Whoops, forgot the list on first iteration, try again. On tirsdag 03 mai 2005, 15:19, you wrote: > So, you mention in your bug report that this is not a problem for > Postgres...
Eh, no, to the contrary, it is more severe for Pg than for other backends. I wrote "can't overcome". The issue is really transactional integrity. Pg has the strongest code for transactional integrity, equal to that of Oracle. In the case of problems, Pg will wait around and eventually time out (after two minutes), but other backends will go ahead. It will go OK for a while, but may eventually run into problems. > Is it likely to be a problem If I try to use mysql to store the > session files? I don't know, but I think the percieved problems will be less, since MySQL isn't as rigorous as Pg. It may be worth a shot. However, since there is no solution to the real problem, and so, the consequences may be bad. > Is this even possible? > Can you give som pointers as to how I would go about doing this > please? No, sorry... > If it's not possible, what other methods are there that I could use > which would not encounter this problem (excluding Postgres)? Never invalidate the session is the best I can give... It occurs only on invalidation, and you could in principle design the app so that a user comes in, gets a session, and that session is used, assigned values and so on, but never actually invalidated. Actually, I think this is how it is supposed to be done... Best, Kjetil -- Kjetil Kjernsmo Programmer/Astrophysicist/Skeptic/Ski-orienteer/Orienteer/Mountaineer [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] Homepage: http://www.kjetil.kjernsmo.net/ OpenPGP KeyID: 6A6A0BBC --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
