Re: pgsql: Fix pg_type.typlen for newly-revived line type. - Mailing list pgsql-committers

From Bruce Momjian
Subject Re: pgsql: Fix pg_type.typlen for newly-revived line type.
Date
Msg-id 20140514202710.GE31386@momjian.us
Whole thread Raw
In response to Re: pgsql: Fix pg_type.typlen for newly-revived line type.  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-committers
On Mon, May  5, 2014 at 03:00:47PM -0400, Tom Lane wrote:
> Alvaro Herrera <alvherre@2ndquadrant.com> writes:
> > Tom Lane wrote:
> >> Note that if the line type had been in use in the field, this would
> >> break pg_upgrade'ability of databases containing line values; but
> >> it seems unlikely that there are any (they'd have had to be compiled
> >> with -DENABLE_LINE_TYPE).
>
> > Should we add a test to pg_upgrade --check mode?
>
> I doubt it's worth it.  Anybody who was enabling the old code is pretty
> well screwed no matter what.  They can't upgrade with a plain pg_dump
> either, since the I/O format is different.
>
> What we do need, probably, is a warning about the incompatibility in the
> release notes.

I added a check to pg_upgrade to error out on previous 'line' data type
usage because pg_upgrade would have blindly upgraded the
non-binary-compatible data.  I also added a mention of this in the
9.4 release notes.

--
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + Everyone has their own god. +

Attachment

pgsql-committers by date:

Previous
From: Bruce Momjian
Date:
Subject: pgsql: pg_upgrade: error out on 'line' data type usage
Next
From: Michael Paquier
Date:
Subject: Re: pgsql: Replace SYSTEMQUOTEs with Windows-specific wrapper functions.