On 07/05/10 16:46, Raphael Coeffic wrote:
> Frankly, I don't like this approach too much, as there is a million
> cases it could go wrong. I think we should have all the state changes
> tailorized into one function (setState), which would then call some
> onStateChanged(old_state,new_state). This is much closer to our fsm
> strategy.

Yeah, smth. like setState() also crossed my mind.
Some onStateChgd() would also be OK, though there are lots of callbacks
[already :-) ].

> 
> Concerning the specific problem with authentication, Stefan and I were
> in favor of moving the code into trunk. Any other suggestions?

Sure, but would be cool to keep current capability allowing each app to
instantiate it's own "authenticator"; like now, having a registration
agent, but also some caller session, each using different credentials.


Regards,
Bogdan.
_______________________________________________
Semsdev mailing list
[email protected]
http://lists.iptel.org/mailman/listinfo/semsdev

Reply via email to