Mike Rowehl: This is Mobility

I think location based services are a fantastic idea, in theory. However the cost structures the carriers generally want to provide, and the cartel of cartography data keeps most of the interesting stuff that could happen with LBS from really making it out to the public. The map stuff is really a sore point for me, because innovation is being stiffled in order to protect entrenched business models.

Most people tell me to grow up and just deal with it. The only real option is to get a whole bunch of vans with GPS in them and make up your own maps, and no one is going to bother doing that at this point. It’s too expensive. No one except maybe Microsoft it would seem. They’ve had vans spotted in a number of places apparently building up their own cartography information. Interesting move on the part of Microsoft, especially together with their Windows Live Search beta being available for Java in addition to Windows Mobile. Microsoft normally shys away from Java. Looks like they might be genuinely interested in opening up the market here, that’s not normally their way.

Even more interesting however is that this could break the cartel. There were very few sources of US map data, and they gaurded their position agressively. The supposedly “open” Google and Yahoo maps offerings carried heavy restrictions on what kinds of data you could use to drive your web applications. Using GPS info to drive the mapping app was specifically prohibited by the terms of service for both. This was done to protect the revenue model baked into providing mapping info to manufacturers of offline GPS hardware, and crippled the more expansive market to protect the smaller established one.

The Microsoft version of the data won’t bear any such restrictions. If Microsoft wants to bake location stuff into upcoming releases of Windows Mobile in a way that allows developers to access cartography and point of interest info directly from their app they would be in a pretty unique position to be able to do so. It would potentially force the other providers to update their thinking and allow their info to be used in networked applications by third parties, or get gradually squeezed to death. Either way works for me as it turns out. I end up with a smile on my face in both cases.

Last night was the December Silicon Valley Mobile Monday at Tellme (thanks Tellme folks!). They provided an excellent overview of voice based services – output, input, and multimodal. Here are a few things that stuck out for me:

• Hearing a side by side of that standard broken up computer voice (the one you normally hear reading back things like payment amounts) with a generated voice that’s been smoothed is really pretty drastic with respect to your ability to recall what you just heard.
• The Tellme folks kept saying “voice services allow you to whisper in your customer’s ear”, good way to think about things. I think that general way of thinking about providing a service applies outside voice services and generally summarizes the intimate mobile experience that everyone should be thinking about.
• There was a lot of talk about trying to turn the web experience into “linear” audio. Meaning that on the web it’s generally easy to skip around. Scan a page quickly for some info, and hop back to the top to check out things in more detail. When you’re driving a voice interface that’s not really an option. You need to think about how to do customer education while continually progressing in the right direction. This is another one that maps well outside voice too. The mobile web experience is nothing like the fixed web experience, and I would say that one of the major reasons is that the interface tends to be a lot more linear. You can skim around on a mobile, but it’s certainly not nearly as convenient.
• In running their own portal they found that matching up the voice to the expected norms for whatever information was being delivered made a pretty big difference in the user experience (ie. someone who really sounds like a sportscaster while they deliver the sports info), and selecting the right voice is a pretty important branding decision.

Last night was also the kickoff meeting for Mobile Monday in Germany! Russell has a post up about it, looks like it went fantastic. Congratulations to everyone involved over there!

The Homebrew Mobile Phone club meeting on December 13th has a fantastic lineup, including Benoit Schillings talking about some TrollTech efforts like GreenPhone and a discussion of the Open Source Java effort when it comes to ME. Thanks to John Kern for posting it, otherwise I might have missed it. Now I’m subscribed to the calendar though, so that won’t happen any more. John’s own Symbian Programming SIG is the day before and will feature Ravi Belwal talking about mobile TV. Definitely not going to be a slow month.

I really liked the post that Paddy put up recently at OpenGardens, particularly this bit:

In our open garden, the service provider is the garden designer and horticulturist, who will be creative in order to captivate his audience. He doesn’t just hand over his wares at the gate. These open gardens will be enabled not just by openness in platform technology, but openness of ecosystem for service providers.

“openness of ecosystem for service providers” really sums up a whole set of issues quite succinctly. It also got me thinking about open standards and mobile development. Paddy is definitely correct about the advantages of applications in mobile that don’t follow the traditional breakdown of the role of the client and the role of the server in online interaction. The GMail java app certainly doesn’t feel like other applications that talk IMAP or POP back to the server, and Opera Mini also has a very different feel than most web browsers, even on a smartphone.

Of course this is just the reason that operators have always used for not cleaving to open standards for stuff like messaging and until recently web access. I’m wondering if the way to tip the scales is to develop some open infrastructure to begin with, outside the carrier domain but structured in a way that it should be easy for the operators to adopt when the time comes. Essentially additional instances of efforts like Funambol, a company building open tools around the public SyncML standard. The kinds of things that work without a carrier cooperating, but work better when woven into the ecosystem.

