Re: ANSI Compliant Inserts

Hiroshi Inoue wrote:
> > I meant to say there was no question we wanted this item fixed, not that
> > there was no need for implementation discussions.
> >
> > In summary, code changes have three stages:
> >
> > o Do we want this feature?
> > o How do we want the feature to behave?
> > o How do we want the feature implemented?
> >
> > Tom was complaining because the patch appeared without enough discussion
> > on these items. However, from my perspective, this is really trying to
> > micromanage the process. When people post patches, we aren't forced to
> > apply them.
>
> But shouldn't someone check the patch ?
> If the patch is small, making the patch seems
> the simplest way for anyone but if the patch
> is big, it seems painful for anyone to check
> the patch. If no one checks the patch, would
> we apply the patch blindly or reject it ?

Of course, we would review any patch before application. I guess the
full path is:

o Do we want this feature?
o How do we want the feature to behave?
o How do we want the feature implemented?
o Submit patch
o Review patch
o Apply patch

I assume your point is that people shouldn't send in big patches
without going through the discussion first. Yes, that is ideal, but if
they don't, we just discuss it after the patch appears, and then decide
if we want to apply it or ask for modifications.

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026