I would +1 everything you said... if Julian had changed svn_boolean_t.
He changed svn_tristate_t, which is used exactly once in trunk, so
I don't see the need to be as careful with it.
The caveat? This assumes that --- as we ought to --- we not act on
feedback differently because it comes after the change has been
committed. Feedback that would have caused a "No, let's not commit this
yet" reaction if provided early should cause a "OK, let's revert this
for now then" reaction if provided later.
Greg Stein wrote on Tue, Nov 09, 2010 at 00:47:27 -0500:
> On Thu, Nov 4, 2010 at 08:45, Hyrum K. Wright
> <hyrum_wright_at_mail.utexas.edu> wrote:
> > On Thu, Nov 4, 2010 at 7:43 AM, Julian Foad <julian.foad_at_wandisco.com> wrote:
> >> Having said all that, +1 on removing the gratuitous inconsistency by
> >> applying this patch.
> >> Committed r1030909.
> > Gah. Can we please wait a little bit longer on this kind of stuff to
> > allow people in other timezones a chance to weigh in?
> I have raised this before, Julian. For making changes with some
> impact, where feedback from the community is desired, then the
> standard Apache rule is 72 hours. And even if we don't worry about
> rules, it is simply *respectful* to give others a chance to speak up.
> Last time, you bumped the format with something like FOUR hours
> notice. That was really bad. I'm not sure about the extent of the
> badness here (tho changing something as central as svn_types.h seems
> pretty obvious as an "input-required" change), but given that another
> member of the community has said "woah. too quick", then you REALLY
> need to slow down.
> This isn't an attempt to slow down your work. This is an attempt
> (bordering on requirement) for you to work with the rest of the
> community on changes to our codebase.
Received on 2010-11-09 18:19:00 CET