Reflections On 2006 - Accessibility

Whither WCAG 2.0?

It had been expected that the WCAG 2.0 guidelines would be released by W3C WAI this year, to replace WCAG 1.0 which came out way back in 1999. However the reviewing process for WCAG 2.0 seems to have slowed down (if not stopped) due, I suspect, to the huge influx of comments received after Joe Clark published his “To Hell With WCAG 2.0” article published in May on A List Apart. Joe’s posting generated much debate from commentators such as The Pickards, Stephen Downes, Joe Dolson and contributors to Accessify Forum as well as follow up articles and discussions on Joe’s own Blog.

Christmas Quiz - An Answer

In my Christmas Quiz posting I asked which of the following are open standards:

Flash PDF RSS 1.0 RSS 2.0 MS Word

Before giving my thoughts on this, I will comment on the responses.

Your Responses

James Brown felt that “first, my definition of an open standard: publicly available and able to be implemented such that two or more different uses of the standard will be compatible and accessible.” He then went on to apply that rule to the examples I gave and felt that each of them had characteristics of openness but not completely so.

Phil Wilson felt gave his “definition of an open standard: publicly available and able to be implemented such that two or more different uses of the standard will be compatible and accessible.“

Ravis Reddick thinks that “Flash itself isn’t the standard, it’s the official authoring tool and general brand name for the technology. Other authoring tools can author ‘Flash’; the export format is various version of Shockwave Flash. I think this can now be authored in an open way using an intermediary format“

Kevin Ashley’s view is that “Flash, PDF and more recent versions of MS Word are all open in the sense that the file format is published, and it is possible to create tools to read/render content in those formats using nothing except the published standards.” He went on to add that “All of them are not open in the sense that new versions of the formats can be created at the whim of the company that owns them.” Although he is “not so worried about that; if I don’t like PDF 1.7 because of some new license twist, I won’t use it. All the stuff I already have in PDF 1.x (x“.

And finally Steve Nisbett feels that “PDF, FLASH and MS may well be seen as ’standard’ - available all over the place, but they are certain not Open.“

My Views

I can recall back in 1993-4 having similar discussions about open standards. Back then I can recall arguing that a system such as the Web because of the:

• Open standards.
• Client software was available on a variety of platforms.
• Service software was available on a variety of platforms.

There were a number of other bullet points (which I have forgotten) but I do recall that my definition was used in a response to other systems which were competing with the Web (and in my opinion where inferior to the Web). These included such as the Guide hypertext system, developed at the University of Kent and Microcosm, an “Open Hypermedia Environment for Information Integration” developed by Professor Wendy Hall and colleagues at the University of Southampton. As these products were being pushed in 1992/3 as potential alternatives to the Web (and came from well-established Computer Science departments with good reputations for the quality of their research) I came up with my definition of openness, which I gave at a day’s workshop on Hypertext Systems On Unix Platforms at the University of Kent back in 1993 (it was originally intended, I think, as a promotional events for Guide, and I was the token person describing at alternative approach!) However my definition of openness was clearly not an open definition - it was intended to embrace my preferred solution, at the expense of the competitors. I’m aware that others take a similar approach (which all too often seems to resolve to “an open standard is one that Microsoft have no involvement in, no matter how proprietary it may be!”).

The respondents to this quiz take a more honest approach, I’m pleased to find.

So what are my thoughts?

The examples I used were taken from a presentation of a paper on A Contextual Framework For Standards which I gave at the “Workshop on E-Government: Barriers and Opportunities” which was held in Edinburgh in May 2006. The opening speaker at the workshop was Ivan Herman of the W3C (and currently the lead of the W34C’s Semantic Web activity). In his presentation Ivan cited an EU definition of openness:

• The standard is adopted and will be maintained by a not-for-profit organization
• The standard specification document is available either freely or at a nominal charge
• The intellectual property of the standard is made irrevocably available on a royalty free basis
• No constraints on the re-use of the standard.

