(read my previous post about ENUM Summit)
Today the topics covered mostly ENUM deployment status in the telecom carriers in the US and ITSP in the Europe. Very, very informative, and I realized the following facts:
1. Today ITSPs are still exchanging calls via PSTN network (e.g. IP-PSTN-IP). Because of this the quality of VoIP calls and the value-added IP-based services can not be delivered effectively. So actually ENUM could be the best available solution to bridge all those isolated VoIP islands and making VoIP a truly "on-net" environment. XConnect is providing ITSPs free peering and call exchange services. I chatted with its CEO, Eli Katz, yesterday about how Taiwanese ITSPs can join the alliances. I think it's time to introduce our IPOX to them.
2. Jason Livngood from Comcast Cable illustrated how carriers are using private ENUM mechanism to route their VoIP calls in/ out of their network. But I found that peering is a big issue among them, same old business issues as peering among ISPs. ENUM basically performs DNS look-up and tells you where the call should go to. Ideally, you can simply find any other calling party if you're in the public Internet. But actually there's no real public Internet. Take Comcast for example, they have 19,000 miles for fiber, 2500G capacity, a big big Intranet in fact. It makes no sense for them to route other ITSP calls without certain peering policy, a.k.a. who is paying who. So Public ENUM does not work. Private ENUM does. On the other hand, the standards for "carrier.e164.arpa" is truly important. HiNet or Seednet in Taiwan should seriously study this.
3. Doug Ranalli from Netnumber explained how GSM operators are using private ENUM to route MMS message. Tom Kershaw from Verisign later added 50 cent is the one who is making money out of ENUM :-P
4. Last but not the least, Paul Mockapetris, Chairman of Nominum and one of the designer of today's DNS system, said that the ENUM is still evolving. True. 20+ years ago there was debate on root, one side said no ".com", just country code, the other side (Paul's side) said do both, and see how it went. So this is how the story of 30 million dot-com names began. I don't know if we could be comfortable on this "see how it goes" strategy today since the Internet is not the way it used to be, yet I would just keep Paul's words in mind and remind myself of holding on to the core principle (Scott Bradner said that yesterday too). On the other hand, I remember hearing a concept that if Windows need constant "updates" and "patches" to keep it work, then we may consider to replace this OS with a new one. The next question: What does the new DNS look like?
Anyway, up to this point, ENUM works, comes at the bad time, but a right thing to do.
Comments