On Sun, Aug 07, 2005 at 07:36:27PM +0200, Hilmar Berger wrote: > > Another error I encountered is some strange behaviour > > when adding a health issue to an new patient (try: create a > > patient, go to the EMR-tree, add a health issue -> exception > > is thrown).
> Looks like this very error was fixed while I was writing this email by > Karsten. I need to test it again, though. I sincerely hope it is fixed. Mathematically I don't feel I can ascertain it's gone. I have, however, dramatically shortened the window in which it might occur -- IF my theory on how it comes about is correct. The fact that apparently you were able to expose it using one client only while I only ever encountered it using two clients working on the same patient somewhat bolsters my theory of unfortunate timing during concurrent operations triggered by backend updates. At the end of the day I have a nagging feeling that the root of the evil is non-locking access to middleware variables ... 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
