This is a collection of all the Statements of Interest submitted by workshop participants. Statements are alphabetically ordered by the family name of the submitter. Links are provided when a formatted version of the statement of interest or a position paper was submitted by the participant.
W3C
My main interest is in information that can be used in novel ways by a reader, by which I mean ways that the author of the information didn't explicitly provide for. As a concrete example, I have a program that every day extracts weather info from three HTML pages on three different sites and constructs a new page from that data. The program only uses standard Unix tools, such as sed and awk. It runs completely unattended. (This is in fact how I define the "Semantic Web.")
That means that for me, all URLs that give different results based on information outside the URL are evil. I would have to find out what HTTP headers give me the results I want and make sure that I send those. It also means that, ideally, every resource should only do one thing: HTML structures text, SVG provides graphics, PNG bitmaps, CSS style, etc. As soon as things are mixed together, information becomes hard to use.
CSS works on those principle (one URL = one resource, one resource = one small task) and thus I'm interested in extending CSS (as far as possible without diminishing usability) to cover more presentations where authors would otherwise have to resort to server-side content adaptation.
Interestingly, that principle also means that namespaces are bad, at least in so far as they are used to combine different formats in one file (thus namespaces in RDF are OK, since all that is combined is RDF). Putting, e.g., SVG inside HTML deprives the client of the option to skip the SVG. A link (SWITCH, OBJECT or similar) would be better.
Position paper: The place of CSS in device-independent publishing
W3C
As W3C Staff, i've been involved in the W3C Device Independent Working Group since the beginning of this activity at W3C, around mid-2000. Since Feb 2002 i'm the W3C team contact for DIWG.
My major interest in this workshop is about content adaptation and what it requires from the author.
My opinion is that it's possible to make device independent content, like WAI group succeeded to define accessible content, requiring on authors the add of few extra informations (like eg alt info.). IMHO, The goal is to define what are extra informations requested from the author so that a clever system would be able to adapt the content to the client characteristics.
My specific interests in this workshop are :
1- to hear from commercial product implementation : what are
their capabilities, what are constraints on authors, ...
2- to hear from w3c (particularly WAI groups) and non-w3c groups
with wich we have common or related goals : how they define the problem,
how they are actually trying to resolve it, ...
The leadership and credibility of the W3C organization depend not only on its working groups but also on the W3C's exposure to the global community. We should continue to learn from the community by listening to how companies and individuals meet the challenges that we all share - hear their suggested solutions, their difficulties, and their wish lists. We should also welcome their comments on our work.
This workshop is a way to expand our Working Group and include in it, if even for only two days, the larger community.
We have understood last year that Delivery Context and Authoring Techniques hold the keys to technically implement the Device Independence Principles. We have conversed with the community about the Delivery Context. Time is ripe now to discuss the Authoring Techniques.
I expect this workshop to be educational for the Working Group and an opportunity to open up communication channels with all the participants.
Catholique University of Louvain
I am a phd student from the UCL Telecommunication Lab and my interest lies in methodologies to develop multimedia mobile interfaces. I am primarily involved in the Salamandre project whose goal is to design and implement an adaptation mechanism for delivering web content to different platforms and devices. The techniques investigated within Salamandre will be applied to realize the GUIs of ARTHUR (ARchitecture de Télécommunication Hospitalière pour les services d'URgence), an emergency depatment management system that runs on desktops and PDA.
HP Labs
Since I joined HP Laboratories back in August 2000 I worked on Custom Publishing with particular emphasis in the document authoring/engineering and repurposing. My interest is then moved to the definition of a common language that will enable the adaptation for multiple devices broading the content representation from documents to web-content and web-applications. This research has lead to the development of a framework able to show principles as "single authoring" and adaptation techniques.
In the field of document authoring I am working and leading two, now open-source, projects: - WH2FO: a converter from Word Html to XML and XSL-T to produce XSL-FO (http://wh2fo.sourceforge.net); - FOA: Formatting Object Authoring tool, an autoring tool with document preview for creating XSL-T specialised into XSL-FO generation (http://foa.sourceforge.net)
Position paper: Device Independence Web Application Framework (DIWAF)
** Summary of interest in the Workshop **
** What's been happening
* current:
- the digital talking book specifications became an ANSI Standard.
http://www.loc.gov/nls/niso/
- working on a strongly trans-modal interaction representation in INCITS V2
http://trace.wisc.edu/world/v2
(See also Gottfried Zimmermann, he can tell you more).
