[Asterisk-Users] OT (kinda): Justification for adding Asteris kto the business plan

[Colin Anderson]

>The old "lock-in" model of the telecom world was nasty and monopolistic and

>expensive, but by jove it just worked.  

>Because as you've seen from this list ANYTHING and EVERYTHING can, and 
>does, go wrong with this technology.

Yes, true, but this is the nature of the beast. It (Asterisk) is an
infinitely configurable platform that runs on anything. This essentially is
it's greatest strength, and it's greatest weakness. It's easy to make a
crappy Asterisk box. It's hard to make one that's five nines. Unlike a
closed box solution, where you just pick up the manual or go to the boot
camp, there is no replacement for hands on experience with Asterisk. When I
look back at my first naive attempts at implementing Asterisk, I laugh at
the stupid assumptions that I made, and I thought that I knew telecom pretty
good. 

I have come to realize that this is the fundamental problem with documenting
Asterisk. There is so many ways to do things, and so many ways that things
can be acceptable to one guy, and not acceptable to another, that
documenting it ISO-9000 style is impossible. Can you imagine the tombe that
would be produced from the aggregate knowledge of this list? And the
contradictory nature of it?


>When I look at VoIP, I see nothing different between setting up and
>configuring perl scripts for AGI/* Manager interfaces (system admin
>scripting) to plugging the BRI into a Digium card from the old T1 router.
>Granted there are some differences in command configurations and might have
>to label a wire or two "Voice" instead of "Data" but nothing earth
>shattering.

Oh, be careful with that assumption. Because, really, Asterisk functions as
an abstraction layer between data and telecom. It's seductive, because we
know data, and we figure "hey, I just SSH into the box like normal" but the
core of problems shown on this list are usually not Asterisk itself but the
interface between Asterisk and the PSTN and / or the PSTN itself. You have
to have good telecom fundamentals or you will be in trouble very quickly.
Can your guys look at ISDN L2 frames and determine what's going on there?
'Cause if they can't, your rollouts are going to be disastrous. Unless you
specialize at first in SOHO boxes or VoIP boxes only, but there's not a lot
of margin there. You want to replace a MICS or two to make it worth your
time. 

Maybe eat your own dog food. Convert your existing phone system to Asterisk
and make sure all of the consultants have a hard phone or a SIP phone at
home so they can get calls at home. Then they might get excited about it.
This way, you build support and mindshare internally and get experience all
at the same time.