[asterisk-dev] About asterisk development plans

[John Lange]

On Wed, 2009-03-18 at 18:38 +0100, Olle E. Johansson wrote:
> To show the complexity of what we're discussing:
> 
> Let's take blind transfer.

[... snipped for brevity: The point Olle is making here is that testing
everything is very very hard ...]

For certain this task is not easy.

Perhaps the focus of this discussion should be on whether or not we feel
it's possible to identify a "core" set of features and configuration
scenarios which could be implemented in a testing environment.

I truly believe this is quite possible. We don't need everything
included on the first day and once something is up and running we can
expand it to include more and more tests.

If you really stop to think about what things are "core" I don't think
the list would be that huge.

To test the SIP features, two ends of a SIPp would work. That is what
SIPp is designed for.

For analog testing I don't think you need to get into anything so
complicated as robotic phone dialers with analog phones, simply
connecting an FXO to another box with an FXS should work well enough.

And a similar technique for PRIs, two PRI cards with a T1 x-over. I do
this all time to test newly built Asterisk boxes before deploying them
to client sites.

And here is a list of features that I think are worth testing:

Blind Transfer
Supervised (attended) Transfer
Call Forward
Call Monitoring
Call Parking
Call Pickup
Call Queuing
Call Transfer
Call Waiting
Caller ID
Conference Bridging
Three-way Calling
Do Not Disturb

Call Detail Records

DTMF

Music On Hold
Voicemail

This is just "back of a napkin", but 17 features tested on 3 transports
(SIP, Analog, & PRI) is 51 tests (and not all tests are relevant for
every type of signalling).

Sounds doable as a starting point.

-- 
John Lange
http://www.johnlange.ca