Re: Documenting removal of nonnullvalue() and friends

On Wed, Oct 13, 2010 at 9:56 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Tue, Oct 12, 2010 at 9:05 PM, Josh Kupershmidt <schmiddy(at)gmail(dot)com> wrote:
>> On Tue, Oct 12, 2010 at 1:03 PM, Josh Kupershmidt <schmiddy(at)gmail(dot)com> wrote:
>>> Perhaps a mention of this change could be made in the 8.4 release notes?
>>
>> And here's a patch for the 8.4 release notes.
>
> It seems odd to change our release notes after the fact.  I don't
> think we usually do that.

Well, I do count 10 or so updates to release-8.4.sgml since its
release date (2009-07-01). Granted, most of these are minor cleanups
or to document further point releases.

> I'm also not really sure that this is important enough to be worth
> documenting.  It looks fairly self-explanatory.

Oh well.

Would anyone favor instead back-patching the documentation for the
8.3, 8.2, and 8.1 branches to include mentions of these
previously-undocumented functions, instead? In
<http://archives.postgresql.org/pgsql-docs/2004-08/msg00015.php>, Tom
opined that they should be left undocumented, but I really don't agree
with that.

I almost concluded the original -novice poster was simply missing a
user-defined function from her old database, since googling for
"postgresql nonnullvalue" brought up nothing useful, and there was no
mention of it in the (current) source code. It seems pretty harsh to
me to intentionally not document these functions, even if they are
obscure and not terribly useful.

Josh