The Evolving Web
A Pace Layering view of the development of the Web and the W3C
Roger Hudson
19 March 2008
SUMMARY
The Web today is very different to what it was in 1994, when the World Wide Web Consortium was established. This article looks back at the early history of the Web and uses Pace Layering theory to consider some recent developments and the role of the W3C in the process of change.
Perhaps the Membership and structure of the W3C, which were highly effective in establishing the Web during its first decade, are no longer able to adequately reflect the diverse interests and aspirations of the greatly increased number of Web users today. Also, perhaps it is time for the W3C to break away from its financial dependence on large corporate Members and to start taking a more active role in regulating the Web.
This article draws on the keynote presentation "Web 2.0 and the Concept of Universal Design" by Roger Hudson and Russ Weakley at the OzeWAI Conference in December 2007.
1. INTRODUCTION
Undoubtedly, the World Wide Web has changed significantly during the last few years with the explosion in the number of social networking sites one obvious manifestation of this. At the same time, the long, and some might say tortuous, process of developing Version 2 of the Web Content Accessibility Guidelines (WCAG 2.0) has ground on and acrimonious debates have swirled around some of the output of the Working Group considering HTML 5. All of this has caused me to look back at the evolution of the Web and the role of the W3C. And, in the process I found myself increasingly looking at the development of the Web from a Pace Layering perspective.
2. WHAT IS PACE LAYERING?
The first use of the concept of Pace Laying, that I am aware of, is by Steward Brand in his book "How Buildings Learn" that was published in 1994 [1]. He suggested a building could be divided into "layers" and the rate at which the layers change varies. At the bottom, or first layers, such as the building site or the structural core of the building, the rate of change is slow; whereas at the top layer, which incorporates things like the paint, fittings and furnishings, the rate of change is fastest.
("How Buildings Learn: What Happens After They're Built" by Stewart Brand)
Key elements of the Pace Layering view are:
- The lessons learnt and changes that occur at the outer layers, where change happens the fastest, filter down to the inner layer where change happens slowly
- The differing rates of change in each layer are likely to cause friction between the layers
- The different rates of change, and resolving the resulting friction, are beneficial to the overall development of complex systems
"The fast parts learn, propose, and absorb shocks; the slow parts remember, integrate, and constrain. The fast parts get all the attention. The slow parts have all the power." [2]
Steward Brand, The Long Now Foundation
A Pace Layering perspective can be particularly useful for people who build things; be it a multistorey skyscraper or a content management system, since it can help us, Avoid such classic mistakes as solving a five-minute problem with a 40 year solution, or vice versa
. [3]
Over the years, Pace Layering has evolved into a broad concept to help us understand the process of change. Pace Layering theory has been applied to a wide range of situations including, the study of human society and the environment, information architecture and user-centred design.
3. WEAVING THE WEB
I have deliberately used the title of Tim Berners-Lees' book as the heading for this section, since both the book and this section are primarily concerned with the early history of the Web. By the way, for those who haven't read "Weaving the Web: The Original Design and Ultimate Destiny of the World Wide Web", it's a great read, providing an interesting first-hand account of the conception and growth of this amazing communication tool.
Although the Internet and Web are sometimes viewed as being synonymous, the Internet had been around for about 20 years before the Web emerged. I suppose we could consider 1969 as the start date for the Internet, for this was the year 4 host computers joined ARPANET, which had been recently developed by the US Advanced Research Projects Agency (ARPA).
Tim Berners-Lee is often referred to as the inventor of the World Wide Web and certainly, in my view, no one else can make a greater claim to this honour. However, we should not overlook the work of others who laid the ground work, not least Doug Engelbart and Ted Nelson who developed the idea of hypertext and Robert Cailliau who worked for many years with Berners-Lee on developing a system for networking computers at CERN. The Berners-Lee 1989 CERN proposal, "Information Management: A Proposal" became in effect the original proposal for the World Wide Web. [4]
The first Web server was at CERN in Switzerland. By 1991, it was receiving requests for about 10 – 100 page views a day, so Tim Berners-Lee and Robert Cailliau submitted a proposal to present a paper on their work at the Hypertext '91 Conference in San Antonio. The paper was rejected, but they were allowed to set up a demonstration in the exhibition area. At the 1991 conference their display booth was the only one associated with the Web, within two years however, every display at the Hypertext conference was related to the Web and the CERN server was getting more than 10,000 hits per day. [5]
In April 1993, CERN agreed to have the underlying Web technology put in the general public domain, in effect allowing anyone to use the Web protocol and code to do with as they like free of charge. For many years, I have felt that this inspired act of openness and beneficence symbolised the potential and power of the new communication medium. In recent times however, I have begun to have some doubts and wonder whether this move, as important as it was to the early development of the Web, may have also been flawed. And, sometimes the flaw that adds the glittering sparkle to a jewel may also be a weakness that can become a fracture line.
