Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Mon, Jul 25, 2011 at 12:26 PM, Kevin Grittner
> <Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
>> There's no doubt that it would be better the way you're
>> suggesting; but it looks to me like about five times as many
>> lines of code, harder to be sure it's right, and probably forcing
>> me to learn a few new subsystems of PostgreSQL internals to
>> accomplish.
>
> Sorry, I didn't mean to make homework for you. Nor am I sure that
> the solution will pass must all around even if I think it's the
> best thing since sliced bread. I was just throwing it out there
> as what I would like to have happen in an ideal world...
Well, if it can be done, it will be better and less likely to break
existing code, so it's at least worth looking at. I don't object to
broadening my horizons. ;-) Sorry if it sounded like a complaint;
my intention was to communicate that I'm going to be looking at it,
but I've got a few more urgent tasks to deal with first to get our
application release out the door.
By the way, my current patch does break two existing UPDATE
statements in the regression test misc.sql file:
| -- This non-func update stuff needs to be examined
| -- more closely. - jolly (2/22/96)
| --
| UPDATE tmp
| SET stringu1 = reverse_name(onek.stringu1)
| FROM onek
| WHERE onek.stringu1 = 'JBAAAA' and
| onek.stringu1 = tmp.stringu1;
|
| UPDATE tmp
| SET stringu1 = reverse_name(onek2.stringu1)
| FROM onek2
| WHERE onek2.stringu1 = 'JCAAAA' and
| onek2.stringu1 = tmp.stringu1;
Perhaps it's time....
-Kevin