Interview: Rotan Hanrahan on MobileAware Participation in W3C, Role of Standards

As part of a series of interviews with W3C Members to learn more about their support for standards and participation in W3C, I asked Rotan Hanrahan (MobileAware's Advisory Committee Representative at W3C) some questions.

Q. How is MobileAware currently participating in W3C?

A. In addition to nearly a decade of participation in W3C’s ubiquitous Web and device independence work, MobileAware is a founding sponsor of the W3C’s Mobile Web Initiative, in which we chaired the Device Description Working Group (DDWG), culminating in the creation of a new W3C Recommendation (Device Description Repository Simple API). The company is proud to have edited and contributed to several W3C publications over the years and to have participated in many W3C events including the hosting of, or participation in, various Workshops. It is important to MobileAware that participation is active and progressive. Our goal is to help extend the Web beyond the desktop.

Q. How do standards benefit MobileAware, and what W3C standards are of particular importance? Do your customers value standards support?

A. MobileAware is a specialist company, dealing with technical challenges that few appreciate and even fewer understand. We work in an ecosystem of specialities. To make all these work together, we have to establish some common ground. Interoperability doesn’t just happen; you have to make it happen. For us, agreement on the definition and use of the delivery context (for mobile in particular) was a key requirement. The delivery context is the set of all factors that could affect the adaptation and/or delivery of content or services via Web technology. It’s a central concept for many of our products. Our customers appreciate that we have shared our expertise with others and, equally important, we have listened to those who were willing to share with us. Furthermore, the W3C has produced many guides and best practices, which help one to judge the quality of solutions in this increasingly competitive market.

Q. Can you share any success stories about the use of Web standards?

A. Success was quick coming to MobileAware. In the months leading up to the founding of the company, before Y2K fever had set in, we had become aware of this new idea called XHTML modularization. It promised a way to extend into new platforms, and we were excited by the possibilities, especially as we prepared for our bold step to create a new platform for the mobile Web. We built upon XHTML using this new idea of modularity, adding the necessary metadata and structures that would empower authors to create content that adapts to new contexts, especially mobile. It was one of the most important decision we ever made, and extremely successful.

Q. How about any challenges you or your customers have faced when trying to use Web standards?

A. The biggest problem we have faced, and still face today, is decoupling the Web from desktop browsers. For example, we regularly face puzzlement from customers who wonder why a heading on one mobile browser looks so different from the same heading on another mobile browser. Explaining that “heading” is a conceptual notion that may be rendered differently in different contexts can be baffling to people from the pixel-perfect desktop world. Different presentation is an inevitable consequence of differing delivery contexts. Fortunately, our customers quickly learn to embrace this diversity, especially when we provide them the tools to do so.

Q. Many content adaptation products have been introduced or promoted in the mobile Web ecosystem as short-term to mid-term solutions to make the Web at large available on legacy devices with poor browsing capabilities. New mobile devices ship with more powerful browsers. How do you see the role of content adaptation for mobile devices in five years from now?

A. Early content adaptation solutions were motivated by the need to work around deficiencies and extreme variance in new browsers, especially those on mobile devices. These are the short-term to mid-term solutions that you hear about. Adaptation has moved on. Diversity is increasingly seen as a good thing, an opportunity. Deficiencies are becoming a thing of the past. Indeed, many mobile Web-enabled devices have capabilities that you won’t find on many desktops, such as touch-screens, motion sensitivity, geo-location and more. Users are becoming more demanding, looking for personalized experiences, greater interactivity and more combinations of technology/service. Diversity is here and it is growing. In 5 years we will be totally immersed in it. The ever-present challenge is to be able to adapt to this diversity.

Q. Copyright infringements and security/privacy concerns are two examples of issues raised by Web publishers against the deployment of third-party transformation proxies by mobile network operators. How can the diverging needs of the actors of the delivery chain be addressed to the end user's final benefit?

