Re: Sequence Access Method WIP

On Sat, Jan 30, 2016 at 7:37 AM, Petr Jelinek <petr(at)2ndquadrant(dot)com> wrote:
> Just as a note, CREATE SEQUENCE ACCESS METHOD already causes grammar
> conflict now, that's why my proposal was different, I didn't want to
> add more keywords. I think Alvaro's proposal is fine as well.

I missed your proposal, I guess, so please don't take as having any
position on whether it's better or worse than Alvaro's. I was only
intending to vote for the proposition that the type of access method
should follow the name of the access method.

> The other point is that we are creating ACCESS METHOD object so that's
> what should be after CREATE.

Agreed.

> In any case this is slightly premature IMHO as DDL is somewhat unless
> until we have sequence access methods implementation we can agree on,
> or the generic WAL logging so that custom indexes can be crash safe.

Generic WAL logging seems like a *great* idea to me. But I think it's
largely independent from the access method stuff. If we have generic
WAL logging, people can create WAL-logged stuff that is not a new
access method. If we have access methods, they can create new access
methods that are not WAL-logged. If we have both, then they can
create WAL-logged access methods which of course is the payoff pitch,
but I don't see it as necessary or desirable for the two systems to be
tied together in any way.

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