Hmm, not much a response to the last email. No matter, I've made some more
progress.
* The planner now correctly determines that the index is usable without
segfaulting.
Really wasn't that hard. The code that deals with checking the predicate
when inserting still worked fine.
So, as far as I can tell, partial indecies would be completely usable, *if*
I could get the planner to use them. I'm pretty sure it goes wrong in
create_index_paths. The pred_test works fine but somewhere in the lines
below it doesn't realise it can use the index. Since I can't quite work out
what those functions are supposed to do I'm don't see how to fix it either.
Could somebody look into it?
I have a sneaking suspicion the cost estimator will need to be fixed too,
but I havn't got to that yet.
The patch is against 7.1release.
--
Martijn van Oosterhout <kleptog@svana.org>
http://svana.org/kleptog/
> It would be nice if someone came up with a certification system that
> actually separated those who can barely regurgitate what they crammed over
> the last few weeks from those who command secret ninja networking powers.