Yesterday, friend and colleague Ted Wallingford posted H.313 is DEAD and mentioned a post of mine to some tips for VoIP that seemed to indicate H.323. was a viable, competitive alternative to SIP. Not so.

I recently posted a paper on the workings of H.323 to the Realtime VoIP Community Reading Room. The intent of the paper was to describe H.323 and to explain how it works as an umbrella suite of protocols. It was not an endorsement or recommendation of H.323. If you read the explanation of what steps are required for call setup using H.323, you'll understand why.

If you really want to see H.323 in action, fire up Microsoft NetMeeting (if you can find it), and make a voice call. NetMeeting was the closest thing to fully H.323 compliant client software we ever saw.

H.323 produces bloat. Programs and applications using it are overweight, slow, inefficient and sluggish. Put another way, H.323 will lead to a resource pig of a program on your computer.

H.323 is still supported in most VoIP hardware solutions from the big hardware vendors. Sometimes they recommend it for specific business and design needs. That's ok, but you better have a plan for decommissioning it too. Don't implement it and plan to leave it for the long term. And don't let a sales team spin you on H.323 without solid, comprehensible and defensible reasoning.

H.323 really is dead as an umbrella suite for VoIP delivery. Parts of it, the protocols under the umbrella are alive and week. T.120 data sharing standards have a long life left ahead of them. RTP and RTCP are very efficient and useful. What's useless is the umbrella standard of H.323 for VoIP. That dead umbrella can be retired to the archives.


Technorati Tags: H.323. dead standards, dead technologies, VoIP protocols