kfogel@collab.net writes:
> > > I don't understand the significance of this one:
> > > - libsvn_wc/status.c:close_edit
> > > fails to clear traversal_info
> >
> > I haven't looked because I'm pretty sure I won't understand either.
>
> I know the traversal_info stuff somewhat, I'll take a look at this.
Yup, it looks like this was a problem, though a minor one.
See issue #1741 and r8842+r8962 for why the traversal_info needs to be
cleared at all. Unfortunately, clearing it at the beginning of the
function doesn't work, and although it's likely that by returning an
error we'd be making traversal_info irrelevant anyway (because instead
of status being reported, an error would be reported), that's up to
the callers and isn't an assumption close_edit() can make.
It seems like the Most Right Behavior is for close_edit() to ensure
that traversal_info is correct by return time, whether or not we're
returning an error. I've done that in r14709, even though it means
introducing one of those brittle manual-error-checking situations.
-Karl
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu May 12 20:42:25 2005