-=> Quoting James R. Cook to Phil Crown <=-

 -=> Quoting Phil Crown to James R. Cook <=-

 JRC> But does this make a huge deal?  I disremember what SIO uses for

 JRC> buffers, but I would imagine that it's more than adequate.

 

 PC> IMHO: if you want to throw something togather real fast then don't use

 PC> threads, if you want to do it right, then use threads! :-)

 JRC> 

 JRC> :)  I understand.  *sigh*  Is there an alternative to using Mutex

 JRC> semaphores?  They seem to take a *LONG* time to lock/unlock (or is

 JRC> that block/unblock).  I recall seeing some code that used the

 JRC> DosEnterCritSec or something like that to update the global buffers. 

 JRC> That's not "right" but it seemed to be more efficient.



DosEnterCritSec() is worse than Mutex Semaphores.



For example, I saw an anlogy DosEnterCritSec something like this;



(Note: persons = threads)



Say your in the office and you want to go the the bathroom but you don't

want others going to the bathroom while you are.  You go around to each

person and tell them to stop *everything* that they are doing to be sure

that they can't go to the bathroom.  You finish in the bathroom, then go

back around to each person and tell them to resume what they are doing.



A mutex semaphore is like going to the bathroom and locking the door.



:-)



Phil Crown

pcrown{at}airmail.net

http://web2.airmail.net/pcrown/



--- Blue Wave/OS2 v2.30