Shortly after Ivan’s talk I delivered my slides, and, as can be seen, I asked the audience (which included not only Ivan Herman, but also Steve Bratt, CEO of W3C) which of the following were open standards:

XHTML    Flash     PDF     Java     RSS 1.0     RSS 2.0     MS Word

There was universal agreement that XHTML was an open standard and Flash, PDF, Java and MS Word weren’t. But there was ambivalence over both versions of RSS - it may be based on XML - but there are uncertainties over the governance of the standards. As Phil Wilson commented:

RSS 2.0 is published but there is a strict copyright on the usage of the term, and no-one apart from Dave Winer may make changes to it (“Someone has to have the last word, and when it comes to the RSS 2.0 roadmap, that’s me”, “I am banging the gavel”).

Related concerns have been raised over the future development of RSS 1.0. I would argue, therefore, that RSS 1.0 and RSS 2.0 fail the EU’s definition according to the “will be maintained by a not-for-profit organization” - neither seems to being adequately maintained.

Tavis Reddick mentioned that “The question isn’t clear about what is the standard: the interface, the file format, the information model…? I took it to mean something like the data binding format.“

It should be noted that the EU’s definition of a open standards relates to the openness and the governance of the standard itself - it is silent on issues such as the openness of tools to support its usage.

What does this mean to policy makers, developers, funders and users with digital library development programmes? One response is “nothing, this is just splitting hairs”. But if our mantra is “Interoperability through open standards” then surely we need to have an agreed understanding and definition of ‘open standards’? Oleg Liber, Director of CETIS touched on this issue in his talk at the recent JISC-CETIS Conference. His slides reviewed CETIS historical involvement with educational technology standards:

• 1998-1999: Educational technology interoperability standards?
• 2000-2005: Educational technology interoperability standards!
• 2006-2011: Educational technology? Interoperability? Standards?

From the certainties we held at the start of the new century, we are now beginning to challenge some of our basic assumptions.

“Interoperability? Standards?” or “Interoperability! Standards!”? What is your view?

Merry Christmas

Brian

Reflections On 2006 - Standards

This year has seen UKOLN building on its work on ways in which standards can be used to support digital library development programmes. In the past a simplistic approach had been taken, which assumed that standards developed by trusted standards bodies (W3C, IETF, ISO, etc.) would become widely accepted within the market place. This approach, however, has failed in the past (Coloured Book software anyone?) and we currently are seeing a wide range of debates over standards across the Web development community (Web Services Considered Harmful; RSS vs RSS vs Atom debates; the Semantic Web vs the lower case semantic web and microformats and, more recently, a radical vision for the future based on XHTML 2.0 vs an evolutionary development towards HTML 5.0 - as described in Molly Holzschlag’s Blog) .

UKOLN’s contribution to the debate has been the development of a contextual three-layered approach, based on a neutral standards catalogue (containing details of standards, their governance; their maturity and a risk assessment) together with policy layers for selecting relevant standards and for managing non-compliance with the policies. This approach, which has been designed to provide a level of flexibility which is needed in a rapidly changing technical environment is supported by an advocacy strategy (which promotes the benefits of open standards) and an iterative feedback and development approach (in order to learn from patterns of best practices).

We have sought to develop our ideas and gain feedback by papers which have been submitted to a number of peer-reviewed conferences. In May a paper on A Contextual Framework For Standards was presented at the “Workshop on E-Government: Barriers and Opportunities” which was co-located with the International World Wide Web Conference held in Edinburgh.

My colleague Marieke Guy has been engaged in implementing the system which is based on our contextual model. It was pleasing when Marieke and I met with members of the eReSS project to discover that they had taken a similar approach in the area of e-science standards.

This contextual approach has been designed to be usable by the wider community. The information provided in the standards catalogue has a Creative Commons licence associated with the entries, so there should be no legal barriers to the reuse of the content. This will enable developers, policy makers, managers, etc. within institutions to make use of the resources to support institutional development activities. More importantly from a JISC perspective, the approach can be used by JISC’s partners in the Strategic E-Content Alliance (SEA). The SEA is an alliance of bodies such as JISC, MLA. BBC and Becta, which aims to provide seamless access for the public to a wide range of scholarly, cultural and educational resource. The contextual approach to the selection and use of open standards is particularly relevance in this content as, though the bodies will seek agreement where possible on relevant standards, there will be areas in which organisational or political considerations may outweigh technical factors.

Next year will see UKOLN continuing to build on this work - and we are particularly pleased that a paper on Addressing the Limitations of Open Standards has been accepted at the Museums and the Web 2007 conference.