Tue Sep 23 17:30:28 CEST 2003
Short answer: Yes, and RFC 3525 standard (although I haven't used
_every_ feature in megaco, so I'm not qualified to say so :-)
Long answer: We used the OTP Megaco stack for acceptance tests here, and
while I created the tests I did find one problem using topologyRequest,
and another due to the ABNF having a comment which should really have
been part of the ABNF grammar, but that was all.
When we conducted the acceptance test, no issues were raised against OTP
Megaco, and the test revealed numerous protocol issues with the third
party stack implementation, delivered to us by a reputable protocol
stack implementor for our internal use.
So I confidently believe that the rest of OTP Megaco stack is very
solidly compliant to the standard. The issues raised against OTP Megaco
were quickly resolved by the OTP Team at Ericsson.
I also am confident that any interoperability issues will be due to
third-party stacks being non-compliant to the standard.
Why are we not using Erlang/OTP Megaco for our product? The reason is
entirely political, and not based on the superior technical merits of
the Erlang/OTP Megaco stack.
There is only one person in the company (me) who uses Erlang. Having
produced a working prototype of the product mostly using Erlang whereas
all the rest are still struggling to debug C++ code and still have not
matched the performance of the prototype has caused some hurt pride. I'm
not a better programmer, I'm just using a better tool :-)
Carlos Rodríguez Alcalá Villagra wrote:
> Would the Erlang's Megaco Implementation work with other Megaco
> implementations? How standard is it?
More information about the erlang-questions