Mike Rowehl: This is Mobility

I just got an email from Kiran Bellubbi about the Mobile Monday Chicago group which is having it’s first meeting next Monday:

Whatever happened to J2ME and all the talk about Java on small form factor devices. A few important industry trends make this a topic of increased interest - The J2ME stack has matured and Linux has made a push as a reliable OS for mobile devices and it is evident that people are thinking about how to create an effective IDE for developers to build Java products on Linux based devices.

Our speaker for this Monday is Christy Wyatt VP of Ecosystem Development Motorola Mobile Devices and she will be educating us mob-p’s about how Motorola is supporting Eclipse and Linux for Mobile Development.

W00t! It makes me so happy to see new chapters pop up and know that Russ and I have helped contribute to a movement that continues to grow and thrive. Best of luck with the first meetup Chicago MoMoers!

Mobile video is apparently the next big thing. Everyone knows it. Talk to just about anyone working in mobile and they know that everyone is working on video. The commercials on television and the radio keep telling me I can get the video content I want on my phone today!! For the most part I have to admit I’m not really sure what they’re doing.

I don’t really want to pay a few dollars to download a video in some format that I can’t play under Linux of someone (almost anyone really) shaking their junk. Maybe that Salma Hayek scene from Dusk till Dawn, but that’s about it. And getting videos of the kinds of content I want to watch into the formats I need them in to put them on my devices using the operating systems I choose to use has been and definitely continues to be difficult. And no, I’m not going to switch operating systems so that I can watch videos. I think the very concept of “entertainment” escapes a lot of people.

Yet I try it out every once in a while. Cause there is compelling content out there. What compelling content you might ask? Why, this years SchmooCon videos of course! If the website is to be trusted, the videos contain a record low level of moose. You just can’t get that kinda quality in commercial programming! I simply must have it. I don’t think I have to tell you how often I pull down a video on abusing security measures only to find a big hunk of moose in there stinking the whole thing up. Plus, I’m lusting after 3gp posting like MoMo London does. Not that I feel they’re lording their superior 3gp generation skills over us… but they could be! Makes me feel all hackity.

Here’s the step by for getting a 3gp output supporting set of tools onto Ubuntu Dapper:

• Make sure you have the extra repositories in your apt config and install mplayer and mencoder. I had a few repositories on already, and had mplayer installed before I started this adventure. However mencoder wasn’t in there, so I just cranked up the whole thing, apt-get update, and apt-get install mencoder-686.
• You need a ffmpeg that supports AMR in order to generate the 3gp videos. This includes, for some reason, downloading the codec from 3gpp.org. Yep, that’s right. Don’t even get me started. Just do it and steadfastly ignore the idiocy. IGNORE IT! You’ll need the ffmpeg source. The developers suggest pulling directly from subversion, I suggest listening to their suggestion.
• Once you have the ffmpeg source you should look in the libavcodec/amr.c file for info about where to get the necessary bits from 3gpp.org. Do not, under any circumstances, try to visit 3gpp.org itself. It’s worse than useless. I know you’re about to say something like “why the fuck should I be reading source code if all I want to do is create a 3gp video?” But in this case it will actually give you less of a headache if you read the instructions out of the source code of an open source project than if you try to read the website of a group responsible (from what I can tell) mainly for making things simple and easy. The amr.c file will have some links to the 3gpp.org website for some zip files, and it’ll tell you where to put them. I built with the amr_nb-float version. I tried out am_wb also, but that failed to build for me. The download I used was http://www.3gpp.org/ftp/Specs/archive/26_series/26.104/26104-510.zip.
  Create the amr_float directory, download the zip, unzip it, unzip the zip inside the zip, sigh to yourself, and just keep going.
• Standard build and install now: ./configure –enable-amr_nb; make; make install
• At this point you should have an ffmpeg executable, and it should know how to generate 3gp videos! The conversion process definitely isn’t what one would call seamless however. Personally I’ve been using the process described in this post to string together the right sets of commands. It seems to be working well.

ffmpeg, mencoder, and I have crossed paths before. It’s normally not a pretty sight. Sure, I manage to get them to do pretty much what I want. But it’ll take days sometimes of poking around, playing with options, trying suggestions that other random crackpots toss out into discussion forums. And even after that the video and audio get out of sync, the streams pop and hiss, the color levels are off, the volume is either deafening or too soft. But I have to say that this time around it’s looking pretty good actually. I’ve tried a few different videos and they’ve been converting well. The sync seems consistent throughout even a 40 minute video. I can understand what people are saying in the clips. And besides my visit to the 3gpp website (before I realized that the answers I was looking for laid in the ffmpeg source code) I haven’t even felt like crying yet. Ahhhh, the sweet moosey smell of success.