- agreeing with the CSS Working Group that the control of adaptation of
multi-media aggregates needs more work.
http://www.w3.org/Style/2002/css-tv-comments.html
* against a backdrop of ongoing activities or roles:
- Chair Protocols and Formats WG http://www.w3.org/WAI/PF/
- Consult for Trace Center, e.g. http://trace.wisc.edu/docs/ud4grid
** What would I nominate for discussion
* targeted designs as undo-able or elective alternatives to generated forms.
[That's little-f forms as in "derived forms," not fill-in-forms.]
* the idioms that separate disability-adaptive interactions and media from
the mainstream, and the performance scales that motivate the differences.
Examples:
a. The preferred presentation order of the breadcrumb trail is context-dependent.
Top to bottom on the desktop screen, bottom to top in eyes-free audio. All
driven by the linear vs. random access and resulting speed.
b. When is plain text not plainly legible? When you're dyslexic and the text
is in a Semitic language with the vowels suppressed.
http://ubaccess.com/hebrew-access.html
c. Whether to confirm inputs or not depends on a) the persistence of display
and b) the symbol error rate in input channels. It's a kind of variation in
the interaction that should respond to delivery context factors beyond the
hardware characteristics.
http://www.w3.org/TR/UAAG10/guidelines.html#gl-user-control-ui
* the digital talking book as a mature trans-modal content representation.
* the minimum-cognitive-risk website profile as a valuable reference design.
http://www.learningdisabilities.org.uk/html/content/webdesign.cfm
* transformation to Voice Browsing as perhaps the killer tool for checking microclimates of explanation?
.. Brainstorming:
http://lists.w3.org/Archives/Public/wai-xtech/2002Jul/0015.html
.. Related experience in what "microclimate of explanation" actually reaches
the user; html:td.headers in html:table.
http://www.w3.org/2002/Talks/06/24-US_FedStatsWorkshop/slide3-0.html
** some references not to ignore:
some WAI comments on Device Independence
http://lists.w3.org/Archives/Public/www-archive/2001Nov/0069.html
Interaction: User Agent Accessibility Requirements
http://www.w3.org/TR/UAAG10/
Content: Web Content Accessibility Guidelines
http://www.w3.org/TR/WCAG20/
http://www.w3.org/TR/WCAG10/
Authoring: Authoring Tool Accessibility Guidelines and techniques
http://www.w3.org/WAI/AU/#Deliverables
HP Labs
As a researcher at Hewlett-Packard Laboratories, Bristol UK, I have been working on topics related to Device Independence since early 2000. In October 2000, I helped to organise and host the first W3C workshop on Web Device Independent Authoring. I was one of the founding members of the Device Independence Working Group in February 2001, edited the 'Device Independence Principles' working draft of September 2001, and became chair of the group in March 2002.
As one of the organisers and a program committee member of the DIAT workshop, I helped draw up the call for participation, which sets out the goals for the workshop.
My main hope for the workshop is to assess whether device independence is currently best addressed by appropriate use of existing techniques, or whether (and if so, what) new authoring markup is required.
Sony Corporation
Background:
Growing number of Web accesses using non-PC networked appliances, e.g.
TV, mobile phone and PDA, is raising importance of content adaptation
technology. In the meantime the recent advancement of XML technologies is
also predicting the evolution of content management.
Content Adaptation:
XML technology gives us a lot of possibilities to adapt XML descriptions
to devices and users. RSS, XSLT and CC/PP may provide good means to
transform XML descriptions and to create specific Web pages. However,
such technologies still cannot provide general means to automatically
adapt XML descriptions into the context of usage of contents.
Well-Defined Structure:
The recent development of Web service technology gives us another
possibility that service providers and content holders can provide their
services in a well-defined data format and through a well-defined
interface. On the other hand, domain specific XML descriptions, e.g.
NewsML and TV-Anytime, also provide information in well-defined data
structure.
Context Description:
One of the missing technologies may be a standard way to describe the
context of usage environment of contents, information about devices,
network and users. Even if CC/PP provides a way to describe such
information, it has too much flexibility to implement actual products
considering general applications and future advancement.
Work Together:
All these advancements may help us well to some extent. But they are
still not enough to automatically reconstruct XML descriptions into
solid presentation of contents effectively. During this workshop I
expect to share crucial issues lying on the device independent content
creation and its architecture with those who are involved in similar
problems.
