Eight O'Clock Fusion

Woke up at six to catch a conference call at 8:30. Not the best way to start the day in a merry mood, but it was actually a pretty good (if bloody exausting) day.

Came home, surfed the news and thought: Wow. This has got to be "bash the traditional telco day" or something. Yahoo bought DialPad, Motorola is making noises about its post-Bluephone offering around the (yeuch) V3 and UMA, Skype launched a new voicemail offering, and a gazillion people (primed to annoy me about Bluephone) threw me this old piece on Wired.

What struck me, however, was the dismissive and somewhat misguided way in which Guy Kewney tried to compare Fusion/Bluephone and Skype. In comparison, Carlo Longino's piece was much more informative and unbiased.

I've read Kewney's columns for a decade now (possibly more, since I'm not entirely sure when I started buying Personal Computer World and other British computer magazines), and the lack of touch with the reality of the mobile industry that he displayed in that piece was nothing short of remarkable.

Now, I'm as much of a VoIP proponent as the next guy. Or rather, I actually use the technology - both Skype and standard SIP-based stuff, am positive it has a lot of potential, but am highly skeptical of both its mobile applications and 99% of the so-called "business models" around it, probably due to the simple fact that I actually know how it all works.

In the interest of full disclosure, work at Vodafone Portugal. I am not, however, connected to the Fusion stuff - I'm a few thousand Km and several national and organizational boundaries away, so I'm not going against my by posting this.

That said, I really think Kewney (on the technological points he tried to make in his piece) tried to compare peaches with potatoes (just to make the point that he makes it look like he hasn't really caught on to what are apples and oranges).

What Kewney's piece boils down to is that pretty much without any rational argument, he starts by saying that a Skype-based PDA using Wi-Fi would be a better alternative than a mobile phone with mobile-to-fixed handover (UMA-based or not), because:

  • It has a coverage footprint of near-absolute zero when compared, say, to your average GSM operator.
  • It stacks the proprietary Skype protocol on top of an IP connection without any sort of guaranteed QoS, riding on the 2.4GHz unlicensed band and using the PDA's CPU to perform all the processing required.
  • Operators are evil (yes, he does make that point, rather explicitly).

I believe the proper British term to adopt in reply is "Bollocks", considering that:

  • Any GSM or UMTS operator's coverage makes Wi-Fi availability look like a bad joke (and before you rush to the comments link, see to understand precisely why I say this, and the realities of scale involved).
  • Mobile phones use standard, interoperable protocols that just work on any mobile network, and perform far more efficiently than any PDA the tasks of (in, say, UMTS) encoding your voice, doing echo cancellation, slicing it into ATM cells, managing the radio link, allocating enough codes to shunt said cells over the radio with guaranteed end-to-end QoS, and managing hand-overs to any neighboring cells - while you drive on the highway if needed be. No current PDA running Skype has any hope of doing that as efficiently and in a package that small and with as long a battery life, and last I checked (around three weeks back), pretty much all Wi-Fi-based SIP phones had proprietary (and lousy) hand-over abilities. And no, saying that "it's just around the corner" doesn't cut it, I've been hearing it for the last six years. Not to mention that he is simply ignoring the issues involved in terminating calls to and from other networks.
  • Operators have spent the last few years ensuring that the above just works, regardless of where you are and what handset you purchase. They couldn't "sabotage" a technology that wasn't working three years ago, and is barely starting to work properly now. And it's not going to be any sort of success with gimmicks like this.

I could also rant on about the way is trying to ride the on Intel bandwagon by saying it wants in on (yeah, right, like they wanted to license it at this point), but I guess one rant is enough for today.