Remove pg_dump -i option (was Re: Proposed patch: synchronized_scanning GUC variable)

Alvaro Herrera <alvherre@commandprompt.com>

From: Alvaro Herrera <alvherre@commandprompt.com>
To: Simon Riggs <simon@2ndquadrant.com>
Cc: Bruce Momjian <bruce@momjian.us>, Tom Lane <tgl@sss.pgh.pa.us>, Neil Conway <neilc@samurai.com>, Gregory Stark <stark@enterprisedb.com>, "Jonah H. Harris" <jonah.harris@gmail.com>, pgsql-hackers@postgresql.org
Date: 2008-01-31T14:47:26Z
Lists: pgsql-hackers
Simon Riggs escribió:
> On Thu, 2008-01-31 at 11:20 -0300, Alvaro Herrera wrote:
> > > Tom Lane wrote:
> > 
> > > > in fact, personally I'd like to make that case be a hard error,
> > > > rather than something people could override with -i.
> > 
> > +1 to this idea.  TODO for 8.4?
> 
> -1 without some more planning about the effects and implications.

Effect: we would stop receiving complaints that an old pg_dump can talk
to a server that most likely is incompatible with it.  People would
learn on the spot that they must install the newer pg_dump.

Implication: you cannot dump a newer database and expect it to load on
an older server (it would work for certain versions, but not all).  So
if people have a bleeding-edge test server, they cannot migrate stuff
from it to their older production server.

-- 
Alvaro Herrera                                http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.