Bug 185884

An updated patch with a few more plist fixes.
-Nathan

Comment 2 Nathan Whitehorn 2014-01-25 22:16:28 UTC

Here's another patch (an addition to the last rather than a replacement) 
that fixes sphinx. Doing the grammar pickle generation post-build 
happens before 2to3 is run, with the result that the command will fail 
to execute. Running it in the stage directory solves the problem.
-Nathan

Comment 3 Nathan Whitehorn 2014-01-26 15:05:24 UTC

Responsible Changed
From-To: freebsd-ports-bugs->freebsd-python

Here's an updated version of the patch with respect to r344210.
-Nathan

Comment 5 Nathan Whitehorn 2014-06-08 15:03:35 UTC

Created attachment 143532 [details]
Updated Patch

Newer version of this patch. Also see bug 180467.

Comment 6 Kubilay Kocak 2014-09-05 15:19:55 UTC

Canonicalise python@ assignment.

Comment 7 Alfred Perlstein 2016-02-07 21:43:43 UTC

Koobs, is this going to be committed?

Comment 8 Kubilay Kocak 2016-02-10 14:51:18 UTC

@Alfred I'm not sure this change comes under python@ domain, since it makes changes to ports not maintained by the python team. I should/could have noticed this when canonicalising the python@ assignment, but did not inspect the attachment contents at the time.

On first look however it appears that a few updates to the patches are probably necessary, in terms of:

* that the reported issues are still present in the ports
* if so, ensuring the diffs still apply (and are valid) against ports head
* that any python framework changes that have occurred since submission are reflected. For instance, PYTHON_PY3K_PLIST_HACK has been removed.

To ensure that future issues don't fall through the cracks like this one, it's important to create an issue per port (for port changes), or an issue per isolated/standalone change, so that individual maintainers can be automatically informed and assigned.

I have cc'd the relevant maintainers on this issue now, though 'assignment' is the primary way committers (particularly those in teams@) track their open task lists.

Comment 9 Kubilay Kocak 2016-02-10 14:52:41 UTC

@Nathan, It would also be good if at the very least, each port's changes was in a separate attachment, such that maintainer-approval can be requested for each of them individually.

Comment 10 Raphael Kubo da Costa 2016-02-10 15:07:18 UTC

The devel/py-qt4-core changes were also sent in bug 180467 and landed in ports r358702 ages ago.

Comment 11 Kubilay Kocak 2016-02-10 16:03:05 UTC

Thank you Raphael :)

1 down 2 to go.

I'll help coordinate resolution on this (take)

Comment 12 Brooks Davis 2016-02-10 17:52:44 UTC

Comment on attachment 143532 [details]
Updated Patch

Looks fine for llvm33.  Similar fixes are likely needed for other llvm ports (none of the existing ones will likely ever work with python 3).

Comment 13 Kubilay Kocak 2016-02-13 11:48:39 UTC

Comment on attachment 143532 [details]
Updated Patch

Testing, ignore.

I think this is overcome by events.