Re: v7.0.3 Regress Tests Errors

Paul Huppe <huppe(at)NRCan(dot)gc(dot)ca> writes:
> It was acutally Peter Eisentraut who asked me to post to the list. He
> wanted to see the .diff file because he tought that I had a problem with
> broken Unix domain sockets.

Hm, I should think that broken sockets would lead to total non function,
rather than selected (even if extensive) errors.

But, if you're suspicious of that, try running the tests with
environment variable PGHOST set to localhost, so that TCP connections
will be used (and make sure you started the postmaster with -i of
course).

> Platform: Sun SPARCserver 10
> OS: Solaris 8
> Compiler: gcc 2.95.2
> Configuration Options: none, used defaults

That doesn't sound particularly risky. Odd ...

regards, tom lane