I wrote:
> I think it's okay for CREATE INDEX CONCURRENTLY to use bulk-read access
> strategy (that is, seqscan using a limited number of buffers), but it
> has to be able to force the scan to start at page zero.
I've committed a patch to do that. Please test CVS HEAD and see if you
still see problems.
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to [EMAIL PROTECTED] so that your
message can get through to the mailing list cleanly
