For me there is people who believe in the device independent Web and those who believe in the device dependent Web.

Mobile Web, .mobi, that’s all “device dependent” as people think they need to target Web pages for the mobile.

I’ve long argued for the “device independent” Web. For example on my Nokia E65, I browse BBC‘s mobile site and BBC‘s normal news page.

I can view either with Nokia’s excellent Webkit based Web browser, though the mobile site is quicker as it is faster to download. However the ‘BBC normal page’ is often far more up to date! Having a ‘device dependent’ Web will create redundancy. Solution: make the BBC News page a little leaner please and just do away with the “Mobile Web site”.

Then there is some people who think you either get:
# Standards geeks who write boring Web sites
# Or people who write for the best possible user experience on the mobile

The 2nd group argues are that most peoples’ Web browsers suck and/or their network sucks. A lot of modern Web browsers do suck. Also network connections in London aren’t fantastic. :/

Though there is the very real future of capable mobile browsers (Nokia’s Web/Opera Mobile) and fast networks (go to Korea or Japan).

People who care about the “user experience” are not in my good books. They are often making dead applications for big companies (e.g. Sony, BMW, FTSE100) who do not need to have to build a page rank. These ‘one off’ applications with great user experiences have no future.

The ‘standards geeks’ who write device independent minimal Web pages will create the most useful sites. Web applications that quickly discern useful information is what I like to call a great “user experience”. No flash device dependent dead crap thank you.


Here at .mobi we like to think about .mobi as a way to honour the user’s mobile context rather than seeing it as a device-dependant web. There is nothing about .mobi sites that prevents them from working on a desktop browser (they are supposed to default to XHTML-MP, which works on all of the major browsers). So, rather than targeting pages to a mobile device, good mobi sites present views of information that make sense to users in a mobile context (which could easily be someone using a PC/laptop). I personally use .mobi (and other mobile sites) on my laptop quite often, specially when I am using my phone as a modem :-)


Comment by Ronan Cremin

Hi Ronan,

Ah, could you say hi to James Pearce for me? We have been through this .mobi context argument before.

XHTML is not supported by IE or AFAIK Google. It’s a DEAD markup format. DEAD like WAP. DEAD BURIED, BYE WAP. GOODBYE XHTML. Hello HTML. :)

I recall the .mobi site for checking markup had a mobile and desktop version. Could you practice what you preach? Show me an example of a .mobi site you use on a desktop WITHOUT DEVICE DEPENDENT DETECTION MODIFICATION (CSS media tests allowed) would be a fantastic start.

Comment by hendry

Both and use thematic consistency to provide subject-equivalent (although style- and navigation-adapated) pages for a given URL. I believe I explained that when you flamed us before.

We’ve been working hard (like the thousands of others who are trying to make this whole medium a reality) to be as inclusive as possible for multiple types of client. It’s a shame you don’t have anything positive to say on the matter.

For dotMobi’s sites, and our particular user type (developers) there’s little point in expending effort on WAP1.1 content for very legacy devices.

But in general terms, I can’t reconcile your blanket maxims “DEAD BURIED, BYE WAP” and “free of device discrimination”. Spare a thought for the billions on the planet who don’t yet own a mobile with an HTML browser.

What do you mean “Google doesn’t support XHTML”? The URL might give you pause for thought. Nevertheless I agree that XHTML‘s future is not assured.

But the mobile web? Inevitable.

(Sadly, so are the challenges of device diversity, at least for the foresable future – but we’re not going to sit back and wait for that problem to go away before we begin the revolution. You’re welcome to, of course.)

Comment by James Pearce

When I mean Google doesn’t support XHTML. I mean for example one of my test sites, which is correctly served as XHTML. It does actually show up now at last on Google search results, however as File Format: Unrecognized.

The mobile revolution starts when people realise we need to support the mobile Web as device independent HTML, a Universal Web. Forget Web page developers targeting screen sizes, forget “contexts”. Then I see vendors seriously getting down to shipping a serious HTML browser.

Fragmenting efforts into WAP, “mobile web” (by which you mean Web designed for mobiles) or mobile “context” slows us down.