Nokia Research Center
I am interested in device independence in general and specifically technologies that make it easier to adapt applications to the needs of mobile phones. I am personally involved in the work that is outlined in our position paper entitled: "Definition and prototyping of a Renderer-Independent ML". Furthermore, I am the Nokia representative to the W3C DI working group.
Position paper: Definition and prototyping of a Renderer-independent ML
DoCoMo Euro-Labs
Contribution
As a researcher on Man-Machine Interaction for DoCoMo Communications Laboratories Europe GmbH, the presentation of Web-based content on handheld devices is a very relevant topic for me since it requires to consider the devices capabilities such as the screen real estate , the input mechanisms as well as the users' tasks, preferences and location (context-awareness). The conception of usable devices based on technology standards would certainly reduce the need to create device-specific content.
User testing is a method often applied to understand how users interact with devices and what are the usability issues they are confronted to. As part of an iterative process, this method can finally lead to the design of usable devices or the improvement of existing devices. Different usability experts such as Jakob Nielsen agree that the following features contribute to the improvement of handhelds usability:
The only existing products that correspond the most to this description are PDAs (e.g. Palm, Pocket PC). Because of their limited screen real estate and keypad, mobile phones are not considered as suitable as PDAs for presenting Web-based content.
More generally, it is often stated that keypad, roller, selection bar and micro-keyboards are not the appropriate features for enhancing the usability of handheld devices. Hence, this should be taken into consideration during the development of future handheld devices and while authoring content to be presented by these devices.
The second important point is that the web-based content delivered to the user should also be based on his or her location, preferences as well as context of use. The capability to determine the location of the user and his or her context of use can be defined as context-awareness. Considering the physical capabilities of the devices and improving the context awareness could finally help to determine which type and quantity of information can be presented to the user.
Interested Discussion Topics
According to these statements, I wish to see discussed how interaction, navigation and information presentation are considered by Device Independent Authoring Techniques? What kind of standardisation should be required for the content as well as the input and output mechanisms? To which extend a user-centered approach can be applied for the development of Device Independent Authoring Techniques?
In addition, it would be also very interesting to have discussions about the cost of device-specific authoring and the cost of services and devices difficult to use if the content is not correctly tailored for them.
Formatted interest statement (.pdf)
Mobileaware Ltd
Mobileaware is a leading provider of multi-device single-source authoring and transformation solutions. It has always been our position that annotated XHTML coupled with directed transformation is a minimum and sufficient mechanism to enable device independence. Annotation is a key feature of our modular extension to XHTML and we are keen to see the benefits of annotation being made available to all through the standardisation processes.
Annotation is the method by which an author adds structural information to content. This information pertains to the relationships of subsets of the content, their relative importance, their suitability to different contexts etc. Annotation may be provided within the document (embedded) or attached to the document as separate data (external). Mobileaware currently employs an embedded approach to annotation.
A directed XHTML transformation takes its cues from the original content, from the accompanying annotation and from the delivery context. Annotation is represented in the document object model in the same way as other content nodes, but the effects of annotation can span the entire model, leading to very flexible transformation processes.
Annotation techniques have been proposed in the past, but these have been complex and so different from traditional content authoring that mainstream adoption is unlikely. Mobileaware believes that annotation can be made simple and compatible with existing authoring techniques, which will ease adoption and thereby address the multi-device challenges.
As part of a workshop presentation, Mobileaware intends to demonstrate a subset of its annotation technologies, and to explore the future possibilities of the techniques, leading to greater power and flexibility while retaining simplicity and compatibility.
The key to Mobileaware's technology is in its ability to make maximum use of information provided through annotation, but we also believe that annotation has many other uses (e.g. intelligent caching, content aggregation etc.). It should be obvious that agreement on annotation techniques would be of considerable benefit to the Web community. Mobileaware will contribute to any activity that can bring about such an agreement.
Position paper: Position Paper from Mobileaware Ltd
DaimlerChrysler Research and Technology
We, i.e. DaimlerChrylser Research and Technology, are interested in participating W3C's Workshop on Device Independent Authoring Techniques.
Our interest in the workshop is it to share ideas on Web Services as the basis of Device Independent content publishing.
Position paper: Device-Independent Web Applications Based on Web Services
W3C/Keio
As W3C Staff, i've been involved in the W3C Device Independent Activity and been this Activity Lead.
My primary interests is how document profile helps to contents transformation. Idea and concept of document profile might be understandable. There is no doubt we need a profile for contents description. But I don't know what kind of information we be needed in real world and how to write them. We don't have actual language do describe contents description. At the workshop, I would like discuss document profile description in syntax/ontology with participants, and hearing from experience people.
