"Jonathan S. Katz" <jkatz@postgresql.org> writes:
> Just to chime in real quick: from the perspective of the RMT we did not
> look at these as a series of "right/wrong" options but what would make
> the most sense for the v11 release so the community could continue to
> improve support for stored procedures and make it easier for users to
> work with them going forward.
Yeah. I think the real options we had here were:
1. Ship v11 with CALL as it stands, try to improve going forward.
2. Delay v11 for months while we figure out something better.
3. Remove CALL from v11, try again in v12.
Neither #2 nor #3 are at all attractive.
People will need to realize that CALL behavior is still a work in
progress. That may mean that JDBC etc shouldn't try to support it
yet, which isn't great, but none of the above options would have
led to near-term support of CALL in JDBC ...
regards, tom lane