On Sun, May 29, 2005 at 01:58:50AM +0000, Horst Herb wrote: > 2.) Because reference database can be volatile, the prescription module will > *not* just store a reference to a drug in the drugef database, but for each > prescription it will record in full text For each *prescription* or for each *drug* ever used in a prescription ? The former would mean that you'd be storing data on a given drug as many times as you prescribe the drug (a shame) while the latter means you would have a table caching the information from the reference database inside the clinical database - but only for those drugs that were ever used in a prescription and then only once for each. The actual *presription* table would then simply point to that "cache" table.
This is in fact what Hilmar suggested we do in GNUmed and I do think it is a sensible approach. > I used to print via generating a KWord (or formerly Abiword) file, but now I > am really happy with the way printing works in wxPython 2.6.0, so this > weekend I port my printing modules to use wxPython routines Is LaTeX too slow ? > The user interface problem for different national requirements is simple to > solve: > 1.) I "draw" the user interface using wxGlade, one for each special > requirement, but sticking to user interface element naming conventions > 2.) wxGlade produces a Python module for each such user interface design > 3.) the prescribing dialog module inherits one of those generated modules > programmatically depending on locale > 4.) since most functionality (including GUI) will be the same for all > nations, > these "universal" routines are isolated into a separate module I have to see this before believing it. I seriously doubt it is "simple to solve". 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
