Re: Oid registry

On Thu, Sep 27, 2012 at 04:02:45PM -0400, Robert Haas wrote:
> On Thu, Sep 27, 2012 at 2:34 PM, Magnus Hagander <magnus@hagander.net> wrote:
> > I'm not sure that's a way we really want to go down. How do we define which
> > third party vendors would get to reserve oids? And how many? And under what
> > other potential terms?
> >
> > Seems like we'd set ourselves up for endless discussions and bike
> > shedding...
> 
> Not really.  I'm only proposing that it would be nice to have a block
> of OIDs that core agrees not to assign for any other purpose, not that
> we dole out specific ones to specific companies.  There's no reason
> why, for example, EnterpriseDB's fork can't use OIDs from the same
> reserved block as PostgreSQL-XC's fork or Greenplum's fork or Aster
> Data's fork - those are all distinct projects.  All might need private
> OIDs but they can all come from the same range because the code bases
> don't mingle.
> 
> That having been said, we've gotten this far without having any
> terrible trouble about this, so maybe it's not worth worrying about.
> It's a nice-to-have, not a big deal.

Interesting idea, but if plugable data types started using that reserved
range, it could conflict with XC or EDB-reserved oids, making those data
types unusable in those forks.  Maybe we need two reserved ranges.

--  Bruce Momjian  <bruce@momjian.us>        http://momjian.us EnterpriseDB
http://enterprisedb.com
 + It's impossible for everything to be true. +