Fri Mar 16 06:21:33 CET 2007
On 16 Mar 2007, at 2:27 am, Danesh Daroui wrote:
> Also in that file "Barrier Synchronization" problem has been solved
> using semaphores.
Barriers are very easy to solve with message passing.
So are the other examples you mention.
The interesting question is not "are there problems that could be solved
in Erlang using semaphores", but "are there problems that arise in
Erlang programming could be solved BETTER using semaphores".
I note that a certain other programming language, designed over a period
of decades with input from literally thousands of experts for
concurrent programming, while it does have tasks and monitors (which
'protected records') and some other concurrency stuff, does not have
has had anything resembling a semaphore. I am, of course, referring
to Ada 2005
and its predecessors Ada 95, Ada 83, Ada 81, and Ada 79.
The big thing that Erlang is interested in is concurrency mechanisms
reasonable to use across a network as well as within a single UNIX
More information about the erlang-questions