> > Hmm, did you apply the latest patch correctly? My build can produce
> > right results, so I don't see why it isn't fixed. Make sure the lines
> > around 2420-2430 in planner.c like:
> > /*
> > * must copyObject() to avoid args concatenating
> with each other.
> > */
> > pulled_exprs = list_concat(pulled_exprs,
> > where copyObject() is added.
> I'm sitting here away from home with a funny feeling I forgot to make
> I think my home adsl has dropped out so can't confirm that. If it
> works for you and not for me last night then I probably did forget.
> I'll let you know.
Sorry, false alarm. What can I say, it was late. I did make install, just to
the wrong place, so ended up using the old binary.
Sorry for panic. The bug is fixed. Only I still get regression failures on
window. Do they all pass for you?
I have most of the weekend to test Heikki's work, in fact it's building now.
But I'll have to leave it running...
In response to
pgsql-hackers by date
|Next:||From: Simon Riggs||Date: 2008-11-27 19:29:15|
|Subject: Re: New trigger file in pg_standby to promote thestandby to the primary|
|Previous:||From: Simon Riggs||Date: 2008-11-27 17:14:37|
|Subject: Re: Review: Hot standby|