[asterisk-dev] QSIG error -- can somebody explain?

[Matteo Brancaleoni]

On Thu, 2006-02-16 at 16:27 +0100, kpj at junghanns.net wrote:
> >> However:
> >>
> >> The Asterisk libpri is a 'hackup'  as far as Q.931 and EuroISDN goes!
> >> BRI in particular is problematic. Asterisk will not pass EuroISDN
> >> compliance testing etc.
> 

I agree. we're recently passed TBR tests for layer2/3 ,
functional tests (telco specific tests) and "on duty"
tests (again specific tests) with visdn + asterisk.
The whole problem is libpri that doesn't implement
a state machine as on ETSI specs (thus difficult to
extend and implement) and chan_zap that's a driver
that implements all sort of signalling in one driver....
that's scaring.

btw, on our experience in the tests labs, many problems
are also on asterisk side (eg. a wrong dialplan could
make a test fail) and so on....

we're going to buy the TBR tests suites and functional
tests , in order to keep visdn fully compliant
over new versions (is libpri still tested against
TBR on each mod? is bristuff, too ?), and, why not,
we'll test also other implementations like
libpri/zaptel vanilla, bristuff, misdn and so on...

if you need further info on visdn approach, 
please look @ www.visdn.it , or join the visdn-hackers
ML.... currently we support hfc-pci, hfc-4s, hfc-8s, hfc-e1,
vGsm and.... upcoming new cards :)

cya,
Matteo.

-- 
Come to visit us @ CeBit 2006
>From 9 to 15 March 2006
Hall 13  Stand no. E25/1

Matteo Brancaleoni
System Administrator
Tel  +39.02.70633354
Sip  matteo at sip.voismart.it
Iax2 matteo at iax.voismart.it