One of the difficulties [for Palm developers] is that there's no one killer program Palm owners want to go into a store to buy. In fact, aside from the built-in programs for the Palm, no application is on more than 10 percent of handhelds.

-- CNet article, "Can developers pull dollars from Palms?", Feb. 8, 2002

The Nokia Communicator. The Handspring Treo. The Samsung i300. Combined with about a thousand other mobile phones offering "Wireless Web" access in various shapes and forms, it's clear that the Internet is finally going mobile. The bandwidth is a bit low yet, but that should be changing as it becomes more popular. But the problem is, it's not all that popular. People everywhere are using mobile phones for SMS -- the small-screen equivalent of e-mail -- but the Internet and all its content is getting ignored.

Not for lack of services, mind you. Yahoo! is entirely handheld-friendly, and people still aren't using it in droves, despite the availability of things like weather, stocks, sports scores, movie listings, the latest news and so forth. Part of this is because the screens are so darned tiny, but I don't think that would stop people if they wanted to read badly enough. No, the real problem is that Yahoo! doesn't offer any compelling content online that a mobile phone user couldn't get just as easily by picking up a local newspaper -- either today's or tomorrow's. Sure, I can get sports scores for cities halfway around the globe, but why would people want to? I have Yahoo! on AvantGo on my handheld, but the only thing I ever use it for is to look up movie listings occasionally -- and that's because I'm just too lazy to find the newspaper.

Despite all the hype about the wonders of the wireless Web, the available mobile devices are not able to provide a wireless Web adventure that would impress anybody. Transmission speeds lag behind the dial-up access rates the average AOL user gets at home. And the big question remains, who needs it?

-- NewsFactor.com article, "Where Is the Wireless Web?," Feb. 13, 2002

So portal-type content isn't really driving people to use the Web wirelessly. What would? The kind of content you can't get in a newspaper, obviously -- niche reviews, obscure factoids, interactive discussions. These things are available on the Web, but most of those sites aren't designed for small screens. Browsers for those devices, like Handspring's Blazer browser, try to optimize 800x600-resolution Web pages for 160x160 screens, but they can only do so much. Try surfing the Web using the text-only Lynx browser and you'll see why: complex layouts using tables and JavaScript-dependent features cripple browsers that provide anything less.

The Wireless Web should be the mobile device's "killer app," but it's not. All the truly compelling sites are slow and badly formatted, and all the fast and well-formatted sites lack compelling content. Most sites try to offer as much information and navigation as possible in one page, filling up the left or right columns (or both) of every page with the same stuff and leaving the main page available for specific content. Everything2 does this, too. E2 is generally easy to use with Lynx, but all those wonderful nodelets still sit at the bottom of the page while the Search stays stubbornly at the top, forcing users of non-graphical browsers to scroll and squint to identify either one.

In many of the more relaxed civilizations on the Outer Eastern Rim of the Galaxy, the Hitchhiker's Guide has already supplanted the great Encyclopedia Galactica as the standard repository of all knowledge and wisdom, for though it has many omissions and contains much that is apocryphal, or at least wildly inaccurate, it scores over the older, more pedestrian work in two important respects. First, it is slightly cheaper; and second, it has the words DON'T PANIC inscribed in large friendly letters on its cover.

-- "The Hitchhiker's Guide to the Galaxy" by Douglas Adams

E2 is never going to be popular with mobile users in its current state. But it should be. After all, E2 offers all the things the portal sites don't, and all in one place. Movie and book reviews, by people who aren't paid to hype them. Biographical information on celebrities and historical figures. Popular culture. Definitions. Textbook knowledge. Famous poetry and prose, both original and not. Humor. Blogging. Eclectica.

H2G2 tries to be what Douglas Adams' Hitchhiker's Guide wanted to be, but fails for a number of very good reasons. Plus, E2 offers a community spirit and a very open system for adding and editing content. At the time of this writing, E2 has nearly half a million distinct writeups by some 40,000 distinct users. And none of it is available in any other, single location.

