Accepting foreignkey assignment by PK on QuerySet.update() (#11521)

5 views
Skip to first unread message

Jerome Leclanche

unread,
Sep 11, 2009, 9:29:22 AM9/11/09
to django-d...@googlegroups.com
Could someone give some feedback on #11521?
(http://code.djangoproject.com/ticket/11521)
It's been sitting for almost 3 months without any kind of answer. The
current patch is a 2-line change which looks very reasonable, but I'd
love to be aware of any possible side effect of that change.

J. Leclanche / Adys

Karen Tracey

unread,
Sep 11, 2009, 10:03:31 AM9/11/09
to django-d...@googlegroups.com

Feedback from someone who isn't an ORM expert:

Why doesn't it have a test?

Mentioning in the description that the fix may be a hack might not have been the best idea. 

Karen

Jerome Leclanche

unread,
Sep 11, 2009, 10:06:38 AM9/11/09
to django-d...@googlegroups.com
On Fri, Sep 11, 2009 at 5:03 PM, Karen Tracey <kmtr...@gmail.com> wrote:
>
> Why doesn't it have a test?

I'll write one then.

> Mentioning in the description that the fix may be a hack might not have been
> the best idea.

It's not "may", it's "I don't know". I'm not an ORM expert either,
hence the feedback request.

Cheers

> Karen
>
> >
>

Karen Tracey

unread,
Sep 11, 2009, 10:51:20 AM9/11/09
to django-d...@googlegroups.com

Fixes presented with the statement "It may or may not be a hack" will likely take longer to get accepted.  It's a clear signal to any committer not intimately familiar with the area of code in question that tackling that particular ticket will require doing some significant amount of investigation to verify that it is not in fact a hack fix or just symptom patching.  Thus a statement like that likely limits the number of commiters who are going to look seriously at the fix to those who are experts in the area.  (Or those who have some extra time on their hands and are interested in becoming an expert in the area.)

I'm not suggesting presenting fixes as "surely right" when you are not in fact sure of that is a good idea.  But if you are looking to move things along, doing some independent research into the code and perhaps verifying the approach taken in the fix is also used elsewhere, verifying that the full test suite runs correctly with the fix, etc. are things you can do and state in the ticket that might help with getting the fix accepted.

I realize now you're looking more for feedback on the technical issues that may be associated with the fix but that wasn't clear to me from the first note, where it sounded like you could be looking for general feedback on why the ticket hadn't moved along.  Sorry I can't provide feedback on the former but I did see some reasons for the latter so that's what I responded with.

Karen
Reply all
Reply to author
Forward
0 new messages