dotMobimobiThinkingmobiForgemobiReadyDeviceAtlasgoMobi
background image

One Web - Why does this stir up such emotion?

Jo Rabin's picture
Posted by Jo Rabin - 02 Nov 2006
Twitter share icon Facebook share icon Google Plus share icon

The subject of "One Web" stirs up much passion, though in general it's not clear at all what people mean by it. There is a wide choice of meanings to choose from:

  • a universal information resource created and maintained for the benefit of humanity;
  • free access to information because "information wants to be free";
  • a code word for not wanting the Web to be taken over by evil mobile carriers, and especially not their carving out "walled gardens" accessible only to their consumers;
  • another word for "net neutrality" i.e. not wanting internet bandwidth to be paid for differentially according to usage or priority, one class of travel only;
  • a silly idea invented by people who didn't realise that user interfaces need to be designed differently for different types of information, different circumstances of use and differences in presentational ability of devices;
  • the value of allowing all information to be interlinked far exceeds the value of the same information as islands.
  • and so on...

W3C Mobile Web Initiative and One Web

Over slightly more than a year I have had the privilege of working at the W3C editing the Mobile Web Best Practices (MWBP) document on behalf of dotMobi. In the course of this work the subject of One Web was hotly debated. In the end, we needed to create some words around what we thought One Web meant - and what we meant by saying that the Mobile Web is indeed part of the One Web. Along the way, as you can imagine, we had a lot of fun misquoting Lord of the Rings, Bob Marley and anything else that we could think of that included some phrase that started with the word "One".

Incidentally, you'll note that there is a semantic disconnect here - the document is entitled "Mobile Web Best Practices" - rather than being entitled "Best Practices for the Delivery of One Web to Mobile Devices", which is more what it should be. A bit more of a mouthful, perhaps, as well as having a bit of an unfortunate interpretation as an acronym.

In the MWBP doc we devote some space to a discussion of One Web - and my interpretation of our conclusions is:

  1. Content must be delivered as some form of HTML over HTTP.

    In this technical sense the One Web is self defining. If you have a User Agent that implements HTML transferred over HTTP then the Web should be accessible to you.

  2. Do not deny access to content on account of your presumptions about the user.

    Some applications are particularly suitable for use while mobile, and don't - on the face of it - make much sense if seen on a desktop. Others are the other way round. A lot of applications fall between these two extremes - i.e. they have some aspects that make sense while mobile and other aspects that make sense from the desktop.

    However, do not deny people the opportunity to access your content if they try to access some aspect of it from a context that you would not normally consider appropriate to that aspect.

  3. Devices must have a certain minimum level of characteristics to make them useful

    They must have a certain minimum screen size, some ability to select hyperlinks and input data and so on. This is called the Default Delivery Context.

  4. Thematic Consistency

    A key point here is that One Web doesn't say that you must serve exactly the same page format for mobile as for desktop use. That would not be sensible. What it does say is that when you serve content it should the thematically similar - i.e. that a page served from a particular URI should be about the same thing, even if the format or the exact expression is not the same on different devices.

One Web Detractors

Of course not everyone agrees that this expression of what One Web means is right, or even that One Web is a desirable goal.

Barbara Ballard presented recently at a Mobile Monday London event that I organised on "The Real Mobile Web" and talked about her views. Other presenters gave some different views, Tom Hume recorded some impressions of the speakers (in real time I believe).

It seems that to Barbara, we must understand that the circumstances of use while mobile are different - and she thinks the idea of One Web does not make sense. She's made a number of particularly cogent posts on this, especially her series "What's wrong with the Mobile Web", and she has written very comprehensive guidelines for developing mobile experiences - which you should definitely check out at her site if you haven't done so already.

