Sony Ericsson position paper - Mobile Web Initiative Workshop

Peter Stark, Application Manager, Sony Ericsson
18-19 November, Barcelona


Sony Ericsson supports the creation of the W3C Mobile Web Initiative and looks forward to Recommendations covering rich Web-based multimedia applications for mobile phones, and a permanent place for mobile phones in the W3C.

Our experience of the mobile Web is that few actually use it. Despite operator portals easily available from dedicated hard-keys (the 'operator key'), most users are reluctant to go online and browse the Web.

One problem is the lack of high-quality mobile Web sites. Although mobile phones is no longer a small minority on the Internet, the mobile version of most Web sites today is often just a dumbed down version of the 'real site'. Automatic content adaption (eg, use of UAPROF) is rarely implemented. As a result, there is little useful content on the mobile Web.

The lack of useful mobile Web sites is the primary reason why technologies that fit large Web pages onto small screens (with some clever layout strategy) are popular - something is better than nothing. New mobile Web browsers are 'bug-compatible' with de facto standards ('street HTML') and let users view the same Web pages they can on their PC. So far this has been possible mostly in the high-end segments (ie, Smartphones and PDAs), but the trend is coming to mid and low segments as well. Soon it will be possible from all phones.

In 2006 we expect most phones to have a mobile Web browser that is able to render almost any Web page on the Internet. WAP is history and, from a technology point of view, the mobile Web has converged with the de facto Internet standards.

But although the technologies converge, the mobile Web remains different because of different user experience (screen sizes, interaction models, etc.), different use cases.

Web 'browsing' is not the key use case

More fundamental than the lack of mobile Web sites is that Web browsing - the act of navigating HTML pages and following links - simply is not a successful use case on mobile phones.

In most regions, messaging (text and picture) and gaming are the key services that drive mobile data. The fraction of 'portal users' is typically much less than 10% of all users. The lack of high-quality mobile Web pages cannot alone explain why browsing has failed to become successful. Probably many other factors such that users are in different context and have different goals compared with when using a PC from home or from the office. Also, constraints such as screen size and interaction methods, also explain why Web browsing is so inconvenient that most users avoid it completely.

New mobile Web browsers render Web sites designed for PCs in 'smart way' on a mobile phone. This will make mobile browsing more popular. But we do not think this is the final solution to the mobile Web.

There is a concern that users will be as disappointed with a 176x220 version of their favourite Web site, rendered in a 'smart way' with all frames stacked on top of each other (some stuff removed), that takes several minutes to download over a 3G network and offers poor usability, as they were with with WAP in 1998. Of course, some users like this, especially those that have tried WAP, and now finally can do what they thought they would be able to do in 1998. The majority of users will be confused when they try to view a 1024x768 page on a 176x220 screen, regardless of how smart the smart rendering is. The Web usability that the PC-user experience, does not make itself available to the mobile user. And usability is critical for mobile services.

High-quality mobile Web sites are rare. Smart rendering of 'real' Web sites will make mobile browsing more popular but the smart rendering is unable to reproduce the usability that the PC-user experience. Despite significant progress in this field, 'browsing' from a mobile phone still remains an uncertain case.

Mobile Web use cases #1: applications

Getting users online is critical to operators and content providers. Otherwise users will not download content or use online services - online users spend more money than others. If browsing is not the key use case for getting users online - what is?

There are many ways of using the Web, in addition to browsing. Here are some examples of use cases, within two major trends.

The above cases have in common that the Web is viewed as an application-oriented platform instead of a document-oriented information network. The Sony Ericsson PlayNow(tm) music download service, for example, is a Web-based application, the user is unaware of the fact that the user interface is Web pages. We expect the Web to be used as a platform for similar, but more advanced, applications in the future.

To think in terms of 'applications' instead of 'documents' is nothing new. Already WML was heading in this direction. But the problem with WML, and XHTML/WCSS that came after, is that they are text-based formats. Today mobile phone user interfaces have evolved far beyond text towards advanced graphics. Users expect Web applications to have a similar look and feel. This is why Sony Ericsson supports SVG Tiny 1.1 in most phones since the K700i and supports the new W3C working group on Compound Document Formats (CDF) that we believe will enable a new kind of mobile Web applications. We expect more animations and multimedia being integrated into mobile Web portals, and the combination of all Web technologies (SVG, SMIL, XHTML, CSS, and ECMAScript) is the key to a standard solution for Web applications.

Mobile Web use cases #2: information updates

To further address the issue of how to get users online, and their reluctance to browse the Web in the traditional meaning, we look at another major trend.

Push services are on the rise on the Internet, based on the de facto standard RSS. We believe that RSS has a great potential in mobile phones, as a technology to automatically provide updated content to users - accessing the Web without browsing.

Standardization activities

Between 1996 and 1999 the mobile Web was based on its own standards (eg, WML and cHTML). Since 2000 it has moved towards W3C standards such as XHTML, CSS, and SVG. Today, XHTML is widely adopted on the mobile Web; In 2005 we expect most mobile phones to support SVG Tiny. Today we have reached a point where the mobile Web is based on the same technologies as the Web, and even goes beyond the Web with wider support for XHTML and SVG.

Standardization has mainly worked so that the core technology has been developed by the W3C (XHTML, SMIL, SVG, CSS), then 'mobile profiles' have been developed in the OMA (XHTML Mobile Profile, Wireless CSS) and 3GPP (SMIL/MMS), and sometimes by subgroups in the W3C (SVG Tiny). In some cases it is a problem that the profiles are too many and developed by separate groups in different organizations.

Sony Ericsson welcomes a mobile Web initiative in the W3C and hopes that it will take of the following things:

The relation to Device Independance must be clarified.

And finally we believe that a mobile Web initiaive in the W3C will establish mobile phones as first class citizens of the Web, which is what they are.

Sony Ericsson has been working with mobile Web standards for many years, in the W3C as editor for [XHTML Basic] and [CSS Mobile Profile], and as editor for several specifications and requirements in the OMA Mobile Applications Environment (MAE) working group.