Om writes that cable companies have had a “whopper” year. He goes on to say, “Cable telephony subscribers currently represent roughly 9% of telephony-ready homes, 10% of basic cable subscribers, 24% of cable-modem subscribers or roughly 7% of Bell households, UBS says, which means that things are going to get a tad difficult for independents like Vonage and Sun Rocket, and also for Bells which are scrambling to roll out their triple play offerings.” Being in the arms supply business, it is natural for me to suggest strategies for the independents and the Bells.
Since the cable companies are focused on POTS replacement business, the independents should migrate away from ATA=based telephony and offer more advanced features. For obvious reasons, I can enumerate these features. But one example will be the ability to control the calls from a “foreign” location – not just via the ATA. Even the early adapters prefer to use the phone form factor; but their contacts are stored in the computer. So why not allow the call be dialed through the computer, but use the ATA for the actual conversation – good old third party call setup? A thoughtful exercise will produce a long list of such features. Of course, the incumbents offer many of these features to their POTS subscribers as well.
But both the incumbents and the cable companies have an handicap – their signaling architectures. Both POTS signaling and NCS, used by the cable cos are stimulus in nature. This means all the features need to be developed in the Class 5 switch or the Call Agent. In other words new features have a long development cycle. But the independents use SIP and so the features could be implemented in the ATA. Development of features can be outsourced if they define a good set of APIs. I think this is where the independents should focus on now.
Om informs us of a new startup called Vyatta that has developed open source routing software that can be run on commercially available components. An interesting implication is that such routers will cost just one fifth of the commercial ones. Om also summarizes other open source efforts that have implications to the telecom industry. He does not say this in his post, but I think Vyatta can impact VoIP as well. Session Border Controller (SBC) is an important component of VoIP architecture. In its core SBC is a “twice NAT”. So it is very likely the open source router software used by Vyatta could be used to build SBC as well. This is an important development indeed.
Jonty interviewed me yesterday and it is available at VoIP Central.
It is a benchmark day for two of my blogosphere friends. Alec found out that his company iotum was named DEMOgod. Video of the demo is available for your viewing. Congratulations, Alec. Paul Kapustka has restarted his blog. Welcome back, Kapustka.
Only a couple of days back I once more wrote that charging a premium for click-to-call is feasible only if the service provider artificially injects itself in the end-to-end flow. Yesterday, it was reported in NYT that AOL and Yahoo are planning to offer a tiered email delivery service. In a nutshell, those that pay extra will have their emails bypass spam filters and will be delivered with a “stamp of approval”. This is a change in the service model and hence needs a closer scrutiny.
Ostensibly, this will reduce the spam problem. But for practical reasons, if not anything else, they will continue to accept other emails as well, except they will be processed by spam filters. This means that there is not going to be any reduction in spam; indeed it will only increase from the users’ perspective because each user will consider at least some of the “approved” emails to be spam. So the real reason must be revenue enhancement. Tom Evslin suggests that if anybody, it is the recipient who should be compensated and that there are mechanisms in place for that.
This proposal from AOL/Yahoo is “evil” because it breaks the end-to-end argument. One of the first casualties is the long-tail phenomenon, which should be THE celebrated cause in Web 2.0. In some respect, insertion of mail server broke the end-to-end principle; but it was voluntarily done by the user. Even spam filtering was at the behest of the user. But this one looks like an unilateral move. Both Tom and Richi Jennings (in the NYT story) suggest that users can walk away from such providers. It sure looks like one more example of Netheads behaving like Bellheads. A nethead will loath to break the end-to-end principle.
In any event, we do not have to worry much. Given always-on connection and plenty of choices to have presence in the public internet with plenty of storage, there is no need for dedicated email servers anyway.
In the raging discussion that is commonly known as Network neutrality, the incumbents suggest that companies like Google are using their pipes for free. But a long time back Tom Evslin (para 4) had pointed out that the way traffic is passed between “subnetworks” that make up the Internet: essentially a subnetwork that is carrying another subnetwork’s traffic gets credit from the latter; between equals the credits cancel each other and no real money is exchanged and between unequals, I suppose there will be some form of settlement. Hence, Tom’s point that the incumbents are not carrying the traffic for free. Indeed, the incumbents should rejoice because most of their customers are receiving traffic from the Internet, thereby allowing the incumbents to accumulate credit.
But it also points out the error in current charging scheme for Internet access. In many countries, the Internet access is tariffed based on the amount of downloads. But it shouldn’t; it is in the interest of the ISPs if their subscribers download a lot. They should be worried only about the upload. So they should change the tariff, charge the customer for uploads and should increase the revenue potential by deploying access systems that have huge upload capacity. It seems that such systems are inherently cheaper to deploy as well.
While commenting on one of Om 's post, John Allsopp suggests that the real target is VoIP providers when AT&T wants to charge the content providers. He continues on to say that "text, even lots of it doesn't suck bandwidth like voice."
In the scheme of things voice doesn't require much bandwidth. A wideband codec like Speex requires only about 40 kbps one way. The problem is in the collective perception that voice can demand a premium. This includes those who claim, correctly I may add, that voice is just an application in IP world. But then when companies like eBay and Google say that they can charge premium for click-to-call, there is a scant murmur to be heard. Shouldn't the stalwarts of "everything is an application" shout at the top of their lungs or at least hammer away in their blogs? Why the double standard?
So here is a request for any blogger who is going to lament about AT&T "double dipping": please add at least a passing comment challenging the click-to-call business model. If you disagree with me, then tell me why I am mistaken.
The announcement from Tello is more than one week old. I m sure you would have read about it from many sources by now. Jeff has neatly assembled many of these references. Based on his conversations with Jeff and Doug Renert, CEO of Tello, Andy has summarized the idea behind Tello. It took me this long to fully grasp and internalize what Tello is all about. This post is a summary of my understanding.
Tello service has three major components:
These descriptive titles may give the impression that there is nothing new here. For sure, all the features have been known for a while; indeed many of them have been available in FWD Communicator (or pulver Communicator, as it was previously known). The newness is in the whole packaging as a commercial service.
For example, the Presence feature collects presence information from many sources - “public” IMs like MSN and Yahoo, mobile device like Blackberry and status of phones connected to a PBX from one of the partner companies. As Om has observed, this is lots of information and the presentation of this vast amount of information in a usable manner is critical. I think Tello recognizes that and presents only a synthesized version. For example it combines the presence information from all the IMs to a single value.
Tello client can be used for text messaging and also for real time communication like voice and video. The communication can take place as long as both the parties belong to the same “public” IM service or they belong to a partner service provider (“federation partner”). I suppose they will build up their partner network over a period of time. The interesting twist is that there is no need to launch the native client but use Tello client alone.
They use Persony’s web sharing application for the screen sharing feature.
Alec has commented that Tello and iotum are complimentary. I am not sure whether I agree or disagree. In some respects they have different approach in presenting presence information to the user. Tello gives information on different modes of communication and let the user select a mode of communication for a session. iotum on the other hand selects a mode of communication based on multiple factors. In other words, with Tello the calling party has the choice and with relevance engine, the called party decides. I am not suggesting one is better; but there is a clear difference in the approach.