[Asterisk-Users] NetworkWorld article on Open Source Telephony

[John Todd]

At 7:06 PM -0700 on 6/8/04, George Pajari wrote:
>An interesting article for those needing ammunition to sell Asterisk within
>their organisation or to others:
>
>"Is open source IP telephony ready for prime time? Yes"
>by Zenas Hutcheson, St. Paul Venture Capital
>Network World, 06/07/04
>
>http://www.nwfusion.com/columnists/2004/0607faceoffyes.html
>
>On a related note, they also have an article arguing the contrary position
>(see link within article). I'm too busy right now to write up a response
>showing the flaws in that column but others on the list might wish to
>contribute to the fray.
>
>George Pajari
>www.netvoice.ca
>www.IP-Centrex.ca

The opposing view had some good points, though I don't agree with 
many of his comments.

http://www.nwfusion.com/columnists/2004/0607faceoffno.html

I'm not even going to try to post a reply on NetworkWorld's broken, 
ad-strewn, and ambiguous forum manager.

I think I can disagree with all of Zeus' comments except this: 
management(*) for IP telephony is just as important as the telephony 
itself.  Without the ability to measure, manage, and examine 
performance, it is a tough sell for open-source software in the 
enterprise.

Perhaps that doesn't matter, actually.  Enterprise isn't really where 
Asterisk is written and supported, so we don't see the robust 
features that an enterprise would require.  Remember: there are no 
sales brochures for Asterisk, and the CTO who is looking to implement 
Solution C or Asterisk will not have anything to use in the Asterisk 
column except for (maybe) my feature spreadsheet and an enthusiastic 
network admin who runs it at home.  This will not typically lead to 
Asterisk as the winner.

I am not saying that this is good or bad, actually.  It's neutral. 
The purpose of Open Source is not to defeat commercial 
implementations of the same features, but to provide a "better" 
solution for some people who want to get in there and make things 
work exactly they way they wanted, if they have the spare time, clue, 
and don't have any money to pay someone else to do it.

JT



(*): for a quick definition of what "management" means, here are some 
concepts: provisioning interfaces, per-stream QoS examination, 
overall QoS examination, call routing interfaces (GUI or otherwise), 
cost control and cost examination tools, etc.   You're saying "Well, 
all of that can be easily built!"  Sure it can, but careful with that 
word "easily."   The question is: are these components a patchwork of 
third-party tools, or is it a well-planned whole-system design?  Is 
management an afterthought?   As an example of what enterprise users 
might need, view this post and note that there have been no movements 
towards answering these items:

http://lists.digium.com/pipermail/asterisk-users/2003-July/014965.html

Again, this is not a fault that these management reports don't exist. 
If nobody develops these reports, then maybe they're not used by the 
people that use Asterisk.  Enterprise users aren't so hot on 
developing things themselves, so maybe this just languishes, and so 
they don't use Asterisk (yet?) because the combined effort of doing 
all that stuff is just more than it's worth when they can have the 
CFO sign a check for Vendor A to get it all done.