Re: alter table schema, default sequences stay the same

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Sim Zacks <sim(at)compulab(dot)co(dot)il>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: alter table schema, default sequences stay the same
Date: 2010-06-30 14:22:55
Message-ID: 9547.1277907775@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Sim Zacks <sim(at)compulab(dot)co(dot)il> writes:
>> I haven't consumed enough caffeine today to recall the details, but
>> I think you could have ended up with default expressions like the above
>> if the database had been dumped and reloaded from 8.0 or earlier.
>> nextval(regclass) was introduced in 8.1 precisely to solve this type
>> of problem.

> My first database was an 8.0 and at some point we upgraded to 8.2 so
> that must have caused the ?corruption?

Yeah, that would fit. The upgrade process was intentionally designed to
work that way, because the pre-8.1 behavior of DEFAULT nextval('foo')
was that it always referred to the sequence currently named foo, because
'foo' was just a text constant. We didn't want to risk breaking things
for anyone who was depending on that behavior. But now you have to
explicitly put in ::text if that's how you want it to behave ---
otherwise you get a regclass constant, which will track the original
sequence's identity despite renamings.

regards, tom lane

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Merlin Moncure 2010-06-30 14:46:55 Re: Problem with temporary tables
Previous Message sam mulube 2010-06-30 14:20:35 Re: Postgresql partitioning - single hot table or distributed