> What I'd like to see it do is grab the dependency data in pg_depend and
> do a topological sort using that.
At the end though, we'd need to dump stuff not caught be the topsort, for
cases where pg_depend has been messed with.
> This leaves some issues still to be
> resolved ... like what to do when dumping a pre-7.3 database ... but I
> think it's the core of a maintainable solution.
Problem is you'd need to sort tables by the youngest column in the table,
which is a pain. Because the main problem is this:
CREATE TABLE...
CREATE TYPE newtype
ALTER TABLE ADD COLUMN newtype
That always breaks...
CHris