4. W3C
In 1994, the World Wide Web Consortium (W3C) came into existence with Tim Berners-Lees as Director. Putting together a Consortium that contained competing companies in the emerging Internet business would have been no mean feat, but in the first few years of the Web there were worrying signs that unless some agreement on how to proceed could be found the dreams of a world wide web might not be met.
At this time, the Web was still on the periphery of the computing and networking industry sector and few imagined the growth that would occur over the next 10 years. Members of the World Wide Web Consortium in the early days were in the main large businesses from the IT sector and a few academic institutions. As Members, they were entitled to be represented on the different W3C Working Groups that were concerned with developing web-related standards or, in W3C speak, "Recommendations", such as HTML, CSS and sometime later WCAG.
"The representatives came together with concerns over the potential fragmentation of HTML. This was seen as a huge threat to the entire community. There were so many proposed extensions for HTML that a standard really was needed… Indeed, the true art for the consortium would be in finding the minimum agreements, or protocols, everybody would need in order to make the Web work across the Internet. This process did not put the consortium in a position of control; it was just providing a place for people to come and reach consensus." [6]
Tim Berners-Lee
In Pace Layering terms, the actions of the W3C and its Working Groups back then were not yet at the core of the computing, or even Internet, community. The notion of the Web and the work of the nascent W3C were swirling around the surface of a Pace Layering model of the computing industry; the outer layers where new ideas emerge and the rate of change is fastest. Although I don't know it for a fact, it would seem reasonable to assume that in those days, the Members developing Recommendations would have resisted any attempt from the computing or networking "establishment" that tried to force them to adopt a particular protocol, approach or proprietary product – for example VAX/VMS.
The Web matured quickly and was soon at the core of the Internet industry and community with a resulting change in the position of the W3C. Today, the W3C has over 400 members and although any organisation or individual can apply to become a Member of the W3C, Membership still mainly comprises technology related businesses including many large companies with a direct involvement in the Web industry such as Microsoft, Google, Adobe, IBM, Sun etc.
About 85% of W3C Members are from the developed 'first' world with nearly 40% from the US. The restricted nature of the membership is perhaps not surprising given the relatively high cost of becoming a member. For example, the annual membership fee for very large companies from countries like Australia, America and New Zealand is US$63,500. For medium and small businesses, government departments and not-for-profit organisations in these countries the fee is $6,350. Even for a not-for-profit organisation in a poor country like Cambodia the fee is nearly $1,000.
There is no real provision or fee structure for individuals to become members of the W3C. And, while individuals can be invited onto Working Groups as "Experts", the requirements for continuing participation very heavily favour representatives who are able to get the organisation they represent to cover the sometimes not insignificant costs in terms of time and travel.
The W3C's stated mission is, To lead the World Wide Web to its full potential by developing protocols and guidelines that ensure long-term growth for the Web
[7]. A mission that the Consortium primarily pursues through the creation of web-related recommendations and guidelines: More than 110 since 1994. However, the reluctance on the part of the W3C to exert any control over the operations of the Web could undermine these efforts and ultimately the relevance of the Consortium. It is likely that at the end of 1994, a far higher proportion of the people using the Web, or putting content on the Web for that matter, were aware of the W3C than is the case today.
I'll venture to suggest that the vast majority of website proprietors and developers today are not aware of the W3C and when it comes to standards, they just want their site or application to work for them and their clients; and they are not in the least bit concerned with noble ideals like "Web interoperability" and "accessibility". I recently used the W3C Markup Validation Service to check the validity of the page code for the Home page and one internal page of 5 large web-related organisations who are Members of the W3C. Of these 10 pages, 7 failed to comply with between 10 and 250 errors per page identified. If after more than a decade active W3C Members can't or won't comply with W3C Recommendations, what hope the rest of us?
5. THE WEB TODAY
Berners-Lee originally conceived of the Web as medium with little or no barriers between the providers and users of content; a two-way street that anyone could use to share the material of others or put up their own material for others to share. In fact, as we know, it wasn't long before the traffic on Web Street was largely one-way (similar to the long established media): A place for the relatively small number of people with the knowledge and/or capacity to put material for everyone else to access.
The proliferation of social networking sites over the last few years however is helping to return the Web to the original notion of a two-way street, where a 10 year old skateboarder and a media mogul are both able mark-out some territory and, with relative ease, present their content for the rest of the world to share. It was this growth in the use of social networking sites that partly stimulated me to look at the development of the Web and the role of W3C from a Pace Layering perspective.
