arc.1
103

Nevermind. After re-reading, I believe what you have is correct.

Tweak the description a bit.

markj added inline comments.
tools/tools/git/git-arc.1
103

I added a bit more clarification there.

tools/tools/git/git-arc.1
100

I tested this by creating a test commit and review with git arc create -d -l -r markj HEAD. It successfully created the review in draft mode. I then used a private browser window to view the new revision URI, and I could see it. Is this expected?

markj added inline comments.
tools/tools/git/git-arc.1
103

Hum. I guess it's "expected", but I didn't realize it worked that way. One has to guess the URL in order to see it (i.e., the draft doesn't show up in reviews.f.o/p/jrm), and auto-added reviewers and subscribers don't get email for a draft, so this approach is still better than nothing, but yes the text here is misleading.

I think the proper way to do what I want is to configure a separate phabricator project with different default settings for visibility. I'd need an admin to set that up. :(

This will be a useful verification step. I sometimes submit a review and then notice that the summary isn't displaying as I intended.

I think the documentation is clear, but I still wonder whether In this mode, the review is not visible to others at the top might mislead some people. I'm struggling to come up with something clearer, but this is my best attempt.

Creates the diff as a draft.
Suppresses notifications to reviewers and subscribers until publication.
The review remains accessible via direct URL.
Use the web UI to modify visibility or policies before publishing.
This revision is now accepted and ready to land.Mon, Apr 27, 5:40 PM

Update man page text following the suggestion from jrm

This revision now requires review to proceed.Mon, Apr 27, 6:42 PM
This revision is now accepted and ready to land.Mon, Apr 27, 6:43 PM
In D56664#1298047, @jrm wrote:

This will be a useful verification step. I sometimes submit a review and then notice that the summary isn't displaying as I intended.

I think the documentation is clear, but I still wonder whether In this mode, the review is not visible to others at the top might mislead some people. I'm struggling to come up with something clearer, but this is my best attempt.

Creates the diff as a draft.
Suppresses notifications to reviewers and subscribers until publication.
The review remains accessible via direct URL.
Use the web UI to modify visibility or policies before publishing.

I expanded your prose a fair bit but tried to follow your structure.

This revision was automatically updated to reflect the committed changes.