I’ve been lusting after the E70 for a long long time. Everyone knows, it’s definitely not a secret. I’ve got this thing I’m working on and it just needs a real keyboard on a device capable of getting SMS style messages and the occasional blip of high bandwidth activity. E70 seemed perfect for that, but the US version is still nowhere to be found.

I considered the E61, but it wasn’t as compelling for me. Then Russ went and got one and started spouting off about it. And spouting and spouting, and just literally gushing over how cool it was. So I went over to check it out. And it has a nice browser, and SSH via Putty on the thing rocks, and Agile works, and the calendar kicks some ass, and 2gig MiniSD cards are only $80. The final nail in the coffin, it’s available right now from Amazon. 30% off. Madness!!! Mine arrives on Friday damnit.

Some of the folks long involved with the W3C have been wandering or stomping away as of late. The problem seems to be detachment from the “ordinary designers and developers” on the part of the organization. I’m struggling to figure out what that means for the Mobile Web Initiative. Good or bad?

Being beholden to corporate paymasters might be seen as a good thing by the companies most directly related to making the mobile web available to the general public (carriers and manufacturers). But on the other hand I think mobile reality is already lagging behind mobile vision once again, just like it was 6 years ago. Mobile services aren’t going to be able to promise the world and deliver 10% of it at 56k. We need to actually deliver this time around or we might never get a crack the the consumer public again.

That means that the infrastructure (of which I consider the browser and mobile web standards two of the most key bits) is going to have to show the kind of malleability and rapid evolution that made the fixed web spread like wildfire. Put that way I suppose the W3C never really stood a chance.

The video from the July meeting of Mobile Monday London has been posted in multiple formats. The topic was funding, and included a few presentations about the world of venture capital and what I thought were a fantastic set of presentations. The presentations were given by companies with mobile solutions and products, and critiqued by a panel of VCs. Great setup, looks like it was a fatastic event. I haven’t watched the whole thing yet, it’s more than an hour and a half of video, but I’m working on it! And it kicks ass that they publish in 3gp, I want to do that. I downloaded it to my phone and had it sitting there for the choice moments when I could watch a few seconds.

I can’t get to sleep, and there’s still a while left before Paranoia Agent comes on. So of course the only logical thing to do is hack around with some Wordpress WAP goodness. The plugin worked out quite well to get a basic version of WML content up and going quickly. I’ll have to play around with it some.

I’ve been thinking a bit about making blogs mobile accessable. The plugin is great as long as someone is running Wordpress. But another level to aim at is the RSS/Atom output. Phonifier takes RSS and Atom feeds and outputs them in very simple XHTML for use on phones. It would be nice to have the same kind of thing with WML output instead. Maybe I’ll add that into my transcoder.

This is self serving of course. We would love to make mobile versions of blog content so that we can get AdMob ads on there. So of course there has to be an easy way for the owner of the blog to set that up. If they install their own plugin they can just copy the normal publisher code in there. But if we go with some kinda transcoder style tool I assume we should do something like run it at AdMob for the publishers. Like we provide the Mobpages for advertisers. Just thinking out loud here, I’ll have to play around some. Ping me if you’re interested or have any particularly strong feelings one way or the other about potential solutions.

Charlie is exactly right about everyone needing to check out the talk given by Tom Hume at d.Construct 2005. Fantastic stuff in there. I liked this one:

There’s lots of technical similarities between mobile and web. There’s lots of places where it overlaps. Lots of mobile services use basically a web server with pages and files sitting out there. They use similar sorts of markup languages, maybe XHTML. Java’s been moved quite successfully from - well Java’s moved all over the place. But from applets onto mobile phones. But this is just a tiny slice of the pie. If you’re building any sort of mobile service, or in fact any sort of web service, the time you spend writing that markup lanaguage compared to the time you spend doing everything else is absolutely minimal. When you think about what’s involved in building a modern website these days. You’ve got the planning side of it, doing interaction design, information architecture, database design for the backend, middleware, integrating all that stuff together, testing it across different browsers, installing documentation. The actual bit of that where you’re writing the markup langauge is very very tiny. So the fact that you’re having to use different markup lanaguages in different places doesn’t actually matter all that much. It’s more the thinking behind the application which matters.

Tom makes some great points, and helps to emerge some of the things I believe a lot of us find frusterating about the current state of mobile. There’s so much potential there, but what we keep seeing are folks trying to jam a web experience into the mobile environment. And of course everyone involved in the experience ends up frusterated. We haven’t seen too many instances of “mobile applications”, for the most part just applications which can be interacted with from a handset.

However I also have to consider the Sapir-Whorf hypothesis, and if we should be inverting the way we think about this. If the current crop of technologies yeild crap applications because developers thinking in terms of markup languages generate complex and cumbersome interfaces does that mean we need a different way to specify mobile markup? Is the move to unify the fixed and mobile web on a single markup style hurting the mobile web instead of helping it? Is it making folks mistakenly classify their website as “mobile” because their markup might pass a W3C validator - focusing attention on the wrong areas?

