[asterisk-users] Stable-Stable Asterisk

[Diego Iastrubni]

On Friday 24 August 2007 20:27, Russell Bryant wrote:
> Let *me* ask a question.  :)  What level of heavy regression testing would
> you *expect* of an open source development team?
Unit tests. What unit tests are available for me to try?

Can you share the tests your lab is doing with the Asterisk community? We 
might be able to test it with you and help you with it (even add some tests).

> We really do try very hard to test all of our changes.  We have community
> members that work very hard to help test out the more invasive changes.
> Furthermore, we have a lot of people run the code from the release branches
> directly so that regressions are caught quicker, and hopefully before they
> make it in to a release.
And yet people complain a lot. (see the post of  
Steve Totaro bellow me). 

> At Digium, we have a department dedicated to doing testing of our products,
> including Asterisk.  Every bug that is found as a part of this testing gets
> fixed in the open source branches as well.
Well, I am a n00b in the Asterisk world (only 2 years in it). But I do have an 
active @kde.org email account for 5-6 years, and I do see what they are 
doing. The KDE4 code is 50-200 times bigger (*) (I am not even counting the 
extra* modules, or the dedicated branches which might get the number WAY up), 
and yet they are doing 50 times more testing. Both insecurity and bug fixes 
(think EBN, or the unit-tests been made in kdelibs).

... and that is even before we speak of the huge work Trollteck is making in 
stabilizing Qt releases *before* KDE gets them, and not the huge amount of 
work Novel is investing in usability tests, or the work Mandriva is investing 
in coding, or the infrastructure that RH is investing in linux itself.

What I would like to know is : how I as a person using Asteriks, or deploying 
it, can test it for you. What steps do you recommend to me?. Just for 
reference I just left a mini-system here at Xorcom running with ~30 calls 
SIP/ZAP for 2.5 days, playing Vivaldi to echo test, giving me a load average 
of 2.5-3 for 2 and a half days. It's a start, but I am not sure what else can 
I do.

Can you give us, the community, a standard set of tests we can make to our PBX 
for which we can claim "we tested this system with the same tests Digium are 
testing, it's stable, we and Digium guarantee it".

(*) just for reference, I think Amarok's code is a little bigger then Astresik 
1.4 code. Well at least 1.2. I did not "measure" it, it's just a hunch.