Lamar Owen <lamar.owen@wgcr.org> writes:
> On Wednesday 18 September 2002 12:55 am, Tom Lane wrote:
>> But the system catalogs *store* that metadata.
> They _currently_ store the user's metadata. But that's my point -- does the
> user metadata that isn't typically substantially different after going
> through a dump/reload _have_ to coexist with the system data which is
> intrinsic to the basic backend operation?
I think we're talking at cross-purposes. When I said we can't freeze
the system catalogs yet, I meant that we cannot freeze the format/schema
in which metadata is stored. That affects both system and user entries.
You seem to be envisioning moving user metadata into a separate set of
tables from the predefined entries --- but that will help not one whit
as far as easing upgrades goes.
> Given the desireability for opaque to go away soon, if
> the 7.3 pg_dump Does The Right Thing and creates an opaque-free dump,
The present proposal for that has the 7.3 backend patching things up
during reload; it won't matter whether you use 7.2 or 7.3 pg_dump to
dump from a 7.2 database.
> And it may prove to not be
> that bad -- I'll know in a few days, hopefully.
If you find that it's not too painful then I do agree with doing it.
There will doubtless be future cycles where it's more valuable to be
able to use the up-to-date pg_dump than it is in this one.
regards, tom lane