>> I would like to sneak this fix into 9.2, though. Does anyone think >> it's already too late to be touching these APIs for 9.2?
> I'd like us to stick to the standard practice of not changing features/API > in beta releases.
This is a bug fix, not a feature addition, and sometimes you can't fix bugs without touching APIs that might be used by third party code. So the question here is whether this bug fix is sufficiently important, and on the other side how likely it is that anyone has already built extensions for 9.2 that depend on IndexStmt or DefineIndex. I don't think trying to treat it as a "policy" matter is helpful -- it's a tradeoff.
I was hoping that we could fix the bug in released code without having to change the structure or the API, but if that's not feasible, I will withdraw my objection.
If you happen to know of EDB-private code that would be broken by this change, telling us so (and why a mid-beta change would be problematic) would be helpful.
I checked, and I don't see any EDB code that would be affected by this change.
Best regards, --
Gurjeet Singh EnterpriseDB Corporation The Enterprise PostgreSQL Company