Luca Passani, one of the co-creators of WURFL (much kudos to him and Andrea Trasatti for this work - see Andrea's article on WURFL), became frustrated with what he considered the W3C group's wrong-headed notions on One Web and went off and wrote his own best practices, which he calls GAP. To my mind GAP is worth a read - there's much to agree with, reinforcing as it does many of the things we espouse in the Best Practices document - though equally I would strongly advocate caution when it comes to implementing stuff that is only present in the OMA XHTML-MP world - as the worlds of the OMA specs and W3C are converging rapidly. It's not hard to write stuff that validates as both XHTML-MP and XHTML Basic 1.1 [which we are hoping will be exactly the same as each other]. As a general principle I believe one should write for the future and retrofit the past, where necessary, rather than writing for the past and retrofitting the future.

I'd also advocate caution as to the advice to "use adaptation", which is Luca's mantra. If you don't know what adaptation is, don't worry for now (or better still look it up in the MWBP) adapatation has its place but is not the answer to everthing. You can create some terrible sites with adaptation. Some of the most effective sites are very simple. There is nothing inherently bad in a one-size-fits-all solution, providing that the application is suitable for that kind of presentation. Equally, I'd say that if your application and its intended audience demand something of more than a basic level of sophistication, then one-size-fits-all is simply not the way to do it and some form of adaptation really must be used.

Strong One Webbers

I suppose that you could consider the W3C the 'Church of One Web' and is not at all opposed to adaptation, which is what you might infer from Luca's document. Indeed, quite the contrary. There is an entire activity devoted to it called "Device Independence", which recognises that if the Web is truly to be accessible from anywhere then content authors need to create their content in way that allows different sorts of devices to access it, and that adaptation processes may indeed be needed in the course of delivery of that content.

Nic Brisbourne, who attended the Mobile Monday event I mention above, posted strongly in favour of One Web on his blog The Equity Kicker. Nic cites the argument that people don't want different interfaces on their mobile to what they are used to on the Web.

Nic expresses this argument eloquently. However, I find myself quite sceptical about it. I believe that as Barbara says, you must be sensitive to the context of use. Not only are the device capabilities in many cases radically different, but different aspects of applications will appeal for mobile use as opposed to the desktop use.

Further, I'm not actually convinced that users demand the same interfaces to applications in different contexts. For example, you don't expect your car radio to work in the same way as the portable in the bathroom, or the tuner in the living room. They are all radios, you can get the same content on them. You can listen to FM classical music in the bathroom but you will have a better experience if you use your 7.1 speakers in the sitting room.

Weak One Webbers

Tom Hume (where does he find the time?) posted on this a little while back, and he and I exchanged some views which I expand on a bit, here:

A key point about the Web is that you can bookmark it and exchange bookmarks. Increasingly you will want to transfer those bookmarks between your various devices, especially between your desktop and your mobile. Consequently I think it is very important indeed that if you do transfer bookmarks, or if you send someone a link, you should have a reasonable expectation that it will work on the device they are using. As I mentioned above, the property of Thematic Consistency is what enables this. Adaptation is one of the ways of achieving it.

Some people, and among them representatives of Nokia, think that there is so much stuff out there on the Web, and that there's so much inertia as to the desktop format that the best think to do is to sell people more capable devices that have clever browsing technology on them, so that you can experience a desktop presentation on a smaller mobile format device. This raises the question of whether you do indeed need to create customised mobile vs desktop experiences. To my mind this is actually quite a complex area, introducing as it does a number of questions - among them the question of whether you want a customised mobile experience if you can get the desktop experience. I'm planning to post on this theme shortly - The Quart in a Pint Pot Browser.

I think there is also more to say about a custom mobile experience than that it is for use when mobile. One of the important characteristics of the mobile context is that you do not expect to capture the users entire attention, as you may do when designing desktop applications. There's at least some, if not a lot of similarity between that type of usage and the kind of use you expect from a desktop Widget.

Conclusions

I strongly agree with the idea that applications should, where appropriate, be tailored to the user's context and that there is a self-selecting set of applications that appeal while mobile. But how much of that perception is habit, and dare I say it, age? How will younger people see it?

