Or we can just use our own search system and not use the wiki search
system. I can (and have planned to for some time, just out of
$freetime) revive
searchfedora.org using the existing datapark search
(
http://dataparksearch.org/) code or we can use an experimental indexing
system I'm working on in python that uses Xapian and some custom python
to feed the indexer, as Xapian doesn't have it's own
robots (
http://www.xapian.org/). Datapark search scales very well. I've
not been the most impressed with xapian, but moreso with it's indexing
strategy rather then it's true purpose as an indexing framework (as
best I can tell.) IIRC, MoinMoin has xapian plugins, but I would
recommend a completely external search system. Let me know if reviving
searchfedora.org should be moved up on my todo list.
Why not just let Google do it? I know I've run into sites that do this.
I don't know the details though. Presumably you put a