"David E. Wheeler" <david@kineticode.com> writes:
> Well now that was cool to see. I got some failures, of course, but
> nothing stands out to me as an obvious bug. I attach the diffs file
> (with the citext.sql failure removed) for your perusal. What would be
> the best way for me to resolve those permission issues?
Don't run the tests in a read-only directory, perhaps.
> Or do they matter for sanity-checking citext?
Hard to tell --- I'd suggest trying to get a clean run. As for what you
have, the first diff hunk suggests you've got the wrong function
properties for citextsend/citextrecv.
regards, tom lane