Graves on SOHO Technology

VoIP Supply’s Garrett Smith has an interesting post the other day. Actually, we can cross reference a couple of posts to follow his line of thinking. Firstly, he offers some advice on selecting an analog gateway in a post titled, “Four Keys To A Successful Voip Gateway Purchase.” It’s a good basic overview on VoIP gateways for the beginner.

I’ve made my thoughts on gateway/ATA devices well known in the past (here, here and here). I see them as relics of a prior era of VoIP, and to be avoided if at all possible. However, some people feel that they are truly necessary. Whatever floats your boat.

A little later on over at VoIP Supply’s VoIP Insider Blog Garrett poses the question, “Where Are The Open Source VoIP Gateways?” That is a fine question and one the deserves an answer. It also needs to be broken down a bit to get to a clear answer.

What constitutes an open source gateway? Is the hardware design open source in the same fashion as David Rowes excellent Free Telephony Project? Or is this really all about the software being open source? Clearly Asterisk and Freeswitch, the open source VoIP leaders, are both capable of performing such gateway functions given suitable supporting hardware.

There’s the trouble…”suitable supporting hardware.” If someone were to offer an off-the-shelf package of software, a sort of DIY kit for a gateway device, it would need to address a wide range of available interface boards to be useful to a significant group of people.

There are the well known products from Digium and Sangoma that’d be top of the list. However, I suspect that underlying Garrett’s inquiry about open source is some concern for cost. That suggests that a gateway would likely also need to support cheaper clone cards from Asian manufacturers to gain a cost advantage.

In reality, such a gateway can be assembled right now, and it’s not that difficult. But it is 100% completely DIY, so not something that a totally non-technical person would undertake. It’s the kind of thing that a VAR or consultant would do for you. Come to think of it, an awful lot of the open source realm is like this. Very capable, but not a finished product…it’s a kit. Some assembly required.

In the end it’s hard to know if there’s any merit in building one’s own VoIP gateway. There could be if you have specific needs that are not met by existing commercial products, for example transcoding between wideband codecs. That you have such a requirement suggest that you are not likely a VoIP newby, so just dig in and get it done. The time and effort may not be justified if your needs are more basic.

In the end I suspect that there are enough gateways commercially available to address any common need. Could they be cheaper? Certainly. Could they be more flexible or have better firmware. Absolutely! Is open source a path to get there? I’m not so sure.