Back in the early days of email, I remember that people said that no-one would want to read a document on-screen. The "paperless office" was a joke, because it introduced volumes of information that often 'needed' to be printed off to be consumed.

All that has changed. People's expectations about where, how and why they wish to consume information changes constantly. Before asserting that there is a complete difference between the mobile experience and the desktop experience, consider how things change - and how absurd it seems today to print off emails before reading them.

The Web is FOR mobile, stupid.

All this is compounded in parts of the world in which computers with large screens have yet to arrive. It is possible that in the not too distant future we will find that the majority, if not the vast majority of individuals who access the Web do so by a mobile device. At a minimum we should not be surprised if this turns out to be so.

Meanwhile, let's make sure our Web applications work on a reasonable range of mobile devices and have appropriately tailored interfaces to provide at least a functional user experience on those devices.

Check out what will be a growing body of explanatory tools, explanatory documents, how-tos and so on here at mobiForge. There are tools that help you get an idea of what your site looks like, how "mobileReady" its mark-up is, size, page load times and so on. Check out also the work of the W3C Mobile Web Best Practices Group to which dotMobi is a major contributor.

Jo Rabin
[Independent Consultant inter alia on Contract to dotMobi]


Posted by Jo Rabin - 02 Nov 2006

Jo Rabin's picture

CTO at dotMobi, co-Chair of the W3C Mobile Web Best Practices Working Group, co-Organiser of MobileMonday London.

Posted by passani 7 years ago

Jo, first, let me thank you for pointing to GAP after I criticized BP so much :)

Secondly, you wrote:

"one should write for the future and retrofit the past, where necessary, rather than writing for the past and retrofitting the future."

what about: writing for the present, as GAP suggests, and adjusting in the future when the technology allows new solutions?
WURFL was succesful because it used what works as a compass for the direction to take in each and every moment. GAP adopts the same approach. BP is about how mobile technology may or may not evolve in a more or less distant future. I think developers are better served by the first approach.

Luca

Posted by Jo Rabin 7 years ago

Luca

When I start a project I know the world will have changed by the time it is done. Doesn't matter if that is months, weeks or days. So what is the 'present' I might design to?

Good design anticipates and accommodates change - that much I expect we agree on - and much more besides, no doubt.

Jo

Posted by passani 7 years ago

>Doesn't matter if that is months, weeks or days.

but it matters when it is years!!!
I have been in this business long enough to know that what is discussed today will happen min 2 years down the road or, more likely, 3 or 4.

They were talking about XHTML MP in 2000. I saw the first proof-of-concept XHTML device in first half 2002. Vodalive launched in late 2002. It took another 2 or 3 years before it made sense for developers to ignore WML (and it still does not make sense totally).
Same thing happened with WAPPush, MMS and J2ME: 2+ years between hype and reality.
Now everyone is talking about Ajax, wanna bet it's going to take more than 2+ years before Ajax goes out of proof-of-concept phase?
Now, my point is that if a technology is proof of concept, it cannot be best practice....a best practice is solid, safe and consolidated in the industry today (give or take 2 years, of course ;)

Luca

Posted by melisscp4 2 years ago

passani wrote:

>Doesn't matter if that is months, weeks or days.

but it matters when it is years!!!
I have been in this business long enough to know that what is discussed today will happen min 2 years down the road or, more likely, 3 or 4.

They were talking about XHTML MP in 2000. I saw the first proof-of-concept XHTML device in first half 2002. Vodalive launched in late 2002. It took another 2 or 3 years before it made sense for developers to ignore WML (and it still does not make sense totally).
Same thing happened with WAPPush, MMS and J2ME: 2+ years between hype and reality.
Now everyone is talking about Ajax, wanna bet it's going to take more than 2+ years before Ajax goes out of proof-of-concept phase?
Now, my point is that if a technology is proof of concept, it cannot be best practice....a best practice is solid, safe and consolidated in the industry today (give or take 2 years, of course ;)

Luca

Sales Jobs