Re: ALTER TYPE 3: add facility to identify further no-work cases

Robert Haas <robertmhaas@gmail.com>

From: Robert Haas <robertmhaas@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Noah Misch <noah@leadboat.com>, jim@nasby.net, pgsql-hackers@postgresql.org
Date: 2011-01-27T01:12:53Z
Lists: pgsql-hackers
On Wed, Jan 26, 2011 at 7:44 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> But how often do those really come up?  And do you really save that
> much?  The table still has to be locked against other users, so you're
> still down, and you're still doing all the reads and computation.  I
> don't deny that saving the writes is worth something; I just don't agree
> that it's worth the development and maintenance effort that such a wart
> is going to cost us.  User-exposed features are *expensive*.

I would think that text -> [something that's still a varlena but with
tighter validation] would be quite common.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company