Nelz's Blog

Mah blogginess

MobileCampSF Recap


Today, I attended the first San Francisco incarnation of the Mobile Camps un-conferences, called MobileCampSF. These "un-conferences" are built upon the BarCamp model. I attended my first BarCamp (BarCampBlock) event in August, which I blogged about here, here, and here.

"The Mobile Web"

Today, I learned tons about "the mobile web", which is a term for accessing internet information via mobile devices, such as web-enabled phones and PDA’s.

I come from a Java web development frame of reference. In my line of work, browser incompatibilities are a constant headache, and cause lots of frustrations. Luckily, the major browsers are trending towards standards-compliance in things like JavaScript and CSS support. And where the browsers don’t meet up, there has been a lot of work in common libraries that abstract out the differences.

But not so in mobile devices. Do you think the tens of browsers and the 3 or 4 Operating Systems we have to deal with on the "desktop web" are a pain? Try dealing with hundreds of browsers and platforms, each with their own quirks, abilities, and default renderings! That is a concept that could keep me up at night!

Plus, consider how little space there is on a mobile device. A three-column layout is completely illegible on such a small screen. Add on some data-size concerns (since carriers are still charging by the kilobyte), and a ‘normal’ page becomes untenable for many users. This leaves us with two choices, either create pages specifically with mobile-targeted content, or use a "transcoding" technology which chunks up your full page into (mostly correct) bite-sized pieces.

Now you may ask, like I did, "What about emulators?" Sure, there are emulators, but they are not always guaranteed to be up-to-date with the handsets themselves. But, emulator versioning is not even the tip of the iceberg, as the carriers (a.k.a. the cellular providers) may "take liberties" with the actual messages being passed, either by making undocumented modifications to the content or placing undocumented size limits certain types of data.

My "BarCamp Experience"

Compared to BarCampBlock which had around 900 attendees, this was a much smaller un-conference with an attendance capped at 150. It was held at the beautiful Swedish American Hall in the Castro District of San Francisco.

Since this event was smaller, it had some challenges filling up its session list. Until around 10:30, there were only enough sessions for one full "track" during the day. The number of sessions did pick up later in the day, but I did overhear some of the coordinators wondering how they could have reduced the attrition rate from those who had originally RSVP’d.

Another aspect of the smaller size was that the session start and ending times were a bit more fluid. (At BarCampBlock, if you session went over, you knew it by the line of people forming at your door.) This wasn’t a horrible thing, as the conversations were allowed to end naturally.

In regards to my own participation, I took on two self-appointed tasks: maintaining the session list on the wiki page, and organizing a discussion on security and privacy in the times of the mobile web.

One last contrast to BarCampBlock: sponsorship. For BarCampBlock, the organizers tried to cap monetary donations at $300, with an open accounting. Today’s MobileCamp had no apparent expense disclosure and only one single sponsor in Nokia. Nokia even provided the organizers with several N95’s and a couple of N800’s to give away as prizes. (Disclaimer: I was one of the lucky few to walk away with an N95. Thank you, Nokia!)

I was able to talk to Alexis Rondeau for a little bit about Nokia’s sponsorship. He said that Nokia had been a great great sponsor, and never once did they try to force any kind of marketing upon the coordinators. Alexis also said that the coordinators had to ask for the prizes to give away, and Nokia happily obliged, but at no time did Nokia try to force their hand at doling out the merch. It seems to me that Nokia is taking a very healthy and organic approach to supporting this community.

The Sessions


From the MegaPhone website:

  • "MegaPhone is a phone-controlled, real-time, multi-player collaborative gaming platform for big screens in public spaces."

This session was presented by Dan Albritton, who is one of the co-founders of MegaPhone. MegaPhone was originally a graduate student project by the other co-founder (who’s name I don’t remember), and then turned into a startup after graduation.

Basically, if you have a large display you can hook it up to a MegaPhone server which will allow passersby to call in and control games via their cell phone, either via keypad presses or waveform analysis (not speech recognition).

The big marketing potential for this is that the companies who are showing these big screen interactive games can then harvest the information to send SMS or voicemail promotional materials. Also, this concept is much more engaging than some passive display, as it effectively "closes the loop" between the advertiser and the consumers.

The demo was quite fun. Dan showed two different games, one where you could control a "ship" via keypad that shot the other ships on the screen,, and the other game was a voice-activated version of Hungry-Hungry Hippo’s.

My biggest concern over all this is from a personal security and privacy aspect. Dan said that he himself is a pretty staunch privacy advocate, and so far he’s worked with pretty level businesses, but I get creeped out considering the potential for less-than-honest businesses to abuse this information.

A Discussion of Privacy & Security in the Mobile Era

This is the session that I assembled. We had only 4 participants for a while, and a couple more trickled in toward the end. I didn’t have content per se to present, but I was hoping to generate some good discussion around the privacy and security concepts. I was happy that the conversation was very free-flowing and meandered over many different topics.

David Harper really liked this idea, and in fact he said he had tried to start similar conversations at the previous MobileCamps.

David started the conversation by proposing some kind of "Consumer Bill of Data Rights", with some sort of body that would try to certify adherence to these policies. We all agreed that something like that would be nice, but the cynic in me has a hard time because there are so many holes to poke at a system like that.

One of the ideas that was really new to me was the concept of maintaining both an "A-phone" and a "B-phone". Evidently some Hollywood-types one phone whose number is known by fewer high-importance people (the "A-phone") and one phone (the "B-phone") whose number is known by a wider audience. After the A-phone number gets too wide of a distribution, the A-phone number becomes the B-phone number, the former B-phone number gets retired, and the person starts using a brand new A-phone number. Evidently, this A-phone/B-phone thing is so prevalent, that some GSM carriers (in some markets) are offering phones with two SIM card slots.

This rotating A-phone/B-phone brings about an interesting concept: transitory obscurity. Yes, we all know that in a computing paradigm "security through obscurity" is ineffective. However, since information flows quite a bit slower in a social paradigm, security through obscurity can be an effective tool, as long as we realize the tendency for information to become less obscure (and therefor less secure over time). This is where transitory obscurity shows its strength, because if you are willing to manage the transitory access to your information, you can effectively achieve some semblance of "zeroing out" your information availability.

This is a change in what I see as our society’s previous behavior. Phone numbers used to stay with us for years, and due to phone number portability we can still maintain this illusion. Personally I am proud that I have been able to keep my main email address over the last four years, even with my entrance to the public internet. But, according to David Harper, his own kids are embracing transitory obscurity. Once their profiles on whatever social networking site become too connected or too visible, they just abandon it and start over.

(This also brings up the disjunction between reality and marketing speak. Social networking site marketers love to tout their "member count", but how many of will provide (and define) a count of "active members"? None that I know of…)

There are other challenges on the mobile web in regards to security. Some carriers leak per-user information when the user accesses off-portal websites. (Supposedly Nextel provides the unique subscriber id in a header.) And, some carriers obfuscate other user information that is usually available on the web, such as IP addresses. Leaking subscriber information could help provide customized experiences but could also be a security breach. Obscuring IP addresses may be a boon for privacy, but it makes it really difficult to filter certain users from online discussions.

Another problem in the mobile web: how do mobile web providers disclose their privacy practices? The desktop web is bad enough with legal-ese privacy statements that no one ever reads, how should that shake out on the mobile web where real estate and data resources are highly throttled?

As for a tangential learning about the mobile web, it was said that "basic and useable" are much more valued rather than "rich experiences".

Introduction to Mobile Processing

Francis Li presented his open-sourced J2ME platform called Mobile Processing. It’s a pretty cool project that provides some basic UI and connectivity functions in an abstraction layer to reduce complexity for developing end-user mobile Java apps.

Mobile Processing is a pretty cool project, and now that I have a Nokia N95, maybe I’ll play around with writing some basic apps.

Hyperlink Your World!

Alexis Rondeau presented his project web site which uses "QR-Code" barcodes to link physical locations to their respective Wikipedia entries.

A QR Code is a short bit of text encoded into an open standard, two-dimensional bar code that is easily readable (and actionable) by many cell phones. You can encode alphanumeric text, binary, or even Kanji into a QR Code, and is currently used throughout Asia as a way to encode URLs for instant access on internet connected mobile devices. One of the interesting characteristics about QR Codes is their dimensions change as you encode more data into them, so a densely populated QR Code will be bigger and contain more pixels.

The Semapedia project adopted the use of QR Code because it is an open format, though there is at least one other less-well-known open format two-dimensional barcode, and several proprietary ones.

Ability for cell phones to read QR Codes is well documented and heavily used in Asian markets, though carriers in the U.S. have been slow to adopt these capabilities.

The New Walled Garden - Header Crippling Carriers & Transcoding Services

This session really brought home the frustrations of developing content for the mobile web. For years, carriers (cell phone companies) locked mobile devices into using only the information provided on the carrier’s own platform. It has been a rather recent development that these carriers have allowed mobile users to access "off-portal" (open web) content.

This was a very good development for the end users. But, in the recent years there have been some carriers (most notably Vodafone networks in Europe) that have been stripping HTTP requests of some of their basic information such as the User-Agent headers. This robs providers the ability to present content designed specifically to match the end user’s device capabilities.

(Comcast recently tried tampering with end user’s information flow similarly to this, which resulted in all sorts of outrage and legal action. The actions of these cell phone carriers gotten much media attention yet, but it would benefit end users to know that these types of information tampering are still prevalent.)

Without the User-Agent headers, some sites (example: Wall Street Journal) just default to presenting content that was optimized for desktop usage. This content is much more dense than mobile-targeted content, so the carrier then applies their own transcoding technology to the content. In several instances the transcoding breaks the content into little pieces, and redirects the end user back to the carrier’s own portal to access further pieces of the originating site. This draws the end user away from the original source content, and artificially inflates the number of ad impressions on the carrier’s own portal.

The Wonderful World of MAEMO

I only caught the last bit of this presentation, but it was arguably the coolest part. Nokia had lent the presenters two demo versions of the upcoming Nokia N810. It was very nice. The slide-out QWERTY keyboard was a nice upgrade from the N800 (of which there were a couple given out as prizes), and there was much talk about the increased "solid" feel from the N800. One of the presenters brought it all home by saying the best way to showcase the tablet to a techie is to say "It’s a Linux box." (Yay!)