Pixelpark AG
Pixelpark's Interests
How multi-device delivery affects:
How existing authoring techniques can be used to achieve greater device independence:
How to minimize the cost of authoring:
How to control the adaptation process:
We could give a presentation about the DICA 2 project (see summary in the position paper).
Position paper: DICA: HTTP-Based Content Negotiation with Cocoon
SAP AG
As a research leader at SAP AG - Corporate Research in Germany, I coordinate our work in the area of web engineering for device independent application development. In this area I am also the project manager of CONSENSUS (a project partially funded by the European Commission) with partners from IBM, Nokia, Fujitsu-Invia, UBICALL and CURE (www.consensus-online.org). CONSENSUS is to work on the cost-efficient development of usable, device-independent applications. Major goals in this project are (1) the understanding of usability specific aspects of mobile applications, (2) the standardization of a mark-up language to be used for the specification of device independent applications, and (3) to build a prototype architecture adapting application based on the context of use. Besides the CONSENSUS project we have several other activities in the Advanced Customer Interface research program where we support the development of (mobile) applications.
As one of the co-chairs of the DIAT workshop, I participated in the definition of the goals for the workshop. My major interest in the workshop is to collect the state-of-the-art and new concepts in device independent application authoring. The results of the workshop should finally contribute to our work at SAP and in the W3C Device Independence Working Group.
Position paper: Definition and prototyping of a Renderer-independent ML
INRIA Rhône-Alpes
Nowadays, there is a plethora of exotic electronic devices such as pagers, PDAs, color cellular phones and there is no sign that the diversity of their characteristics will diminish anytime soon. Developing sophisticated solutions and architectures becomes so indispensable to enable the access of the Web content by heterogeneous devices.
The first step of the adaptation process is to start with a content model used in the content authoring. In the Device Independent Authoring Techniques Workshop, we are particularly interested by the new authoring mechanisms and technique that can be used to perform and ameliorate a universal access of the Web content. Adaptation solutions that we try to design encounter usually many problems related the absence of adaptable and flexible content. Ensuring efficient authoring techniques, which should be device independent, will make easier the application of future content adaptations.
Our contribution is directly related to the generation of the content by adaptation using: media adaptation and structural transformation (that includes the use of XSLT). In our approach these techniques are delegated to a third entity (the proxy) placed between clients and content servers. The proxy controls the adaptation process and makes the delivery of adapted content more efficient. The proxy optimizes the effort of original servers related to services customization and tries to use efficiently the network resources.
Position paper: Content Adaptation and Generation Principles for Heterogeneous Clients
Volantis Systems Ltd
Volantis would like to contribute to the workshop by describing the principles that underly its approach to device independent authoring. These include a device independent markup language, based on XHTML, and mechanisms for capturing and using device dependent information for stylistic control, physical placement of material and use of media resources.
In addition, Volantis would like to contribute its practical experience to discussions on the topics suggested for inclusion in the workshop. In particular, Volantis has experience of how multi-device delivery affects interaction, navigation, visual and audio style, visual layout and the granularity of authoring units. It also has expertise in the use of CSS and XSLT style sheets in multi-device environments, of device independent authoring languages, and of many techniques associated with minimising the authoring effort associated with developing multi-device sites.
Formatted interest statementPosition paper: A Device-Independent Method for Web Site Authoring
Covigo Inc
I am very interested in open and standards-based techniques for Single Authoring. Although device detection is a well understood problem implemented with standards such as CC/PP and UA/Prof, there has not been enough work on the concept of authoring device independent content. My colleague (Krishna) and I will be presenting a position paper on a Single Authoring Framework with advanced adaptation concepts such as device specific Transformation, Assembly, Pagination, Morphing, Media adaptations, that especially help to deliver multi-channel presentations.
Position paper: User Interface Adaptations: Indispensible for Single Authoring
University of Salzburg
My name is Guido Menkhaus and I am member of the Software Research Lab of the University of Salzburg, Austria. I am looking to attend the workshop on behalf of the Software Research Lab.
The Software Research Lab focuses on software architectures in mobile computing environments. It is engaged in projects concerning location awareness and user interface adaptation to mobile computing device characteristics.
User interface transcoding is the process where content is translated, adapted from one representation to another. However, the content is generally visualized unaltered on each platform, from a compositional, navigational layout point of view. The task of transcoding does not address a main obstacle to device-independent content authoring: the target devices' different presentation capacity (here the different screen size).