For example I know didn’t go with the Nokia Webkit based Web browser by default, because they fealt they had to support WAP. Hence what CRAPPY BACKWARDS DEFAULT&SUPPORTED BROWSER is on their 3G phones? The Nokia “Services” browser. Do you understand now why I have no time for WAP and XHTML?

The more you advertise for special mobile treatment, the more bullshit like this will happen.

Comment by hendry

Kai, I’m also not sure what you mean. Your example is an XHTML page that is correctly indexed and served. It uses a DTD that is, as far as I can tell, an experimental draft XHTML + MathML DTD. The “unrecognized” file format is probably a result of that.

But we’re talking about mobile and, yeah, have you ever seen :) We’ve served XHTML MP and Basic along with cHTML and WML there since early 2005. If you’ve followed the evolution of mobile markup, you won’t be surprised to learn that the amount of XHTML mobile content out there is increasing fastest by far.

To answer one of your questions—I have never seen an XHTML mobile site that doesn’t work on a desktop browser. It’s not particularly suitable for a desktop browser, but it displays fine. does for example. Can you give an example of mobile site that doesn’t render?

I think the Universal Web idea is simply unrealistic right now. I’d turn your question around and show me a single HTML document that provides a compelling desktop experience while working on most mobile devices too.

Comment by Sean Owen

1) Sean, I stand corrected. Though I thought Google didn’t index application/xhtml+xml for some time (ago).

2) I think I’ve used Though I forgot the address. I find easier to remember. I don’t want to remember the mobile versions of Gmail, Google Maps (or any site for that matter) et al. I just want to use the same URL scheme on any device.

XHTML is nonsense anyway, as I know from conducting several tests that the mobile XHTML browsers don’t parse XHTML correctly, in the sense should stop processing when they encounter an error. It’s a bad joke.

You really shouldn’t promote its use!

3) Well I don’t think desktop browsers support WAP. Also for example BBC‘s How do I browse that XHTML site in Firefox? Can’t you see device dependence causes pain & suffering!

4) has been working universally well on any device for several years now. ;)

Comment by hendry

( and other properties will redirect phones to the mobile version, BTW. And returns XHTML (Transitional actually, ugh) to mobile devices. Change your User Agent to see it.)

I take your point about browsers not enforcing valid XHTML docs, that obviously this well-formedness requirement is followed so little that browsers don’t insist on it. I don’t think you’re arguing against the idea that documents should follow a defined standard, so maybe you mean you prefer SGML-based markup like HTML? I don’t know enough to argue that point. But I think you’re really arguing against the idea of a mobile-specific markup standard, and not whether it’s XML-based.

This very site probably nearly validates as MP or Basic, so I don’t think you object to the spec itself since you’re nearly using it. It’s the idea of a separate mobile spec that’s a problem.

This site, a blog, with mostly text, is easily implementable with one document such that it works reasonably on mobile and desktop. I don’t think that’s true for all or even most sites. Try making a compelling Gmail that works everywhere without modification. It does add complexity—how do you know what the user agent is for example?

In the end I think we are merely differing over implementation details. I don’t think you disagree in theory that different user agents may need to present one resource differently. This page’s stylesheet is restricted to “screen” for example. So you have context-specific CSS there already. It’s a question of whether markup should vary. I would also prefer to have one document rather than two where possible, but I don’t think it’s always possible. The difference between the mobile and desktop presentation of a site is not just a bit of CSS trickery—especially considering how limited CSS support is on phones!

I don’t see the big theoretical difference between presenting different markup and presenting the same markup but different CSS, for example. Seems like an implementation detail. I also just don’t see that this “problem” is the big one holding up advancement of the web and mobile web. People don’t even use basic HTML correctly in most cases, let alone elegantly weave together HTML and CSS in the way you say is desirable.

Comment by Sean Owen

You’re joking right? I need to change my UA to see XHTML of This is my example of a mobile site that doesn’t render on desktops. Device sniffing sucks.

Gmail in a decent mobile browser is just as fast as a native Gmail java client. Push for a decent HTML browser based on Presto or WebKit or Gecko for the mobile device and Gmail will work on that mobile device without modification.

