Mario Weilguni <mweilguni@sime.com> writes:
> Am Samstag, 15. Februar 2003 16:49 schrieb Tom Lane:
>> It might be sufficient to insist on standalone reindexing just for the
>> three or four tables that have nailed-in-cache indexes. I'm not sure
>> though. Hiroshi, what do you think?
> And what about adding a column "indissystem" to pg_index and setting this to true for system indexes, false
otherwise?
> Then there would be no need to hardwire the reindex logic.
If the nailed-in-cache property is what matters, then we need to look at
that, not at some boolean column that might or might not agree with it.
I'm not entirely sure what does matter here, though.
regards, tom lane