Assuming a general model that partitions user interfaces into windows and widgets, we believe that it is insufficient splitting a window uniquely on the basis of the size of the widgets into a new set of windows. Without additional semantic information this process results in a set of windows that can only be navigated in a linear, sequential way. To access the last widget of the original window, each window of the linear set has to be traversed. This process is often called fragmentation [1] or pagination [2].
The challenge is to remodel the widgets of a window into a new composition of windows with a reasonable flow of transitions between them by integrating semantic information between widgets. Semantic information may be expressed with relations such as " A depends on B ", " A contains B " etc.. The semantic information is inserted similar to the integration of link information with XLink [3]. The new structural layout is generated using a clustering mechanism that is based on size and semantic information.
We believe that an intelligent navigation and layout adaptation process requires semantic information about the relation between user interface elements. The Software Research Lab is about to start working on XML Semantic Linking Language (XSLink), a language describing the relation between elements in XML documents.
Formatted interest statementIBM
I have been working on the subject of how to economically develop applications that need to target multiple types of device and multiple classes of user for some years. Much of this work has taken place in IBM but for the last couple of years I have worked as a member of the W3C XForms Working Group which has specified a language which supports user interaction rather that just information rendering and is abstract in the sense that it specifies the "intent" of a user interface construct rather than what it looks like. My experience of XForms and the thinking that went into it should be valuable to the Workshop.
Fraunhofer Institute for Applied Information Technology
We are writing you on behalf of the Competence Center Barrier-Free Information and Communication Technologies (BIKA) from the Fraunhofer Institute for Applied Information Technology. We are interested in participating in the W3C Device Independence Authoring Techniques Workshop.
One of our main research topics is related to *accessibility issues* of Web Publishing Frameworks, Content Management Systems and Multimodal Interaction. We are active participants of several WAI WGS, and we would like to present a position paper in the topic (related to our research work), and participate in the workshop. We will send a more detailed brief description at the end of the week.
Position paper: Supporting device independent and accessible authoring by a next generation Web Publishing Framework
University of Karlsruhe
The IT-Management and Web Engineering Research Group (MW) is one of the six groups making up the Institute of Telematics at the University of Karlsruhe. The staff of the group includes a core of full-time employees, together with faculty and students from the University of Karlsruhe and professionals from the affiliated organizations. The group works in close collaboration with the Computing Center of the University of Karlsruhe and the third-party funded ubiquitous computing group Telecooperation Office (TecO).
MW aims to address key research, education, and technology development challenges created by rapid advancements in Web technologies, and the explosive demand for cost-effective principles to produce and provide high-quality Web-based systems and applications. The main competence of MW is in methods, techniques and tools that deal with component-based development and service-oriented architectures for the construction and evolution of Web applications. The mission of MW is to design and build the technologies and principles for enabling advances in custom-made and industrial off-the-shelf productions of Web-based information technology on the one hand and modern Web-based systems and applications in manufacturing, e-universities, e-business and virtual service organizations on the other. The central focus of MW's research activity is the development of powerful principles and systematic approaches to successfully construct, deploy, and evolve Web applications with emphasis on ubiquitous access, reuse, mass-customization, configuration management, and federation of Web-Services. An integral part of such Web applications represents user interaction. According to this MW designed and developed the user interaction model PetrIX. PetrIX connects the abstract model properties of XForms with the graphical notation of petri nets. One important facet thereby is the specification of device independent user interfaces.
The goal of the group's current development effort is the creation of an e-University prototype, based on a new system that enables the federation of services towards ubiquitously accessible Web applications, and on approaches that support the semantic interoperability of these sources and respect to all kinds of information appliances. The group's affiliates are working to provide the runtime-system, core-components and generic services needed for this effort. MW strives for balance between basic research, proof of concept by prototyping systems and applications, and collaboration with other disciplines in science and industry. An important purpose for these collaborations is to gain new insights into practical problems and to stimulate new ideas for our academic research.
xCA NV
Topics I'd like to see discussed:
What I bring to the discussion:
One of the biggest challenges in device independent authoring is delivering a high level of usability across devices. As devices proliferate, authors are less likely to have usability expertise for all devices. Usability resources will likely be spread thinner introducing a serious risk that usability and accessibility will suffer as the device landscape becomes more complex.
Currently, I am working on extending authoring techniques and tools to help authors create usable sites across devices and contexts without needing expertise in the usability aspects of a specific device class or usage context.
