Andrew Dunstan <andrew@dunslane.net> writes:
> Tom Lane wrote:
>> That's not really the most preferable solution, I think, seeing that it
>> still leaves the user with the problem of having to create the types in
>> the right order to start with.
> I'm not sure we can keep the _foo convention and avoid that.
Auto-rename. I'm working on a patch now, and it doesn't look like it'll
be too awful. Will post it for comments when it's working.
> ... I'd vote to revert the new name
> mangling piece (but keep the typarray mapping column), deprecate the use
> of the _foo convention, and abandon it next release.
I came across a comment in the source that says PG has been using _foo
for arrays since 3.1 (!). I don't think we can get away with changing
it, certainly not with only one release cycle's notice.
The current code is OK from a compatibility point of view, since it only
changes _foo to something else in situations where the old way would've
failed outright. I think we need to preserve that property ...
regards, tom lane
