Il Thursday 25 October 2007 16:29:33 Scott Marlowe ha scritto:
> On 10/24/07, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> > Joe Conway <mail@joeconway.com> writes:
> > > Tom Lane wrote:
> > >> 1. Treat NULL rowid as a category in its own right. This would
> > >> conform with the behavior of GROUP BY and DISTINCT, for instance.
> > >
> > > In any case, the attached changes the behavior to #1 for both flavors
> > > of crosstab (the original crosstab(text, int) and the usually more
> > > useful crosstab(text, text)).
> > >
> > > It is appropriate for 8.3 but not back-patching as it changes behavior
> > > in a non-backward compatible way and is probably too invasive anyway.
> >
> > Um, if the previous code crashed in this case, why would you worry about
> > being backward-compatible with it? You're effectively changing the
> > behavior anyway, so you might as well make it do what you've decided is
> > the right thing.
>
> As a crosstab user, I agree with Tom.
If I can throw in my EUR 0.01 contrib, I would agree with Joe (thanks for your
wonderful crosstab).
If crosstab in 8.3 will have a different behaviour *and* it's not part of the
core features, then I'd prefer to correct it.
In any case developers will have to cope with discrepancies when going to 8.3
and you can bet they won't remain with 8.2 when 8.3 will be rolled out.
And, by the way, why not including the crosstab as a standard feature?
I think it deserves it!