Informix/DB2 extended indexes [was: some long thread]

Obnoxio The Clown <obnoxio@hotmail.com> wrote in message news:<bl3fbf$7giod$1@ID-64669.news.uni-berlin.de>... > Database Guy wrote: > > > Obnoxio The Clown <obnoxio@hotmail.com> wrote in message > > news:<bl1sb2$5r2kn$1@ID-64669.news.uni-berlin.de>... > > > >> > Could we call them "datablades"? > >> > >> Having used a couple, I beg to differ. > > > > Actually, I did once hear that datablades were one of the technologies > > IBM most wanted from Informix. Not quite sure why since DB2 has long > > had extenders - but it was suggested that the Informix implementation > > was in some way architected differently (better)? > > > > Might have been a baseless rumour, might have been factual. Anyone > > know more about what the difference in approach to extenders is, > > between Informix and DB2? > > IMHO, the primary difference is that IDS allows you to create your own > indexing technology, DB2 doesn't. DB2 allows you to create index extensions on user-defined types - quite powerful - but I don't think that's what you meant? Presumably Informix allows you to create the entire index separate from the DBMS yet inform the optimizer that the relevant functions implement a useable index? IMHO this is not a big enough difference. I got the impression this was something more fundamental. However, it could just have been a total red herring. DG