eTel: The open source phone crowd talks to itself

The Emerging Telephony (eTel) conference brings together the open source and web telephony community. It's a bit of a geekfest, with a lot more focus on technology than on business issues, but some of the trends and ideas I saw at it this year are worth sharing.

I wrote a business-oriented summary for the Rubicon weblog, focusing on three issues I saw at the conference that I think have broad relevance to tech companies: the emerging standards for identity management, the vulnerability of mashups to unexpected failures and security holes, and the integration of voice services into websites. I won't repeat that discussion here; you can click here to read that post. What I'll do here is go into more detail about the other interesting things I saw.


First, a couple of overall comments about the conference:

We're breathing our own exhaust. There were about 280 people on the attendee list for the conference. Several speakers asked for a show of hands from all who worked at operators and handset vendors. Virtually no hands went up. Looking through the attendee list, I saw one or two people each from BT, Sprint, DoCoMo, and Vodafone. France Telecom/Orange had several people and was a sponsor. On the handset side, there were individuals from Nokia, Motorola, UT Starcom, and Palm. But absolutely no Verizon, TMobile, Cingular/ATT, Samsung, LG, RIM, SonyEricsson, etc. Basically, the conference consisted of open source telephony enthusiasts and Internet companies speaking to themselves and confidently predicting the downfall of the operators.

In Silicon Valley we call this "breathing your own exhaust" -- you bring together a bunch of people who already agree with each other, and they reinforce each others' opinions. A lot of conferences are like that, and I don't want to ding the O'Reilly folks who ran the conference because they can't really control who comes to their events. But it was symptomatic of the lack of communication between Silicon Valley and the operators. That missing dialog affected the presentations of some of the speakers.

An example: An otherwise excellent speaker on identity, Kaliya Hamlin, tried to suggest some potential win-win strategies in identity management that would help users while still enabling the operators to make money. She suggested that the operators offer identity services and tie them to a commerce engine, so users could buy things and charge them through their wireless bills. It's a great idea, and the Japanese operators are already doing it. But I know from personal experience that as soon as you mention "billing" to most of the US and European operators they run screaming from the room. Their billing systems are already too complex, held together by chewing gum and spiderwebs, and the thought of making a big change to them is terrifying. Kaiya gets an A for effort, but in a forum that had a balanced representation her idea would have been discussed and debated rather than just tossed out there.

Somewhere, sometime I'd like to see a forum where the operators and the Internet folks could engage as equals and attempt to find common ground. Agreement might be impossible, but at least it'd be fun to watch.

The ideas for mobile social communities give me the creeps. People who go to conferences tend to be outgoing; one of the reasons they go is because they enjoy interacting with other people. The O'Reilly folks cultivate this attitude very well, with a lot of breaks and discussion sessions that run late into the night. The energy level is great, and reminded me a bit of some of the early Mac and Palm developer conferences.

But when you take those same extroverts and ask them to design social networking software, the results are creepy. Don't get me wrong, I think online communities are one of the building blocks of the future economy. But the type of community I prefer is one where you interact only on particular topics that you want to share. What the extroverts are designing is communities in which deep knowledge about everyone else is a goal in and of itself.

Case in point: Jyrki Engestrom of Jaiku showed an S60 presence client designed to "bring your address book to life." What it does is collect information about your current status and relay that to everyone else on your contact list. Jyrki showed us the status report on his wife -- at the time of his presentation, she was asleep, with the ringer turned off, at home (the software picks up where the user is through location services). He could also check his wife's calendar to see who she's meeting the next day. In the future, the client will be extended to show what music the user is listening to, and will use Bluetooth to report on anyone else who's nearby. The client also has an API, of course, so we can all extend it to give even greater layers of intimacy.

And of course it'll all be paid for by (drumroll, please) advertising that gets tailored to all that personal information the phone is collecting about you.

Jyrki says the effect is like having a blog, but with smaller chunks of content -- a lifestream, a "living address book." To me, it was more like a system for making everyone the star of their own little reality TV show. Think about it, an entire world that works like LiveJournal. I've got no problem with the folks who enjoy that, but I suspect this will be yet another mobile feature that cuts the mobile market into segments -- some people will like it, and some people would rather bodysurf in a gravel pit.

