Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes:
>> AFAIK the GiST modifications suggested Aoki have not yet been
>> implemented, so this should not have been removed.
> I thought that was an extremely vague limitation and could easily be
> removed... It's like having a limitations section on btrees and have
> mentioning in passing that it can't do full text indexing or something.
Yeah, I agree. It'd be appropriate to mention the Aoki paper in TODO,
but we don't normally document "stuff we'd like to do someday" as
serious limitations.
regards, tom lane