-=> Quoting Roger Sen Montero to Darin Mcbride
 TM>> How do you make it volatile, and how does the compiler complain?
 
 DM> On the marked line (!!!), the compiler complains that it cannot find
 DM> a function "queue >::push(char) volatile".  The function,
 DM> without volatile, exists, though.
 RSM> STL (specially HP's one) has a lot of problems with multi-threading.
 RSM> SGI's STL is much better, but it was not designed with multi-threaded
 RSM> concepts in mind.
True - which is kinda why I was hoping to make it volatile and force my own
semaphoring around it.  In fact, the SafeQueue I posted originally "HAD-A"
dequeue as a member variable, and merely enacted Queue's adaptation, with
semaphores surrounding every call.  That didn't work, and thus the problem...
:-)
 RSM> Recently there was some posts in comp.c++.std about multi-threaded
 RSM> STL (and some companion files, including bstring.h)
Hopefully that'll make it into the platforms that have threads...
 RSM> I have to do some tests with multi-threading issues and STL (probably
 RSM> using OS/2) so I'll post a message as soon as I finish them.
That would be appreciated.
 RSM> Moreover, you cannot push chars in a queue >. You should
 RSM> push deques of chars.
As I've recently said, my STL is as outdated as my C++ compiler... and it
uses that format because it doesn't support default template parameters.
 RSM> ... Without the letter U, units would be nits.
Sounds like a little nit-picking... ;-)
... Back up my hard disk? I can't find the reverse switch!
--- FastEcho 1.46
---------------