On Sun, Jan 2, 2011 at 4:45 PM, Peter Eisentraut <pete...@gmx.net> wrote: > On lör, 2011-01-01 at 17:21 -0500, Robert Haas wrote: >> > I don't see anything wrong with having 20 or 30 messages of variants of >> > >> > "foo cannot be used on bar" >> > >> > without placeholders. >> >> Well, that's OK with me. It seems a little grotty, but manageably so. >> Questions: >> >> 1. Should we try to include the name of the object? If so, how? > > Hmm. There is a bit of a difference in my mind between, say, > > constraints cannot be used on sequences > > constraint "foo" cannot be used on sequence "bar" > > the latter leaving open the question whether some other combination > might work.
Yeah, that's no good. Maybe there's a good way to clear things up with an errdetail(), though I'm having a hard time thinking how to phrase it. ERROR: sequence "%s" does not support the requested operation DETAIL: Constraints are not supported on sequences. ERROR: constraints are not supported on sequences DETAIL: "%s" is a sequence. ERROR: "%s" is a sequence DETAIL: Constraints and sequences are like water and oil, dude. >> 2. Can we have a variant with an SQL-command-fragment parameter? >> >> %s cannot be used on views >> where %s might be CLUSTER, DROP COLUMN, etc. > > That's OK; we do that in several other places. Cool. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers