Mike Rowehl: This is Mobility

One of the questions I keep hearing over and over again is “Are better devices and more capable browsers going to lead to the death of the mobile web? Are we going to stop thinking about the mobile web as a separate medium and just do web development?” Eventually that might be the case, but a number of things would have to happen first. So instead of trying to answer that emotionally charged question with a relatively meaningless yes or no, let me lay down a few factors that hopefully indicate why people are asking it so frequently now.

The impetus of course has been the iPhone, and the ability of the browser on that device to render any web page out there pretty much and present a fantastic version to the user. Does that mean that there won’t be a mobile web any more however? That depends on what you define as mobile web, which is ultimately a somewhat boring question when compared to how to make pleasing applications for mobile phones. What the iPhone is calling into light is that one of the conditions that brought about the split between authoring for traditional browsers and authoring for mobile devices is starting to fall away.

The split between mobile web development and traditional web development was supposedly based on two factors : technology and user experience. The technology aspect was that the networks were much slower than wired networks at the time and the devices had less ram and slower processors. What that led to was using different markup for the pages themselves. Something that was stricter in the computer science sense and easier to interpret on the device and validate in the network. Something that dropped out a lot of features and functions normally found on the web as it was. The user experience end was about presenting the information in a format that was more suitable to viewing on a smaller screen and interacting with one handed while on the go.

The iPhone proves that with current technology and on relatively recent networks the technology argument for mobile web content as something distinct from traditional web content is dead. Leave aside all the questions about Apple’s market strategy, pricing, carrier reactions, etc. Stripped down to just a technology question the iPhone has demonstrated that a handheld device can consume and present existing web content in a way that most users find at least acceptable. Done.

However does that mean that this thing we think of as the mobile web goes away? I don’t think so, I think it just shifts purpose. Take an example right from the iPhone itself. The iPhone renders the front page of the Digg news site no problem, but still there’s an iPhone optimized version of Digg available just the same. This is the mobile ecosystem working just as it should. Now that the question of what is technically possible has fallen away, the new issues becomes what is most useful and pleasing to the end user. In the case of Digg they apparently saw a way to put together the content they had in a way that better served the group of mobile users who are on iPhones.

I think that’s going to be the general trend. And it’s a good thing, it means mobile is moving up Maslow’s pyramid. We’re out of just simple survival and life functions and starting to think about how to make things more useful and pleasant. I think in the early days we had only two kinds of apps: mobile and not mobile. With things starting to blend between the two environments I’m seeing things starting to break down into four major areas (with admittedly very fuzzy boundaries):

• Mobile specific applications
• Mobile enhanced applications
• Mobile aware applications
• Mobile unaware applications

They’re actually overlapping group of apps, looking like this if we ignore the number of apps in different groups and think just about their relationships:

The mobile specific applications are the ones that really only make sense when they have their mobile interface as the primary driver. They’re based on being primarily realtime and rely on some bit of info that you can only get at easily while mobile, such as current location. Take MizPee for example. Sure, you might be paranoid and check the location of public toilets before every outing. But most likely you’re going to use it when you need it and you’re in an unfamiliar area.

Mobile enhanced applications are ones that have taken their service and transformed interaction to suit usage from mobile devices better or have included some features mainly focused on mobile users. That can either be something like the mobile version of Yelp, which has a mobile version that changes the amount of info and presentation drastically from the desktop version. Or an application like Jaiku, which pretty evenly splits the difference between mobile app and desktop app in my opinion (even though I think everyone though of it as a mobile app at first, I think it’s found a lot of utility even among the wired only crowd). It has a traditional website, mobile specific pages, SMS integration, and IM capabilities.

Mobile aware applications are ones that have an ability to tune what they return to a browser to make it more appealing but don’t really make any changes to the content or add features. A Wordpress blog running a plugin that provides a mobile version, like the one that Andy Moore has been releasing I would consider in this category. Lots of the folks who are using it are concentrated on the traditional web, but wanted to have a version available that would work well on mobiles.

Finally there are the mobile unaware applications, which are just what they seem, do nothing in reaction to a mobile device. That doesn’t mean they don’t work on mobile handsets, just that they only work on mobile handsets that are able to render their pages and present them.

I think there’s a pretty strong resemblance between the levels of support for mobile that we see now and the evolution of the web itself. There were web specific plays like Amazon and Google, web enhanced companies like Dell and IBM who started going direct to consumer while keeping their traditional distribution channels, and simply web aware companies that put up their brochureware site so that they would have something to point to for the other members of the web ecosystem. I think the same thing is happening in mobile. And like what happened on the web the most value exists in the system when the set of sites out there that are mobile aware completely overlaps with all available sites.

