I noticed that I am having a difficult time identifying if and when a token is removed or added through the muscle interface.. I am sure
> that many others are having the same issue.. yes you can use the > callback with MSCCallbackForTokenEvent() but this API can cause all > sorts of threading issues. especially if you try and cancel them > while they are waiting.
Is there really a bug with the callback API, or is the "bug" that people have problems doing multi-threaded programming properly?
What's the issue with cancelling callbacks?
It occurred to me that a simple way to identify tokens coming and> going would simply be to add a field to the MSCTokenInfo structure
Callbacks are much more system and resource friendly, especially when software is used on a multi-user system (anyone here remember those :-). Polling is OK when you're on your own dedicated 3GHz XEON Linux box, but it really sucks the cycles when you're sharing a server running multiple users.
Multi-threaded programming is more difficult if you're not used to it, but once you get the hang of it, it's really cool. And, you can almost always convert a callback/wait mechanism to look like a polling loop without incuring the overhead of polling.
I'm wary of "fixing" an API because a common method of programming is "hard". I'd much rather see Dave teach an on-line course in multu-threaded programming :-).
mike
-- ---------------------------------------------------------------------------- Michael Bender E-Mail: [EMAIL PROTECTED] Sun Microsystems, Inc. Tel: 831-401-9510 14 Network Circle Tel: x.31807 Menlo Park, Ca. 94025 Mailstop: UMPK14-260 MD: VPN/IMAP
Never make assumptions... they are usually wrong.
----------------------------------------------------------------------------
_______________________________________________ Muscle mailing list [EMAIL PROTECTED] http://lists.drizzle.com/mailman/listinfo/muscle