A. Transformation proxies generally sit between the origin server (author) and the Web client (user). They act on behalf of the user or on behalf of the provider of the communication service. Adapting proxies seldom act on behalf of the author. In the case of legacy content, the authors did not anticipate the possibility of adaptation and so left no guidance as to how they might like this to occur. They did not, for example, identify which parts of their documents might be optionally removed in order to fit onto smaller screens. Some proxies may take this absence of author guidance to mean that they have a free license to do what they want with the content. Meanwhile, there are many adapting sites on the Web today that will adapt content in accordance with the author’s wishes (and the browser’s abilities and user’s preferences) to deliver a good Web experience. But this is being thwarted by some proxies that will masquerade as different devices, confusing the server by presenting it with false delivery contexts. What the proxy eventually delivers to the user is often poor in comparison to what the adapting origin server would deliver, because the proxy does not have the advantage of knowing the author’s intentions, including presentation, security and privacy concerns. In many cases the key problem is lack of respect for the efforts, rights and preferences of authors and users. Thankfully, efforts within the W3C MWI are proving guidance on how proxies can use available technology to be respectful of these concerns, though we will have to rely on the goodwill of the industry for such guidance to be universally applied.

Q. Most Web developments (think applications, widgets) are still either targeting the desktop world, or developed for specific classes of mobile devices from scratch. On the one hand, adaptation of a desktop site to fit mobile devices is often viewed as a costly and constrained work item. On the other hand, new capabilities offered by mobile devices seem to be mobile and platform specific. Can a true ubiquitous approach be considered? How is the mobile world impacting the desktop world?

A. Part of the problem is assuming that the starting point is the desktop. Sure, if a desktop site already exists, there is a huge temptation to assume that you will lose a lot of investment if you don’t make it the starting point of your mobile site. The results of this sort of thinking can be disastrous. Generally, when you look closely at a desktop site and you strip away the features that make it particularly attractive on desktop browsers, you reveal the true essence of your site. This is where you start, and in many cases this essence is in your back-end solution, not the presentation layer. The mobile Web experience is not just a miniaturized version of the desktop. You have to consider not only the context of the device (size, capabilities, limits etc.) but also the context of the user (in transit, in a hurry, focused etc.). While you may be re-using many of the familiar Web technologies such as HTTP and various markup languages, you will find that the workflow needs to change, the emphasis will change, and expectations need to change. The Web technologies will be ubiquitous, but not necessarily used in the same way. Certainly, for business users of the mobile Web, the browser is less of a surfboard and more of a machete.

Q. Content transcoding proxies deployed by operators can improve the user experience of many Web sites. However, when content providers simultaneously make efforts to make their site work on a mobile device, conflicts can arise. How would you advise both operators and content providers to manage this situation?

A. There are some simple ways that a transcoding proxy can avoid interfering with sites that already deliver contextually-appropriate content. Sampling is the easiest solution. For example, a test request from what appears to be a mobile browser may result in a mobile content response, in which case the proxy can record that this site should not be adapted further. Non-adapting sites can be re-sampled at intervals to see if they have upgraded to adaptability. Observing the returned HTTP headers also provides vital clues, especially the Vary header and No-Transform directive. Generally, proxies should err towards the least interference. Servers should, of course, also emit headers that can be used by proxies to influence their behaviour. Proxies should also announce their presence, so that origin servers have some way of recording proxied traffic (which may encourage desktop site owners to give more attention to the needs of mobile Web users). Users should also be made aware of the proxy, especially if there are security/privacy concerns. The W3C Mobile Web Best Practices group has being doing a lot of work in this area, and has been encouraging public discourse to highlight the issues and possible solutions.

Q. There is no precise definition of the term "mobile" but it seems to be shifting from a concept tied to a physical device to something closer to freedom, as it carries the expectation that a task can be started on a device and pursued on another one. Are new technologies needed to cope with this use case?