One aspect of this work is identifying how current device independent technologies, like XForms and device independent mark-up can be leveraged to help authors create usable sites/applications through integration with authoring tools. In particular, I am looking at how the semantics of device independence can be leveraged to automatically check a richer set of usability criteria then is currently possible. The other side of this coin is, of course, what usability engineering has to contribute to the definition of device independent technologies.
Another aspect of this work involves understanding the impact of device independence on the practice of interaction and visual design. The abstraction inherent in device independent design is at variance with the visual tools used by designers today. How will this affect adoption and the cost of authoring? How can current practices be extended to smooth the transition?
I hope my experience in the design and evaluation of user interfaces for PCs, PDAs, Phones and iTV coupled with a strongly human centered perspective will provide a productive counterpoint to the other participants.
ICS FORTH
We are currently involved with the development of an Apache module that can handle CC/PP headers. These headers are part of the HTTP headers. This module transforms the Apache server to a transcoding server while remaining a normal web server when serving requests without CC/PP headers.
The transcoding process uses ImageMagick library for the transcoding of the GIF images and the FFmpeg library for the transcoding of the MPEG and MP3 files. Furthermore, we have began to study ways of building a plugin for the Netscape browser in order to make it capable of sending CC/PP profiles to a CC/PP compliant web server.
We are interested in the web technologies that are involved with the Device Independent activity. Authoring techniques are important when many devices with different capabilities have to share the large amount of data currently existing in the web. By attending a discussion about authoring techniques a new dimension of the Device Independent activity will be presented to us and may give us new ideas on how to continue our current research.
xCA
As a market watcher concerning device proliferation and mark-up adoption for xCA - the multi-device enabler - I'm part of a team working on our solution for the device independence puzzle.
My main interest in this area is the further abstraction of the content from the presentation.
A first attempt to DI was to be XSL (and XSLT) which allows XML documents that information devoid of presentation be transformed and formatted. The ambition of XSL to be able to deal with most transformation and formatting issues makes it complex. This complexity drove the separation of the formatting part and the transformation part. It is my opinion that this movement in itself, together with the inherent complexity, caused too much confusion and uncertainty. As a result XSL and XSLT have been labeled as too difficult.
The UIML and XiML projects are another take on the problem. These new mark-up languages were created in order to effectively mark-up 'a page' irrespective of devices it would need to be shown on. For each target device or device family a translator is then created that translates the UIML/XiML in a document that can be rendered on the target device. This translation would be based on a rule set that incorporates information concerning both the device capabilities, the user preferences and the semantic relationship between the different page elements, so that these page elements can be ordered and marked-up in the best possible way.
It is my opinion that although the different technologies (RDF, CC/PP) that can assist in this translation process are valid approaches, they make the translation process a very difficult one. This is mostly because information from several locations needs combining and managing (manufacturer CC/PP profile, user preferences (possibly located on the device), rdf information).
As has been shown by WML, a mark-up language/technology that is too complex and restrictive hampers its easy adoption and use (although I fully agree that other factors have contributed to the slow adoption of WAP/WML). The problem in which RDF, CC/PP and UIML (or XiML) need to be combined to abstractly describe pages so that they can be used everywhere will prove insurmountable for many developers.
It is therefore my belief that DI will only become adopted when a language exists that combines the information needed for DI in one technology and one location. The requirements for such a language would be:
As an example I envisage a page designed that incorporates streaming media, high end graphics, end user interaction, ... to be built out of several sub-pages. It is then upon the device when requesting the page to determine what parts it can show/use and what the user wants. This places a big part of the DI problem to be resolved by the manufacturer of the device by building the needed functionality in the browser. But this should be not too big a problem as the manufacturer is the best placed 'person' to know what its device is capable of (rendering, network connection, user preferences,....)
Should such a language exists that can be directly served towards the devices then this would greatly simplify the problem for designers. Off course designers would not be exempt of making good pages and taking in account what devices are available and what they can achieve, so that the most basic page component will cater for the most basic device needs.
I would very much like to hear other inputs and opinions on this point of view, as I am sure there are many oversights and things not accounted for yet.
I.S.T.I.-C.N.R.
My name is Carmen Santoro, I am currently working in the CAMELEON Project (http://giove.cnuce.cnr.it/cameleon.html) whose objective is to build methods and environments supporting the design and development of highly usable context-sensitive interactive software systems. I would like to participate in this workshop in order to share our experiences in the subject of methods and tools supporting design and development of multi-platform interactive applications.
