Bug 247795
| Summary: | [git-webkit] git webkit pr should not create a bug for unreviewed commits with no bug associated | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Carlos Garcia Campos <cgarcia> |
| Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | Normal | CC: | jbedard, webkit-bug-importer |
| Priority: | P2 | Keywords: | InRadar |
| Version: | WebKit Nightly Build | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
Carlos Garcia Campos
It currently creates a bug, but it's not filled with pr info nor closed when pr is merged, I think we should just avoid creating the bug at all.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Radar WebKit Bug Importer
<rdar://problem/102507277>
Jonathan Bedard
We create the bug at branch time, the prompt is:
`Enter issue URL or title of new issue: ` if we have an issue tracker
`Enter name of new branch (or issue URL): ` if we do not.
You can short-circuit the issue creation by providing a branch name instead of issue description (so something like `eng/foobar`) or create the branch before invoking `git-webkit pr`.
The trouble here is that I don't think we want to encourage the practice of commits detached from a bug. It's ok for multiple commits to be attached to a single bug (often, this is even preferable, when talking about something like test gardening or a follow-up fix to an existing commit), but I struggle to think of situation where a commit totally unattached to a bug is acceptable...the only one that comes to mind adding a contributor to contributors.json.
I think the honest answer to this bug is "won't fix", although I can see us making some modification to our prompt to inform the user of the existence of the short-circuit option.