We are the killer app. The world just doesn't know it yet.

I am pleased and proud to announce E2's new better XML output. This increases our current set of XML ticker functionality, plus provides a better and cleaner set of XML (well formed, but not yet validating) displays for this site.

clientdev: new XML ticker output, by JayBonci

I would like to put forth a vision for all the developers who use and enjoy E2: Build a wireless client to the full database that is small, fast, and easy to use. And, of course, freely available to anyone who wants to use it. This would be the killer app for wireless handhelds throughout the English-speaking world.

Imagine, for a second, being able to look up honest movie, music or book reviews right in the store before you buy. Or accessing a travel guide to New York City while you're waiting for the subway. Or getting some fun information on your favorite sports team's opponent rather than just statistics from last season. Or hearing a strange piece of slang while visiting a distant city, or a technical term while reading a magazine article, and being able to look it right up. Or just wanting something random to read until your class starts or your meeting ends without having to download it first. Imagine how much the rest of the world would like to be able to do all that, too.

But E2 isn't good for wireless use yet, because it requires a Web browser and because it's too friggin' slow. I don't mean server loads. I mean the nodes that contain too many writeups with too much text for a slow mobile modem to handle, the search tools that are too far away for a small screen, the soft links and nodelets that provide quick navigation for a user of a graphical browser, but just slow things down needlessly for those who don't.

E2 has all the content it needs to be a wireless hit. It just needs an optimized interface.


Things To Do

These are the major issues I think a wireless, mobile version of E2 would need. Many of them are things that would or should never be done on the graphical, PC-based version of E2. Nevertheless, a different environment requires certain changes.

  • Forget the Wireless Web. A slimmed-down version of E2 for portable browsers is not the answer. Certain elements, like the Search bar, need to be on-screen at all times or at most a single click away. A user who has his User Preferences set up one way for the graphical E2 might want them a completely different way for the mobile E2. What we need is an entirely new client that accesses E2 using XML, and then displays it in a interface designed for small screens.
  • Trim the Extras. Nodelets must go. Softlinks must go. Useful features like adding a writeup or gabbing in the Chatterbox should be available by selecting a menu or clicking a fixed, non-scrolling button. Bookmarks of favorite writeups should be tucked away in their own menu. The text display of the client should display nothing but writeup headers and text, using the same HTML formatting that we support now.
  • Optimize the Writeup Display. E2's current node layout is great for the WWW, bad for the Wireless Web. Getting several writeups at once would be prohibitively time-consuming on low-bandwidth, mobile connections. Only one writeup should be displayed at a time, with single-button access to the last or next writeup. Writeup titles should be static while writeup headers and content should scroll. Voting up or down should be as easy as pressing or tapping two buttons (one would be prone to accidents). Writeup headers should be kept to the bare minimum -- author, type, date, and reputation.
  • Get It Everywhere. PalmOS. PocketPC. Whatever the Nokia Communicator uses. As long as they're looking for developers and "killer apps" to sell their new wireless devices, we should take them up on it. This client should work on every popular mobile wireless device that's on the market, and it should work as identically on every one of them as possible. Obviously, different screen sizes and resolutions (especially the elongated Communicator screen) will require some major changes to the interface layout, but there's no reason we can't keep the same on-screen tools and menus, just stored in different places. UPDATE: There's a book on the way titled Flash Enabled: Flash Design and Development for Devices (http://www.amazon.com/exec/obidos/ASIN/0735711771) that discusses designing Macromedia Flash for portable devices. I'm convinced that Flash, especially the newly-released Flash MX, will be the way to go. Many smart phones can use it, PocketPC can use it, and I believe it's coming to the PalmOS very soon. With built-in XML support, this will be the fastest way to produce a multi-platform client (if not exactly the most backwards-compatible).
  • Let It Make Money. Yes, by all means, give the client away, especially to device manufacturers. And keep E2 on the WWW free for everyone. But the mobile client is a special "extra," and if it becomes popular, it should offer the opportunity to make money for its developers and servers. If we're going to tax the server extra with this, we should get something back, right? So let the client be a 30-day trial, or think of how to sell $2-a-month subscriptions to access the content with it. Or else develop a whole new way to sell ad space on the thing -- let Barnes and Noble put a blurb beside every writeup containing the word "novel," for instance. I can call anyone cheaply from my home phone, but I pay a privilege to do it on the go; it's fair and reasonable for E2 to expect the same.
  • Involve Your Friends. Yeah, maybe you're whiz-bang enough to develop something like this on your own, but why when you can get help? Join the clientdev group here on E2, find out what you need, stick it up on SourceForge, and do your darnedest to let everyone who owns a mobile device know what it is and what it does. They don't have to know or like E2. They don't even have to see the word "Everything2" anywhere on it. They just have to use it and love it. And if they offer suggestions, write 'em down and try to add 'em in. Again, just because E2 on the WWW doesn't do something a certain way doesn't mean a wireless client can't.