Over time if the “mobile specific” things get baked right into the standards that make up the wired web, and the platforms all support all the different mechanisms you might want to use to tweek your site for use from a mobile (or do it themselves) then yes, we stop talking about the mobile web as a separate thing. But I don’t see it disappearing completely even over the long term. And I think the decisions are going to remain explicit, and reaching higher levels of capability is pretty costly. With Mowser we’re trying to make it as dead simple as possible for folks to at least hit that mobile aware stage and start participating in the growth that’s happening in web usage from mobile devices. If the mobile web is going to evolve away from carrier control we need as many participants as possible to balance out the distribution of power.

I was taking a look at a few things this morning along the lines of mobile browser support for extensions and customization. Following along the theme of Russ’s post about using Mowser as a search provider in Opera Mini I just put up a “View in Mowser” bookmarklet for Opera Mini. Thanks again to Dennis at WAPReview for having raised bookmarklet support in Mini to my attention. You can find a collection of Opera Mini friendly bookmarklets at o.yeswap.com.

I have a shortcut setup for this on my install of Mini now. Once you create the bookmarklet and have it working, go into the bookmark manager and select the “Speed Dial” entry. There you can set the bindings for shortcut keys. I have the View in Mowser bookmarklet bound to *1, so I don’t even have to bring up the bookmarks screen to use it.

Opera Mini rocks, I’m surprised to see that the Nokia open source browser still hasn’t grown support for bookmarklets. One other place where it does work however: the Mozilla based browser on the N810. Very cool!

If you take a look at mobile websites on the whole you see a structure very much unlike websites meant for desktop browsers. Desktop websites do a lot of linking to each other, they feel very much like, well, a web. Thus the name. Mobile sites however tend not to cross link very much. In web terms they look very much like closed silos. Much of the recent movement on the desktop web has been away from silos of information and into services that allow users and other developers to make use of the applications you’ve provided in a novel way. We’re not yet seeing this on the mobile web however. Mobile websites are using the technologies of the wired web to deliver applications, but effectively ignoring or contradicting everything else about the web that has made it work long term. I think taking some of those learnings from the desktop web and applying them to mobile is necessary for the progress of the ecosystem and one of the primary reasons I joined Russ at Mowser.

Why is cross linking important behavior? If you take a look at the early discussions about why to use hypertext on the web and the decisions that went into the overall architecture there’s a lot of mentions that “anyone can link to anything else without needing permission”. It’s such an ingrained principal at this point that it’s easy to ignore. You don’t need the permission of the owner of the page you’re linking to, you don’t have to worry about whether the users browser will follow the link or not, and you don’t have to get anyone’s permission to use HTML to make the links in your pages. Simple and stupid right? The overall theme is one of lowering the barrier for a publisher to make available new information on the web, and to be able to benefit from the network effects of that information. Every time one publisher puts a new web page online there’s an incremental increase in the value of the web. There’s a new page that others can link to, a new page to be indexed, a new source of information to be mixed up or referenced directly.

If you take a look at the mobile environment however this isn’t the norm from top to bottom. As Tim O’Reilly mentions in his article about openness in the carrier environment at the NY Times things are starting to get better. But it’s still true that there are carriers out there who are not open. Effectively, yes, you DO need their permission to link out from your website to others or to allow others to link to you. Or they deliver different information to different websites depending on their trust of the publisher. It’s directly in contradiction to one of the base first principles of the web.

But even if we get around those issues and the environment across all the carriers and operators becomes sane there’s still a problem with the publisher effort required. The question shifts however to “will this link work across the different phones that my users have”. If someone putting up a new mobile website is able to link only to the other sites out there that already have mobile versions the value of the site that they can create is much lower than if they were able to link to content at all the existing web sites.

Once the mobile web becomes more webbish the trends that we’re seeing now toward openness in the mobile environment become even harder to control or reverse. One of the big chunks of necessary infrastructure was already cracked with AdMob and AdSense for Mobile offering direct monetization for a mobile audience. But there’s still a discovery and distribution problem in mobile. Evolving the ecosystem requires lighting up more mobile content and encouraging content providers to cross-promote and focus on trying to serve their users (rather than just selling them something). That’s why we’ve made it very easy to use Mowser to create mobile friendly links between pages and to make a mobile version of your site. You don’t need to sign up, you keep all the advertising revenue from your pages, and you can link up whatever you want and we’ll make sure everything works.

I updated the Mowser WordPress Mobile plugin that Russ posted a while ago and added in an admin page where you can configure your AdMob site ID without having to go in and poke at the code. I was considering going back and updating the Wordpress Mobile with Style version that I had as well, but there are some other mobile WordPress plugins floating around already. I’m going to have to poke around and see what the state of the others is before I go and hack mine up some more.