Wednesday’s keynotes at VoIP Developer Conference
were kicked off by Sangoma
President and CEO David Mandelstam speaking on the subject of Open Source Telephony 2006.
As Rich Tehrani introduced David, he engaged the audience a bit, and it was determined that at least one developer had “stayed up all night working on the pigeon problem.”
Mandelstam began with an overview of the state of the Open Source Telephony market today. He pointed out that we are in the midst of a transition from legacy TDM-based telephony to ‘all VoIP,’ and that there was no question that it’s going to happen, it’s just a question of nailing down a specific time frame.
“The transition,” he said, “is being driven by the feature set that VoIP offers; a feature set that is difficult to reproduce by other means.”
Other drivers of this transition include the fact that IP PBXs are becoming ubiquitous, wireless is increasingly replacing copper (e.g., Toronto’s omnipresent WiFi coverage), and PC power is getting more powerful and less expensive. Mandelstam said that the cost per MIP of PC processing power is declining, which makes it easier to throw more power at applications like VoIP, which are very demanding applications indeed.
So where then, does the Open Source Telephony market find itself today?
Currently, according to Mandelstam, Open Source Telephony is a $250 million a year business and it’s enjoying 100% year over year growth.
Why the explosion? Unlike with the battle between Linux and Microsoft, Open Source Telephony developers are fighting a single entrenched incumbent. There is no single winner in the traditional PBX or call center market. There are few if any compatibility issues. And of course, it’s relatively easy to migrate from a key system, add functionality of open source underneath it, and continue using your old systems until they reach the end of their useful life.
The market is wide open.
Today’s software has matured to the extent that it can provide commercial grade reliability and stability at low cost.
Simple telephony hardware is available with optional restricted DSPs to handle PSTN interfaces, with the feature sets needed to make the system work.
Once you nail down the PSTN connectivity, PCs are very good at moving large blocks of data around, which is the main function of telephony.
Mandelstam told the audience that Asterisk is far and away the leader in the Open Source Telephony market — and that the basic code is being improved all the time, with the current emphasis on reliability
Current distributions of Asterisk are focused on adding features, most noticeably with regard to improving the user interfaces. And, Mandelstam said that successful businesses are being started out based on Asterisk systems on generic PC hardware He gave the example of Fonality.
According to Mandelstam, “The larger opportunity seems to be with enhanced systems with easier user interfaces based on selected, tested hardware.”
“There’s definitely a mass market for these kinds of things.”
Mandelstam spoke about HMP, and how a standard sub-$5,000 PC can support more functionality and more users than any thought possible.
In his experience he’s seen such PCs handle PSTN switching of 960 calls on four octal E1 cards; 480 conference calls; unlimited voicemail; VoIP packetizing (1,000+ calls); call routing, IVR, program integration (1,000+ calls)… “These open source solutions can support much more than people originally imagined,” he said.
In closing, Mandelstam shared his positive outlook on the future of the Open Source Telephony market. The market “…will continue to grow at current rates for many years to come,” he said.
There is lots of room to grow, with few barriers to entry. He also went on to say that Open Source Telephony systems will take market share away from traditional PBX manufacturers, and that we should definitely watch for its adoption by mainstream players.
A time is coming when open source projects will be mixed with commercial software as per the Linux model, which is a mix of open source and commercial.
“Watch this space,” Mandelstam concluded.