The UA modifying the content for optimal device display is GREAT. That’s what I want to see happen.

However Web authors modifying content for mobile display, especially using mobile specific markup is something I am really fed up with. The mobile Web is dead.

There is a big difference between markup and CSS. Different CSS OK, different markup NOT OK. This isn’t an implementation detail. This is how the Web works!

Btw I google stalked myself on Google’s Mobile Web (Beta) search.

One relevant link… What you’ve done sets a terrible example for this misguided mobile market. Pull the plug please.

Comment by hendry

Your original point was that mobile (.mobi) sites need “DEVICE DEPENDENT DETECTION MODIFICATION” to display on the desktop. This wasn’t what you meant I suppose since it’s obviously not true; XHTML MP and Basic render fine. chooses to send desktop browsers away from its mobile site for a different reason: it’s hardly optimal for desktop users, not because it doesn’t render.

I didn’t ask whether mobile web Gmail can be as fast as a Gmail J2ME app. I asked whether you can write one Gmail that works nicely on the desktop and mobile, and you dodged that question.

“Gmail” is exactly what you need to address to make your points relevant. This site works well on phones and browsers, no doubt. 99 others won’t even when authored with the latest CSS tricks. I don’t think you can say, sorry, don’t offer Gmail at all, or not on mobile, because it’s not what the web is for, etc.

What do you for people in, say, India, who have a cell phone but no computer? who don’t have an N60 or fancy Blackberry? These are issues that we can’t ignore. We can’t say, sorry India, GET A REAL PHONE/BROWSER! How do we get them Gmail? The practical value trumps any theoretical arguments about how markup and CSS should be used—points I don’t disagree with.

The Google link you provide shows ten results about you (well someone with your name), so not sure I follow your point there? There are two mobile web results and eight transcoded results from the web. I am not surprised that there are better informational links about you on the web. Seriously I’d be happy to hear what you think should have been done better here—I’d assume you like that we’ve finally gotten rid of the Web / Mobile Web distinction here, and only indicate which links are to “native” mobile sites and which are being chewed up by our transcoder because, unfortunately, almost every single desktop site on the planet doesn’t work on a mobile phone directly.

Comment by Sean Owen

Ok the very real problem is that there are LOTS of phones out there in the market place with crap browsers.

Though if you say OK we are going to support your special “mobile environment”, then the world of crap browsers isn’t going to go away is it?

I am worried that you will make the problem worse!

So yes, I am asking India/China to upgrade to a UA that is ‘Web capable’. This will come in time. I am just so tired of false arguments for special device dependent Web authoring because of the screen size, network or now ‘context’. Positive changes will also come when all the players understand what we should be shooting for, the Web. Else, the ‘powers that be’ will still ship WAP browsers like Nokia’s ‘Service’ browser.

It would be nice if Google gave credit where credit is due, and said something like if you want to do business with us, you need to rock with the Web API. Though I guess since you seem to be so desperate for market share, you’re willing to go down any path with Yahoo et al towards a waste of time&money known as the ‘Mobile Web’.

Comment by hendry

You are arguing against the mobile context entirely then. No browser can change the fact that you are browsing on a postage stamp, 10 keys, and dial-up-modem network speeds. Yes, this is a “crap” platform for accessing the Web. It not simply quantitatively different from the desktop experience, it’s qualitatively different. If you think bringing any internet service to these devices is a waste of time, OK, I think that’s actually a defensible position.

Does the Mobile Web harm anything? you’re saying that by making tiny pages, nobody’s incentivized to make a browser that handles bigger pages. I just think this is false. How about Opera Mini? Nokia’s Communicator? iPhone? I see nothing but a scramble to get something like a desktop experience onto something in your pocket. So no, the world of “crap browsers” seems to be going away on schedule.

Yes this is where mobile web access is headed. The Mobile Web is hopefully gone as a distinct thing in 5 years. But what now? I just don’t see the harm, aside from some theoretical point, in producing an experience that works for devices that otherwise can’t access any web content right now. But there is great harm in not providing this experience: you must think of those not privileged enough to have 3G networks and smartphones all over. (Here I mean North America of course. :) )

I do not understand your last paragraph. You still didn’t answer the Gmail question, and still didn’t explain your cryptic criticism of Google search results before, and you’re telling the developing world that it doesn’t deserve any web until it can pony up for an iPhone. Yikes! I don’t think there’s much point in continuing.

