I'd be happy with a usable free text search in DB2.
There are multiple DB2 "text extenders", none of which actually work. And,
yes, we've opened PMRs, and even had somebody from IBM Böblingen on-site.
Some obvious bugs (crashes) got fixed, others (including internal errors)
got the "PER" treatment. The Geman word-stemming issues mostly got "WAD"
(some design!), and the crippling restriction to a single column index was
"documented and hence not addressable"; attempts to open a requirement
basically rejected with a hint that IBM was not going to be doing any
further development on free text search.
We finally implemented free text search by using Oracle: we migrate the
relevant data into Oracle on a weekly basis. I suppose this is why the "what
will it cost your business if we don't do it" question didn't really get "a
lot" as an answer - it only really costs "IBM's business".
"Serge Rielau" <sr*****@ca.ibm.com> wrote in message
news:2t*************@uni-berlin.de...
Uhuh... but are you following the news?
http://news.zdnet.com/2100-3513_22-5198086.html
Cheers
Serge