I don’t think so, I would love to see a unified web. But the talk did make me think and question some things I though I had already thought through completely. Which for me means it’s an excellent bit of commentary. Great work Tom!

One of the continued nasty bits of running Linux on a laptop has been support for WiFi networks. The tools tend to be very capable when they exist, but support for hardware has been spotty (thanks mostly to the hardware manufactuers unfortunately) and the user facing tools have been unpolished at best. Just when things started to get better the security arms escalation required switching to WPA. There just hasn’t been support for the new encryption style at all in user facing tools.

I’m a Linux diehard however, so I do whatever I need to get WPA suport up and going on my laptops. Usually that means grabbing some tools, compiling them, sometimes the occasional bit of kernel hackery. Not bad for me, but really an unacceptable ordeal for lots of users. So I was really happy to see wpasupplicant show up in the Ubuntu Dapper repositories. It still requires editing config files instead of just using some nice squishy interface, and wpa_supplicant itself still looses sync for me every once in a while - requiring a manual restart of the interface. But at least it’s progress in the right direction. And it’s pretty well integrated with the system scripts, so it doesn’t feel like a nasty hack any more when bringing up an interface. The Linux based 770 has supported WPA since I first picked it up. I’m hoping that eventually we’ll see integration like that on general PC hardware.

I would recommend checking out the newest Ubuntu release if you’ve got a relatively high geek quotient but you’ve been staying away because of Wifi support. If the progress keeps up I think the
switching trend that some others have noted is going to keep rolling.

The description that Robin Good has of mobile messaging and swarming (hat tip to SmartMobs) really struck a chord with me. Particularly the points outlined as challenges (routing, interop, open APIs, multiple communications styles). I remember when I first picked up Howard’s book SmartMobs that I felt like we were right at the cusp of a tipping point toward massive decentralization. Blogging was just becomming popular as well, lending to this feeling of peer to peer interaction taking over.

However here it is, years later, and we’re still talking about swarming. But very few people are actually doing it. There’s a seemingly endless number of applications that could be built on top of a generic messaging infrastructure that spans both the wired internet and mobile networks, but that system is still a while in coming. I like C. Enrique Ortiz’s comments on a unified messaging infrastructure, and agree with a lot of what’s in there. Unfortunately that includes the timeline for IM to overtake SMS, which he references John Delaney saying would take 5 years.

One of the things I’m hoping to at least touch on at the July SV Mobile Monday is how we can pull that in somewhat. How can we build these swarming style applications today? With the infrastructure available and without having them cost enough to outweight any but the most niche of benefits? I’m personally not sure it’s possible yet. The ideas are fantastic sometimes, but the existing infrastructure doesn’t support proper development of an efficient solution. The existing usages tend to be interesting because they work in spite of the huge inconvenience their participants have to endure (instead of working because of the support of a complimentary application or service).

I know I’ve pounded my head against issues in mobile messaging a number of times before, and always come away frusterated and disheartened. I’ve heard stories from others that they have as well. Is now the time to build a mobile messaging app? Or would you end up just lining carrier pockets while showing them a potential market they may have overlooked? Would it be best to wait till time and market forces drag down the wall on their own and messaging can be handled like IM? Or can a well constructed play capitalize on the current situation while dragging into a new shape? I’m personally tending toward point solutions built on top of existing IM infrastructure when playing around with side projects. One of the most interesting aspects? Google tying Talk in with GMail means that an XMPP based messaging tool has a tremendous potential audience that can use the service right from an interface they’re already using with an account they already have set up.

THe official release of the new OS for Nokia 770 devices is out. I’ve only been playing with it for a few hours so far, but I have to admit I’m not as blown away as I was hoping to be. There’s a bunch of really interesting great stuff in the new release. The Google talk/messenger integration is fantastic, virtual memory is supported now, the package manger runs against live online repositories. All fantastic stuff!

However there were a bunch of stability problems with OS2006 that I was hoping to see cleared up by official release. In the fooling around I’ve done so far it doesn’t look like they have been. The browser in particular still causes a lot of problems. I’ve had it crash a few times on me while viewing the “tableteer” content that seems to have been created specifically for the 770 device. The package manager was working a bit flakey for me. It had cached a somehow corrupted version of the openssh packages and I had to blow them away via the command line and redownload them to get them working. I even had the device as a whole spontaneously reboot on me when I tried to pull up the streaming audio tool (which has worked since, that seems to have been just a one time thing).

There’s definitely great stuff in there. Fantastic stuff! But it needs some work on the stability and consistency end. I would love to see a point release or two in the next month aimed at heading off these stability issues.