Who We'll Need

  • Anyone familiar with the development tools for existing mobile devices, especially PalmOS, and/or anyone willing to learn how they work.
  • Anyone familiar with XML and how to build and parse it, and/or anyone willing to learn.
  • Edev members to build a few new doctypes that the mobile client can access. Existing doctypes are designed for the graphical WWW; we need new ones just for mobile clients. (Example: a new type of node document that contains only one writeup and a link to previous/next ones.)
  • E2 gods who can make the above things happen on the server side.

Interested? Join clientdev and tell us what you can do.

"Forget the Wireless Web."

You can, but I sure won't. In fact, I won't forget about so much that I started writing a WML client about a week ago. :)

I agree that, to get the full E2 experience, you need more then a phone can offer. But that if you don't want the full experience, just pop in to check a definition, node rep or private messages? I think it fits the bill perfectly.

It also allows you flex your mind creatively because you have to deal with issues no one ever things about on a regular web system (like how do you divide a writeup into chunks but make sure the end out one chunk doesn't fall in the middle of a hyperlink?).

It won't satisfy all, but I think it's not a good idea to rule it out completely.

It's interesting how the flow of E2 actually translates into a small device, too. The very nature of the nested design makes traversing E2Nodes easy and even bearable. The only issue really is one of speed since WAP usually operates over a low speed data network, but since it's just text, it's doable.

This is not an interface to write nodes, obviously, but I've had great success in reading nodes, homenodes, searches and whatnot. Private messages aren't added yet, but are on the list; after the debut of message/email gateway, I got numerous requests from people (though many wanted SMS, which is difficult because you have to pay per message). A chatterbox client, while balky, would be completely feasible.

As for making money, If I wasn't a perl dude who could write it myself and I was told I could browse E2 in even a limited fashion anywhere in the freaking planet... well, I wouldn't pay for that, but I would pay, say, $2 a month to be able to edit nodes (on a telephone keypad? What was I thinking?), monitor the cheddarbox and maybe even chat on #e.

Another thing this has done is point out what I think a re some issues with the XML output, though I'm sure they're only a minor bit of code to deal with.

It currently up and running at http://ajt0.ws/e2/ (A-Jay-Tee-Zero) and is open to all who wish to use and abuse it. It is alpha-extreme, so don't be surprised about 500 errors or WML Compile problems; Source will be gladly given up in a day or two as soon as I make it non-embarrassing.

NOTE: The WAP Proxy has been officially announced to the ungrateful, unwashed and unworthy masses.

I agree with Xunker. (Hmm. That's a bit short, better add some more.)

N.B. This is currently a work in progress.

