On Thu, 2010-01-07 at 18:02 -0500, Robert Treat wrote:
> On Jan 7, 2010, at 5:18 PM, Simon Riggs wrote:
>
> > On Thu, 2010-01-07 at 16:56 -0500, Robert Treat wrote:
> >
> >> Not much more to send really.
> >
> > No problem. I can see what causes it, nothing more required, thanks.
> > What I don't fully understand yet is why the error hasn't shown itself
> > before, because it appears such a basic error.
> >
>
> Is anyone doing regular testing on non-linux? (to the extent that
> this is a problem due to Solaris ideas about snprintf).
OK, understand now. The Solaris test failed in a place where it would
have been good if Linux had done so; this has revealed that the relcache
init file deletion required in some places was actually failing. The
path was never set correctly in the first place, it appears. I've
updated the TODO to reflect the new bug report.
I've almost finished hacking a solution. I'll test and commit tomorrow,
since it's a fairly standalone problem, if fiddly.
-- Simon Riggs www.2ndQuadrant.com
