Open source VoIP has been touted as bigger than Linux, and the competition is hotting up in Europe as one of the US pioneers crosses the Atlantic.
The normally sleepy world of office telephony hardware, the private branch exchange (PBX) systems which put you on hold and transfer you to people’s secretaries, has already been shaken up by the arrival of voice over IP.
VoIP has allowed a few new players such as Cisco to join Nortel, Siemens, Alcatel and the rest of the club of companies which supply proprietary PBX hardware at high margins.
Recently, however, open source platforms are emerging which allow organisations to use a cheap off-the-shelf server to do exactly what a PBX would have done, and in many cases more, at a fraction of the cost.
In March The Reg spoke to Mark Spencer, originator of the best-established open source PBX platform, Asterisk.
Now Pingtel, sponsor of another major open source VoIP initiative, SIPfoundry, is making its debut in Europe. Pingtel is a commercial company which packages and sells products based on code from the SIPfoundry open source community.
“Pingtel contributed a lot of code to SIPfoundry but our business model is very similar to that of Red Hat or SuSE/Novell,” says Bill Rich, CEO of Pingtel. “We take the code that is on the SIPfoundry and make it acceptable to enterprise customers, with support, usability, and reliability, and the features that make it possible for the enterprise customers to use it.”
SIPfoundry was founded fourteen months ago through an amalgamation of code from Pingtel, and the ReSIProcate community. Some members of the Vovida community also joined.
Revenge of the SIP
Pingtel was formerly a proprietary company dedicated to selling systems based on the SIP IP communications standard. It decided to go open source, says Rich, because the general drift of the SIP industry was driving in that direction.
“We have always been a leading proponent of SIP and SIP-based communications,” he says. “One of the hallmarks of this technology is to redefine how people communicate. IT has the power to drive a really strong drive of commoditization through the industry.”
In a SIP based world, when any handset or soft phone can talk to any PBX, there won’t be much money to be made in hardware or software, at least not for companies without a global brand name.
So better to be the ones who kick the bottom out of the hardware market themselves, and make their money selling support and services.
“We can either be the commoditizer or the commoditizee*. We think the market is going that way and we wanted to be the one that does the commoditizing,” says Rich. “When commodification enters the value chain, the number of vendors goes up and there is strong downward pressure on prices, and that is good for the customer.”
PBXes won’t become a commodity overnight, though. Open source PBXes are still an unknown quantity, and the corporate market will take some persuading before it abandons the comfort of established suppliers.
The small and medium business world, however, is already showing some interest, and Pingtel gearing up to satisfy it.
“We have just started building our channel… really in the fourth quarter of last year. We now have 40 resellers signed up around the world. Our core focus is on the SMB market with the product delivered through VARs and resellers.”
Star wars: Asterisk versus SIPfoundry
So now that Asterisk has a serious rival in the open source VoIP world, which is better?
“They compare at a pretty basic level,” says Rich. “They are both open source communication platforms.”
The key difference, he says, lies in the fact that SIPfoundry cleaves much more closely to the SIP standards. SIP is essentially a peer-to-peer communications protocol, with most of the system intelligence distributed to the edge devices on the network.
As Rich puts it, “Asterisk is an open source PBX. It is an architectural model that works in a centralised control system. In SIPfoundry, the smarts enter the system at the end point.”
Mark Spencer, CEO at Digium and Rich’s opposite number in the Asterisk community, agrees, broadly:
“In a sense, I would say that whereas Pingtel views SIP as the be-all, end-all of all telephony, we hope that Asterisk will be that be-all, end-all to the degree that such a thing could possibly exist.”
Asterisk is SIP-compatible, but doesn’t have the same degree of closeness that SIPfoundry has to the Internet Engineering Task Force, which defines the SIP standards.
Rich points out that many of the key members of the key IETF working groups also sit on the board of SIPfoundry.
These include Robert Sparks, SIPfoundry president and co-author of the core SIP specification, who co-chairs the IETF’s SIMPLE working group (on instant-messenger style presence applications), and Cisco alumnus Rohan Mahy, Co-chair of the SIP and SIPPING (investigation of new SIP applications) working groups.
For those who are keen to stick as closely as possible to the SIP standards, that’s an advantage. But Asterisk will be more fleet-footed, says Spencer:
“As new needs come along, Asterisk (like any software implementation) can be completely changed at any level in order to accommodate the new requirements, whereas SIP (like any protocol definition) can only do so within the constraints of retaining backwards compatibility and with lengthy debates in a standards body.”
You pays money (or not, given that it’s free software) and you takes your choice. In a market which is likely to expand rapidly, there is certainly room for both.
“While one could view Pingtel and Digium as competitors,” says Spencer, “I think we both benefit from having someone else in the open source space, as it helps lend additional credibility to what we're doing and pushes us to work harder to deliver what customers and users are demanding.”
* We’re not quite sure what a commoditizee is. It may be like a chimpanzee – a gorilla after a thorough downsizing. Whichever Cisco becomes, it’s sure to raise a few eyebrows on Wall Street.