Comment by Sean Owen

I am super-happy for a browser based mobile context, like CSS media types. Lets be clear about this.

Yes, support for the ‘Mobile Web’ will promote device dependent authoring and hence ‘harm’ with the current confusion reigning the day.

You might not now have a GPRS network and a decent ‘Web capable’ mobile, but you will have. I am not being elitist here. I do not want exclusion, I want the one Web!

Also in the short term if you think there are billions of Asians going to use your XHTML MP ‘mobile web’ services, I think you need to go over there and have a look for your self. I don’t think you’ll ever reach this MBA‘s wet dream.

Your Gmail question? A compelling Web app for any device right now? I’m lost. Just repeat the question. :)

The links are all about me I think. Though they suck. Honestly. Only my CV link might be useful. What about my homepage with my mobile number?

Once again, you think I require 500USD from every developing world citizen before they get the Web? LOL! They will get the Web, when their mobiles and networks evolve to be ‘Web capable’ in time. The non-exclusive Universal Web will take a longer time to be realised if you push this ‘XHTML MP Mobile Web’.

Comment by hendry

Are we talking about the same Asia? which includes Japan and cHTML and i-mode? LOL Asia is already the biggest user of the little mobile web pages, and getting bigger because of developing markets on the other end. I am in the fortunate position of having access to information about mobile web usage globally and can confirm this is really happening. I’d also encourage you to “have a look for yourself” in India. There’s a lot of people happily using the web on phones right now.

Re: Gmail, you say that it’s easy to write one document (with many stylesheets) that renders nicely everywhere. I challenge you to imagine how you would implement Gmail this way. The desktop version’s markup alone won’t fit into memory in some phones. And what about devices that don’t render CSS? or run scripts?

While you can slam-dunk this elegant theory of markup on a static home page with a picture and a few links, or a blog, it won’t do for other sites; I don’t think you’re extending your theories beyond the kinds of pages people were writing in 1995.

I wholeheartedly disagree that the mobile web creates any disincentive to deliver a usable desktop experience on a portable device. I can point to things like Opera Mini that are already in existence. Demand for today’s mobile services, crap or not, stokes demand for tomorrow’s richer, more capable devices. In the meantime, I do not understand why one should deliver no service to lesser devices just because it doesn’t fit a neat model of page authoring. The “Universal Web” is about bringing access to all devices above all else. You personally are welcome to target whatever subset you like, but, why the withering criticism of organizations that are trying not to shut out most of the world?

… and PS your home page is the first result!! look again.

Comment by Sean Owen

I was referring to ‘MBA’ Asia as in the huge Indian and Chinese markets. Btw I used to work for Access Japan, so (I like to think) I have quite a good idea what actually happens. ;)

I do find your “lot of people happily using the web on phones right now” difficult to apply in countries like China&India. Though you are the Google guy, master of all information. :)

Gmail is a cutting edge Web application, which won’t work on most people’s crappy browsers I admit! I can’t give you a better answer, than to push for good HTML ‘Web capable’ browser requirements and hopefully this won’t be a problem. Anyhoo I guess your Java client can “bridge the gap” for now, eh?

And I’ll have to disagree with you about the mobile Web. If people are stuck in the XHTML MP (or worse WAP) mindset, it really hinders the roll out of ‘Web capable’ devices. That’s my experience and UK’s recent choosing of the “Services” browser over the Webkit Nokia ‘Web’ browser is a concrete example of a screw up of this mindset.

If you think by creating XHTML Mobile Web search you are unleashing a world of valuable information into markets like India&China, I think you must be kidding yourself. China&India are very much part of the world without your Mobile Web™. Internet cafes out there are everywhere and cheap.

The first result is my CV. Maybe you are using the Mobile version which is different to my desktop rendering of the XHTML?? :P

Comment by hendry