On 07/02/2012, at 11:57 AM, Chuck Remes wrote:

> While watching this thread, I can't help but think that you wish 0mq was a 
> *much different* library than it is.

I have no "wishes" :) I'm not even using 0MQ at the moment.

> I'm hoping that sustrik stays engaged with it at some level to indicate how 
> much is feasible for his long-term vision (getting 0mq into the kernel) and 
> how much of it is superfluous to that vision.

Putting the locks inside the interface would seem contrary to a kernel 
implementation:
mutex are user space entities.

OTOH getting the API into Posix would probably *require* thread-safety.
I doubt anyone on the Posix committee would want yet another non-thread-safe
API added. If the spec specified thread safety, a kernel implementation would
provide that its own way.

> I think it will be quite difficult to accommodate your proposed changes while 
> providing any kind of backward compatibility. 

I don't understand. The current API is entirely backward compatible.
It breaks nothing. It does not threaten the current model at all.
It's a pure extension (to the API: the implementation is changed a bit
to provide the support).

--
john skaller
[email protected]




_______________________________________________
zeromq-dev mailing list
[email protected]
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Reply via email to