openhatch / oh-bugimporters

Bug importers for the OpenHatch project oh-mainline
https://oh-bugimporters.readthedocs.org/
GNU Affero General Public License v3.0
12 stars 28 forks source link

Exclude listing issues with a patch attached #69

Open ghost opened 10 years ago

ghost commented 10 years ago

Comment by ehashman:

mandarg wrote in issue906: """An unrelated problem is also that the python bugtracker is possibly not always updated, for example, shouldn't http://bugs.python.org/issue1100942 be in Stage "patch-review"? We could also get around this by excluding all issues with a patch attached."""

The listed bug's status is now "needs patch", which in issue 906 would still be considered open, but given the work already done on this bug I'd say excluding it and things like it would still make sense.

However, it may be problematic to exclude issues that have patches attached---if a bug is in status "needs testing", which was listed in the last post, it will probably have a patch attached, and we don't want to filter those out.


Comment by mandarg:

Fair enough. I see two paths here, looking for thoughts on both:

  1. Filter out bugs that have a patch attached and are do not have a status 'needs testing'.

Pro: This gives us bugs for which a patch has already been written, and openhatch contributors can test that patch, and possibly update the bug with the fact that it is working.

Con: There could be an implicit assumption that the person testing the bug patch also is code-reviewing it, for instance the bug that ehashman quoted (http://bugs.python.org/issue1100942) seems to combine reviews and testing. This may create more noise for new contributors.

  1. Totally filter out any bugs with a patch attached, which would probably exclude all bugs that have the status 'needs testing' (since presumably a patch is always attached that would be used for said testing).

Pro: This would restrict the importer to get only those bugs which are open and have no patch and that no one is currently working on. This means that new contributors would always see only "fresh" bugs.

Con: We might be ignoring some low-hanging fruit of bugs in stage 'needs review' with small patches attached, which could easily be tested by new contributors.

I'd suggest going with the first option initially, if possible (to be more conservative about the number of bugs we cut out), and then revisit the issue if we still think the imported bugs look too noisy.

On Fri, Apr 4, 2014 at 12:13 PM, ehashman <bugs@openhatch.org> wrote: > > > New submission from ehashman <elana@hashman.ca>: > > mandarg wrote in issue906: """An unrelated problem is also that the python > bugtracker is possibly not always updated, for example, shouldn't > http://bugs.python.org/issue1100942 be in Stage "patch-review"? We could also > get around this by excluding all issues with a patch attached.""" > > The listed bug's status is now "needs patch", which in issue 906 would still be > considered open, but given the work already done on this bug I'd say excluding > it and things like it would still make sense. > > However, it may be problematic to exclude issues that have patches attached---if > a bug is in status "needs testing", which was listed in the last post, it will > probably have a patch attached, and we don't want to filter those out. > > ---------- > messages: 4183 > nosy: ehashman, mandarg, paulproteus > priority: feature > status: chatting > superseder: Issues on python.org that are actively being worked on or closed remain in our list of 'open issues'' > title: Exclude listing issues with a patch attached > > __ > Roundup issue tracker <bugs@openhatch.org> > <https://openhatch.org/bugs/issue953> > __


Status: chatting Nosy List: ehashman, mandarg, paulproteus Superceder : #Issues on python.org that are actively being worked on or closed remain in our list of 'open issues'' Priority: feature Imported from roundup ID: 953 (view archived page) Last modified: 2014-04-04.17:32:49