On Thu, Dec 10, 2009 at 11:57 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Andrew Gierth <andrew(at)tao11(dot)riddles(dot)org(dot)uk> writes:
>> There's another possible solution (albeit a somewhat nontrivial one)
>> which came up when a bunch of us were talking about this one on IRC;
>> which is to handle the problem in the view deparse: if a column used
>> in a USING clause has been renamed, add an alias to the query that
>> renames it back, e.g.
>> select ... from table1 as table1(v,a) join ... using (v)
> Hmm. Cute, but I wonder why we shouldn't just be throwing an error.
> As I said last night, the only thing I see wrong with the current
> behavior is that the error message isn't phrased in a way that makes
> it obvious it's the user's fault.
I think the problem we're trying to solve is that the view works but
it isn't dumpable.
In response to
pgsql-bugs by date
|Next:||From: Robert Haas||Date: 2009-12-10 17:11:10|
|Subject: Re: BUG #5234: ALTER TABLE ... RENAME COLUMN change view definition incorrectly|
|Previous:||From: Tom Lane||Date: 2009-12-10 16:57:40|
|Subject: Re: BUG #5234: ALTER TABLE ... RENAME COLUMN change view definition incorrectly |