A. The DDWG charter described mobile devices as Web-enabled devices that are normally used away from fixed locations and are manufactured specifically to be portable and usable while being moved. While session mobility (migration) is technically possible, we currently don’t see much demand for it in solutions that are context-aware. This is because the workflow tends to vary, and different contexts will have different entry/exit points, so migrating to a different context changes the dynamics of the workflow. This can be confusing. Dealing with this is a challenge for Web applications, but as the browser becomes more embedded in our surroundings, session migration will be important. At this point we will be moving beyond content adaptation into application adaptation. So yes, we will need new technologies.

Q. What does MobileAware value in its participation in W3C?

A. This is a deep question, one I’ve been asked before in a different context, and I’m happy to repeat what I said at that time: The W3C is a facilitator for product making. The venue (being the environment in which Members operate) is a key factor in this facility and is the primary justification for the basic cost of membership. On top of this are the enablers of product development, namely the resulting specifications, for which the members are expected to expend additional resources for their creation.

Additionally, there are other valuable services of the W3C that act as the "glue logic". To start, the test suites and validators serve to engender quality into the W3C products. The feedback mechanism of public discourse ensures relevance to the Web community while also providing an essential further check on the validity/correctness of specifications.

Then we have the ongoing outreach and education, which in a way is akin to a marketing exercise motivated not by profit but by a sense of mission. The W3C also provides a locus of crystallization for exploratory work through its many workshops, IGs and the many talks by W3C staff as reported in our regular W3C bulletins. The W3C site provides relevant archival services, not only as an authoritative reference but also as an historical record showing how much of the Web architecture has evolved (a history from which we should draw valuable lessons). The W3C sometimes provides a service of industry/community representation, such as the times when W3C takes a message to world governments or defends against unfair patents, which is perhaps something that no individual member would want or be able to do. The W3C provides a badge of trust, which can be a source of confidence and assurance to Member's customers, enabling many niche solutions to emerge without the potentially negative label of "proprietary". In this way, the W3C helps the Web to evolve in novel ways. The W3C even demonstrates how to use the technologies through its open source work, with several very useful free tools. And I could go on...

One of the things that makes these truly valuable is that nobody (as yet) has attempted to place a monetary value on them. From the community perspective, the services of the W3C are orthogonal to the commercial world. The specifications are freely available and unencumbered by IP protection, the validators are always online for immediate use, the archives are searchable by anyone, the workshops are open to anyone who has something worthwhile to contribute (and the results/papers are always free to view), the outreach is fair and global. About the only thing that comes at a price is active W3C membership (i.e. fees and the costs associated with doing real work in the WGs), and MobileAware has been an active member for nearly a decade. We value our participation.

Q. Where would MobileAware like to see W3C improve?

A. There is a danger that W3C tries to take on too much. The Web is now an essential part of the fabric of our technologically advancing global community, and its reach is immense. It would be challenging for the W3C to try to take the lead in everything that has some Web aspect to it. Instead, it may be time to refine the focus, concentrate on the architecture of the Web and be ready to let some activities find a new home in existing or new organisations. The work of the W3C needs to be streamlined to keep up with the demanding pace of the Web. Goals should be more realistic, timeframes shorter, activities more flexible. There should be willingness to stop when necessary. Anything that takes more than two years from the point when the requirements are agreed is either too big or just not worth pursuing. We don’t have to do everything in a single bound; we can make this journey in small steps, and lots of them.

Q. MobileAware chaired one of the W3C working groups for several years. Why would a small company take on such a demanding role?

A. I had the personal privilege of working with many skilled, intelligent, wise and experienced people from many member companies (partners, competitors, observers, enthusiasts and more) over many years in W3C. Whether their respective companies have thousands of employees, or just a few dozen, it makes no difference so long as they can contribute their speciality to help the Web reach its full potential. MobileAware helped to highlight several pressing issues during the formation of the MWI, and when we were asked if we would help solve those issues, it was clear that we were at the right place at the right time. I may be the face of MobileAware in the W3C, but I am supported by a team of experts from our offices around the world, and with the right support any W3C member, large or small, can make a difference. For us, the reward has been to see how the mobile Web has become an integral part of the greater Web, and the challenges are being addressed day by day with abundant enthusiasm. It’s good to be here.

Many thanks to Rotan for his answers.