PSI hashing and relational databases

Alexander Johannesen <[email protected]> Thu, 26 Jan 2012 12:02:03 +1100
Newsgroups gmane.text.xml.xtm.general
Message-ID <CAF89bCB5Jg4SFvzCEJBwrUUGjOJH9oacnmAbPAedhBUZvEeQmA@mail.gmail.com>
Hi all,

I'm in the middle of a big development, and I'm using a relational database
to hold a TMDM/TMRM mashup model, and I'm currently dealing with PSIs and
how to best deal with them.

My design took an early decision to store internal ID's as Int, and have a
separate NAME that was a CHAR(99), both indexed, of course, and then let
the framework deal with the overhead of converting between the two.

Next step is to work out best how to store and deal with PSI, gacing a
similar challenge and I'm wondering what people here have done. I can store
PSIs in a larger char, and create a shorter hashed field which the
framework will work with.

Thoughts, even on the sort of hashing I should do (if at all)?

Alex

_______________________________________________
topicmapmail mailing list
topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected]
http://www.infoloom.com/mailman/listinfo/topicmapmail