What other areas would fit well into something like that? XMPP for messaging obviously, VOIP services, direct support for SIP. Perhaps identity and profile info, LID, OpenID, something of the sort. Just thinking out loud here, I think it would be interesting to put together a vision of what a future mobile network operator could look like. It’s something that we’ve spoken about a lot, but never really laid out in full as far as I know. Anyone know of a pre-existing instance I can take a look at?

This Monday is the December 2006 Silicon Valley Mobile Monday meeting. I haven’t been out and about nearly as much as I used to be, I’ve been busy getting us to half a billion ad impressions served with as few issues as possible. That’s Mike Mettler who wrote that message by the way. So to those who asked, no, I am not a product manager. I remain firmly planted on the engineering side of the house.

I’m definitely looking forward to connecting with a bunch of people who I haven’t been able to keep in touch with, many of them who I met at Mobile 2.0 but haven’t spoken to since. I was actually considering putting together a presentation of my own for this MoMo when the topic was proposed. I’ve been meaning to spend some time digging into SIP and Asterisk in particular, and I figured that trying to put together an overview presentation would be a great way to force myself to take a look at the lay of the land and draw some conclusions. But even as of last month I knew I wouldn’t have the time to do that myself. However if you’re working on stuff of the sort definitely swing by the meeting. I’m sure you’ll find a bunch of interested folks for conversations.

With the official version of Opera Mini 3 out now I removed the beta version and installed 3.0 release over 2. And then a half hour later as I was waiting for breakfast realized that Bloglines mobile isn’t working in the new 3.0 version. If that’s one of your common stops you might want to hold off on the update for a bit.

Mini 3 does have it’s own feed reader, which I fooled around with for a bit. I’m not really sure I like it though. There were a few feeds already in the reader when I started it up and they were kinda cumbersome to delete, I got some errors when trying to load the feed for my blog (which were solved by just hitting reload, server hiccup maybe?), and the functions I use a lot like “keep unread” and “send this link” stuff isn’t there. I’ll probably stick with Bloglines Mobile for now.

That said however, the rendering is great. The small font seems smaller than before, and there’s a TON of readable text on the screen on my E61. The content folding seems to work very well also, pretty impressive.

I found this comment about cellphones kinda amusing:

Today, it almost seems that voice calls are the least-used function of most phones, while Wi-Fi and WiMax use ever-growing amounts of network bandwidth.

What world is that? I wanna go there. To me it seems like everyone I talk to outside of technology and say something like “mobile website” or “mobile web” or “websites meant for mobile phones” the common response is “You can get the web on cellphones?” Penetration is still very very very small, especially in the US. Things get dangerous when people start to assume that the behavior they see out of the four people with whom they share an office in SOMA is the way that everyone behaves. It is most certainly not. I live in Silicon Valley, I know a handful of people with wifi enabled handsets of some form. Taking a look at overall data usage, most of my data still flows through cellular with occasional bursts through wifi.

Don’t get me wrong, I would love for it to be time to change the name of the cell phone to something like Navi. I just don’t think we’re anywhere close to that vision yet.

Since it looks like carriers in the US aren’t really interested in people building SMS applications, I’ve been trying to figure out alternative ways to get messages to my handset. I registered Agile Messenger and played around with that. It used to be free, but now it’s a subscription product. Well worth it for Symbian devices, it’s a great messenger client. However it’s not quite working for what I want to do. I would like to deliver my messages over Jabber (Google Talk), cause I’ve done too much development in the past against Yahoo and AIM only to have the protocol change underneath me and get stranded. So I want to build on top of a public open standard. However my handset does tend to get lost, as in it looses the connection and doesn’t realize it.

Most of the time it’ll reconnect after a while, but not all the time it seems. And I want my handset to let me know when something is up, and not have to send a test message every once in a while to make sure that the link is really active. However, if that connection handling was cleaned up some this would be a decent way for me to get notifications to my phone. Certainly not great for the average consumer, but with Agile up in the background and the notification sound turned on it’s perfect. A little popup comes up saying I have a message in Agile, fantastic. Hopefully the Agile gods will smile and future versions will handle that XMPP connection in a way that detects disconnects more reliably.

I was thinking about trying to hack something together myself to handle the networking, just a simple read from a socket, some kind of ping/pong to make sure we have reliable activity on some schedule, and make a noise when you get real data. I have to imagine people have stuff like this floating around somewhere that I can just grab and use. Pointers anyone?