(For the record, I'm the only luddite who thinks this way. Check out this amusing rant from Steve Bryant regarding Twitter.)

Yahoo is more fun at a conference than Google. If you're ever at a conference and have to choose between attending a talk by a Google exec and one by a Yahoo exec, go to the Yahoo one. Yahoo people sometimes (not always) share ideas and interesting projects they're working on. They seem to hold the same philosophy as many web app companies, that if you share good ideas you'll get back more benefit than you give away. Google, meanwhile, is tighter than Spandex jeans on a 16-year-old. They don't share information on any unannounced projects, so about the only thing they can do at a conference is talk about work they already completed. At a conference I attended last year, a Google manager gave a talk on how they created Google Calendar. Get this -- they interviewed users. Oooooooh. The Google speaker at eTel, Chris Sacca, was much more entertaining. But as expected he talked about history -- Google's work to build a WiFi network in Mountain View, California.


Other highlights

Microsoft has fun with SMS. Sean Blagesvedt and Rajesh Veeraraghavan of Microsoft Labs showed off an experimental project that lets someone send instructions and database queries from a mobile to a PC via SMS. On the PC, you use an Excel template to control how the system will respond to the SMS commands. The PC can send messages back to the phone via SMS. The whole system is effectively a tiny command-line interface between the phone and the computer.

That would be an interesting but not-too-compelling experiment in the US or Europe, but they're aiming it at developing countries where it's difficult for a small business to set up a separate web server, and 3G connections are rare or expensive. Using the existing 2G infrastructure and a low-end PC, a company could set up a basic information access system. I liked the idea.

Coincidentally, Don Norman just wrote an article pointing out the rebirth of the command line in search engines interfaces and in computer operating systems. So this stuff is popping out all over (thanks to David Beers for the link).

Universal identity. I mentioned Kaliya Hamlin's talk above, and I hope it didn't sound like I was picking on her, because I thought she did one of the best talks at the conference. She described the efforts that several organizations are making to create a single, unified system for verifying your identity online. The goal is that you log in once per browsing session, and after that you're automatically logged into every site you visit. There's a lot of work going into making this an open standard, so various competing identity services can operate underneath it. I was impressed with the work that's going into it, and I think it'll be useful infrastructure for the industry, even though I'm not sure if it's solving a burning need for the average user.

Tidbit from the front lines in a municipal wireless deployment. Chris Sacca of Google said the WiFi network it installed in Mountain View, CA, is seeing steady growth. One notable statistic: in any given day, about 90% of the base stations get some traffic. All of them get traffic in a given week. Traffic is highest in the lowest-income parts of the city. That sounds strange, but think about it for a minute -- the rich people all have cable or DSL already. Municipal wireless is often depicted as a benefit for rich people, but maybe by democratizing access to the Web it actually helps poor people the most.

(Of course, "poor" in Mountain View is a relative thing. Maybe "less well off" would be a better term.)

He also showed a lot of amusing pictures of dinosaurs, a pointed reference to the operators. That pretty well reflected the attitude of most of the speakers at the conference.

Progress reports on open source mobile devices. A couple of companies showed works in progress on open source mobile devices. TrollTech demonstrated its Greenphone, which is prototype hardware of a phone running its QTopia Linux (demo devices are, unfortunately, what you have to produce when you don't have a local licensee shipping your latest stuff). Fonav showed a new UI and PIM suite built on the Greenphone, and OpenMoko discussed the Linux phone it's developing. All of the products were interesting, but I kept wondering what unique problems they would solve for users. It was hard to tell, because the time available for demos was very small.

I don't think open source phones that do the same things as the smartphones already on the market are going to excite many users. The thing holding back smartphone adoption isn't the proprietary nature of their operating systems, it's their lack of compelling functionality for most users. If the open source phone guys could turn their energies loose on that problem, I think they'd have a better chance of changing the world.