Dimitri Fontaine <dimitri@2ndQuadrant.fr> writes:
> I think we shouldn't change the content of pg_depend lightly here, and
So here's a patch following that idea.
Even for TIP I don't want us to change how pg_depend tracking is done,
because I want to propose a fix for the pg_dump bug wrt sequences and
pg_extension_config_dump() wherein you can actually register a sequence
(owned by a table or not) but then pg_dump fails to dump it (see report
from Marko Kreen)
http://archives.postgresql.org/message-id/CACMqXCJjauc9jPa64VxskRN67bYjuYmodz-Mgy-_aoZ6ErG3Yg@mail.gmail.com
Regards,
--
Dimitri Fontaine
http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support