University of Kent at Canterbury
We have recently investigated the convergence of digital broadcast and web technologies. This led to contributions to the ISO standards process [1] and a number of papers that viewed interactive TV as but one element in the emerging ubiquitous computing environment.
As part this work, we used XSLT to transform existing web pages into the ISO MHEG-8 standard, for viewing within a TV environment [2]. While XSLT was able to meet many of the transformation requirements, there were limitations to this approach. One such limitation was the need to utilise external functions for some tasks. A more general problem was the lack of sufficient semantic information encapsulated within some web documents to enable a fully effective transformation.
Because authors often assume that documents will render to a PC browser window, automated transformations can result in some information loss. Additional semantic information needs to be explicitly included in the document to capture the author's original intentions.
It is our contention that much of this information can be expressed as constraints. For example, the author might want to indicate the relative positions of two objects within a document so that a piece of text is always displayed below an image. A scalable object might have a minimum display size specified for the user to recognise sufficient detail.
In the general case we consider constraints to be generated by the document context which includes author and user preferences, device capabilities and external information. External context, either generated by external sensors or directly input by the user, might include location information, time or temperature etc. The process of transforming the document thus becomes essentially a constraint resolution problem.
Work by Borning et al. [3] considered some aspects of constraint based document layout for the web. Additional work on constraints extensions to SVG [4] might also be applied to other XML based languages. There are a number of existing constraint-solving algorithms that could be utilised in a more general model for DI Authoring. Such a model would also enable the integration of dynamically acquired content that already contains constraint information.
A system of preferred constraints based on a constraint hierarchy, would enable constraints with stronger weightings to be satisfied in preference to weaker ones. If no solution can be found for a given set of constraints then an author might provide alternative style sheets with an indication of the classes of device for which each is appropriate.
Some questions
[1] ISO/IEC IS 13522-8 (MHEG), Information Technology - Coding of
Multimedia and Hypermedia Information, Part 8: XML Notation for ISO/IEC
13522-5 (MHEG XML), ISO, December 1999.
[2] V Valliapan, D Shrimpton, C Dobbyn and T Casey, 'Transforming Web Pages
for Interactive TV Using XSL', ICME2001, Tokyo, August 25th 2001
[3] Borning A et al., 'Constraint-based document layout for the Web',
Multimedia Systems, Vol 8, pp 177-189, 2000
[4] Badros et al., 'A Constraint Extension to Scalable Vector Graphics', In
Proceedings of WWW10, pp 489-498, Hong Kong, May 2001
SAP AG
I'm working for SAP Corporate Research in the area of mobile computing and gained experience in creating mobile business applications and mobile infrastructures. My special interests as well as my contribution to the workshop are in device independence for mobile business applications, improved usability for mobile applications and software methods for accomplishing those goals with respect to implementation/cost efficiencies.
My interest for the Device Independent Authoring Techniques Workshop is my involvement in the preparation of the "Authoring Scenarios for Device Independence" document. In addition I'm interested using the results from the workshop within our projects.
Position paper: Definition and prototyping of a Renderer-independent ML
Sun Microsystems Inc
I am the project lead and architect of a device independent portal server software product [1] from Sun Microsystems, Inc. I also serve as a member of the W3C Device Independence Working Group [2] and as specification co-lead of the JCP JSR 188 [3].
My contribution to the workshop would be knowledge and experience of implementing a server side solution for content authoring, adaptation, and delivery. I would like to see discussed the current issues that device independent content authors are facing and the various techniques employed to address them. I would also like to propose and receive feedback on how JSR 188 can potentially address the needs of device independent web application developers.
[1] Sun[tm] ONE Portal Server: Mobile Access Pack
http://wwws.sun.com/software/products/portal_map/home_portal_map.html
[2] W3C Device Independence Activity
http://www.w3.org/2001/di/
[3] JCP JSR 188 CC/PP Processing
http://www.jcp.org/jsr/detail/188.jsp
Position paper: Authoring Device Independent Portal Content
Position paper: Developing Device Independent Java Applications with JSR 188
Covigo Inc
I am very interested in single authoring frameworks/techniques for Device Independence authoring and presentation based on delivery context (Connection + device + personalization profiles). My colleague (Srinivas) and I will be presenting a position paper on a Single Authoring Framework for Java Server Pages and an XHTML extension module for device independent multi-channel presentation design.
