Well, at least two people have yessed my question on whether we need to know the original source time zone on timestamps.
As we won't be having a composite type or whatever for that in 0.1 I installed the following which isn't entirely correct but gets us close: In clin_encounter I added a field source_time_zone of type interval. Upon insert a new encounter into clin_encounter that field is set to the time zone of the client owning the session by means of a trigger. On update the initial time zone is kept. Assuming all data for one encounter is entered from locations in the same time zone this will fairly well reflect the source timezone for timestamps. It isn't fool-proof, however, eg when I enter timestamps on behalf of a lab from a different time zone the source time zone will be wrong. Mind you, the timestamp *value* will still be correct. We just don't have the proper time zone for the lab site. That should suffice for now. Karsten -- GPG key ID E4071346 @ wwwkeys.pgp.net E167 67FD A291 2BEA 73BD 4537 78B9 A9F9 E407 1346 _______________________________________________ Gnumed-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnumed-devel