I too think that mblase is too quick to dismiss the 'Wireless Web', making the same mistake that a lot of people did when WAP first came out a couple of years ago1: The point of net access on a mobile device isn't to replicate the experience of browsing the internet on a fully multimedia-capable PC, it's for dipping in for a few seconds or minutes to retrieve a specific piece of information. Good examples include services that offer directions to good pubs and restaurants, and a very popular service in the UK that evaluates used cars.

We should not be chopping up a web application (E2) to fit in a mobile context, we should be designing for the machine. In Europe2 nearly all phones sold today have WML browsing capability (yes, even with images after a fashion.) over WAP or increasingly GPRS. (WML, in a nutshell, is a rather strictly defined XML subset that can be converted into bytecodes to speed things up on 9,600bps GSM circuit connections. It is fairly straightforward to translate to and from HTML, but rather more efficient to write in it natively.)

Guide Mode

I envision that the WML service of E2 would be like a 'Guide mode'. Card3 one would be your search field (or a mode selection menu, if you wanted to choose between different E2 features- e.g. the catbox, but as you could just bookmark the feature or features you intend to use, I'll ignore that for now). Enter your search term and click search. Now E2 (or rather the WML-XML converter doohickey) returns one or more writeups for the node you entered. Here's the clever part4: the savvy noder has previously set up their E2WML settings via the more traditional web interface. (No, this isn't cheating. Think of it as 'docking'.) So now they get a result based on their preferences. (Some examples: return highest rep WU first, return shortest/longest/combination-of-length-and-rep writeup only, prefer writeups by certain noders.) One potential pitfall is that you might have to remove hardlinks and pipelinks altogether from the output, as they would add clutter and balloon the size of the data (that is already almost certainly having to be cut into chunks). But again, this could be an option. The key idea here is that the web-based preferences setup allows the user to tailor the WML output to their needs and their device. (Another example- if someone nodes a book, they can adapt the contents page for mobile users and you have an instant eBook. Albeit a very chopped up one.)

Cheddar Bocks

The catbox could be handled pretty straightforwardly, links and all. A more useful function would be to be able to quickly send private /msgs to other users. This would be little different from SMS, I'll admit, but it might have advantages. (Cheapness- in certain circumstances- for one.)

Other points:

No-one, and I mean no-one, unless they have a laptop, is going to want to submit nodes while on the move.

We can increase the useability of mobile E2 by trying to accomodate it in our writeups or organisation. (This is one of the reasons I'm so keen on nodespheres.) If we had a way of creating 'hints' for mobile E2's parser5, we could give the user quicker routes to get to the data they are looking for (quicker means as few clicks as possible- scrolling is pretty much 'free'. So a massive combobox list is favourable to a text entry field in many cases.) We shouldn't be afraid of putting things on a seperate page (as viewed by the mobile client) if it introduces a speedup and reduces clutter. The web-based noder need not be exposed to any of this, of course.

Finally, learn about the technology. mblase's early versions of the above writeup sparked a tirade of /msgs from me, because it was slanted entirely towards PDAs (expensive and ultimately useless toys at present, IMO) - because personal experience skewed his worldview of what devices are out there and what they're really capable of. (Remember that the press- even the tech press- only have a small amount of experience with this technology, and therefore will have only seen the most widely-known websites, as opposed to the l33t high concept projects- such as multiplayer Elite on WAP). Go deeper.


1. Due in no small part to ill-conceived hype by telecoms companies, e.g. the great satan.

2. I can't comment for the Merkins, because last time I checked mobile technology over there was still in the 'crank-operated field telephone' stage. Or at least, (shudder) analogue. But look at it this way, make E2 WML-accessible now and you can use it when that technology does become commonplace over there. Get the jump on your isolationist information-provision rivals!

3. WML 'pages' consist of 'cards', which are like different stages in a HTML form. Prevents you having to scroll through loads of controls you would rather ignore.

4. Clever may mean stupid. No warranty is implied.

5. Of course ideally the parser will have enough intelligence to make many optimisations automatically....

Log in or registerto write something here or to contact authors.