Position paper: User Interface Adaptations: Indispensible for Single Authoring
IBM
The goal of the Device Independence Working Group in general, and of the Authoring workshop in particular, are very close to the goals of a project called Consensus (http://www.consensus-online.de). IBM is participating in that project. Consensus aims to define mechanisms to augment application data with meta information such that intelligent adaptation and filtering mechanisms can use the meta data in conjunction with context information to optimize the presentation for a particular user with particular preferences who uses a particular device. It is hoped that such mechanisms will improve user experience with mobile devices while keeping authoring effort for a multitued of different devices low.
The meta information will be expressed in XML and could be considered as a preliminary attempt of the DI group work item "3.2.3. Interaction markup". An architecture which captures the necessary context information, augments the application data with metadata and performs the subsequent adaptation and filtering steps, is under discussion within the project. Completion of a prototype implementing that architecture, including an authoring tool which generates the to-be-defined markup, is scheduled for year 2003.
We think that early feedback from our project to the DI group can help in perfecting the results of charter items 3.2. Reciprocally, the different views and aspects brought into the DI group's discussion will help to improve our project's specification, thus accelerating the feedback loop between the Consensus project and the DI group.
In summary, the Consensus project can be considered as a testbed for or instance of a DI authoring environment, to study capabilites and limitations of such an approach. The project is partially funded by the European Commission. Close cooperation with standardisation groups is an explicit goal. Further members of the consortium carrying out the project who are also member of the DI group are SAP and Nokia.
Position paper: Definition and prototyping of a Renderer-independent ML
Web Commerce Group
Web Commerce Group is heavily involved in the device independence arena. We currently offer a product that addresses some of the issues faced by authors wishing to deliver content on the wireless web. One of our technical leads, Jason Williams, is currently serving on the expert group for JSR-188, which is chartered to deliver a Java API for accessing CC/PP profiles.
WCG promotes and uses an authoring technique we refer to as the "Add-on" technique. Our position paper explains this technique in detail and the issues it resolves.
Position paper: Content Adaptation with the Add-on Technique
DoCoMo USA Labs
DoCoMo USA Labs are interested in building device-independent user interface across heterogeneous devices. At present, we are concerning with the web service user interface and the standalone Java application user interface; and we consider devices ranging from PCs, Pocket PCs, to DoCoMo cell phones.
Within our scope of interest, we developed a single-authoring technique for building a device-independent user interface, which can be transformed into different device-specific user interfaces. Our technique is focused on the presentation layout. We believe that presentation layout plays a key role in the usability of web service/application's presentation.
However, in order to generate high quality presentation layout for varies device-specific user interface, authors are required to provide a layout specification for each single device. This requirement demands authors to have thorough knowledge of each device's supported markup languages as well as each device's capabilities. In our proposed method, we try to minimize authors' effort by just requiring a single layout specification from authors. We have tested our technique on different Java profiles, and made several observations from our prototype.
One of our observations is the presentation consistency. All generated device-specific presentations can be totally different or they can be fairly consistent. The extreme case of either approach can highly degrade the usability. We need to conduct further studies to find the balance between the two. There are other observations that are related to deployment of device independence described in the position paper. We would like to share our insight and discuss each observation in the workshop.
Formatted interest statement (.pdf)
Position paper: A Single-Authoring Technique for Building Device-Independent Presentations with Minimum Layout Effort (.pdf)
Trace R&D Center
The Trace R&D Center is an active member of the ANSI / INCITS (InterNational Committee for Information Technology Standards) Technical Committee V2. V2 is charged with developing a standard for alternative user interface access, which in particular will include a specification for a Universal Remote Console (URC). The objective of the URC specification is to let the user employ any access device to control any electronic device or service. V2 aims to define a standard that accommodates the needs of all users in a variety of delivery contexts, including people with disabilities using specialized access devices.
Staff from the Trace Center is currently leading two of the working groups within V2. In addition, Trace is hosting a V2 support project for developing prototypical implementations of the evolving standard. The prototype project helps V2 to explore possible solutions, and to vet draft specifications for feasibility and practicality.
From the perspective of user interface authoring, the problems and solution spaces seem to be congruent for both the Universal Remote Console (which would allow a variety of access devices to be used in order to control electronic devices and services), and for Web services (which are to be rendered on a variety of user agents). Obviously there is a strong interest (on both sides) to harmonize the two standardization efforts in order to have a common set of specifications that are part of both emerging standards.
Position paper: Technical Requirements for a Delivery Context Independent User Interface Model