5.1 Pace Layering model
The following diagram presents a Pace Layering model of the Web today. It contains 6 development layers which are ordered depending on what I perceive to be the likely rate of change in each layer. I have also indicated where the some of the demands or pressures for change in each layer might come from and offer some examples of the outcomes.
I have placed "The Internet" at the bottom, or inner, level with the slowest rate of change; of course I could have started lower, at the emergence of computers or mathematics for example, but the Internet seemed like a good starting point.
At the top, or outer, level with the fastest rate of change are "Users"; in essence people who are looking for new ways to find things or to share their ideas, the weird and wonderful, with the world.
| Development Layers | Demands | Outcome examples | Rate of change |
|---|---|---|---|
| Users |
|
|
FAST |
| Developers / clients |
|
|
|
| Technology |
|
|
|
| Standards and Guidelines |
|
|
|
| The WWW |
|
|
|
| The Internet |
|
|
SLOW |
In common with the general Pace Layering view, differing demands and rates of change in each layer can result in friction or tension between the layers. Also, the outer most layers is where many new ideas are proposed and tried out, and as these filter down through the layers, problems are identified and changes often made with the ultimate aim of influencing the core.
In terms of the Web, the W3C is now at one of the lower (core) layers of the model where the rate of change should be slow. As new ideas for the Web emerge in the upper layers where the rate of change is the fastest, they should filter down to the lower layer where the power lies. The W3C however, which is at the layer responsible for providing stability and continuity, must also be responsive to the upper layers that innovate, learn and propose.
5.2 AJAX and ARIA – a Pace Layering view
Let's look for a moment at the emergence and development of AJAX, which allows developers to introduce new content to a web page by re-writing sections of the DOM. The impetus for its development occurred in the outer layers of the Pace Laying model with the increasing demand by Web users for improvements in the user-experience and greater interactivity. In response, Jesse James Garrett and others at the Developers Layer brought together some established Web technologies in 2005 and popularised the term AJAX. AJAX was soon being used in a wide range of applications – Google maps being one of the best known examples.
One of the interesting things to come out of the move to AJAX has been the relative speed of the response by screen reader manufacturers, who are at the Technology level of the proposed model, to the growing use of AJAX. Since it is JavaScript dependent, AJAX per-se could be considered inaccessible under WCAG 1.0 requiring the inclusion of an accessible alternative and, to the best of my knowledge, no screen reader in 2005 supported AJAX. Eight months later however, a new version of the most widely used screen reader, JAWS, was offering limited support and with each subsequent version the level of support has improved.
In a similar vein, the need for improved accessibility of dynamic Web content for people with disabilities quickly filtered down the layers of the suggested model. In 2006, Firefox implemented an Accessibility Rich Internet Applications (ARIA) extension which would allow more complex navigation systems like tree-views and the content in live updated sections of a page to be more accessible. Members of various W3C Working Groups became actively involved in the project and by 2007 the WAI-ARIA Roadmap was a draft Recommendation of the W3C.
It must be noted however, that the proposed WAI-ARIA Roadmap is not without controversy, in part because some of the suggested techniques do not conform to current specifications or standards for HTML. However since standards are promulgated at a relatively low level of the suggested model, where the slower rate of change allows for greater reflection, we can hope that these differences will be resolved in a stable and enduring way.
Within the Web community the process of considering and accepting the development of AJAX and ARIA has been relatively well ordered. It has also been in general accordance with key elements of the suggested Pace Layering theory of change, namely; ideas are generated at the outer, faster layers, and as these filter down through the various layers to the inner layers, which change slowly and provide stability, the resistance or friction that new ideas often face are addressed and the overall level of support increases.
5.3 Driven change meets resistance
This ordered process has not always been evident. Consider for a moment the disagreement, some might say hostility, which has surrounded some of the output of the Working Group that is considering HTML 5. For example, the Working Group's desire to remove a number of accessibility-related attributes from the specifications concerning data tables, in spite of the fact that these attributes are supported by most assistive devices, and many developers, users and accessibility advocates argued for their retention.
In this case, the stimulus for change appears not to have come from one of the out layers of the Pace Layering model that are usually associated with proposing and testing ideas: But instead, we see radical change being driven by an inner layer that would normally be associated with evaluating ideas, and where necessary constraining them so that they can be effectively integrated into the system. As a result, there has been little incentive, or opportunity, to obtain the support of the people at the outer layers of the model, who have to implement and live with the proposed changes.
Similarly, some of the difficulties associated with finally resolving the development of WCAG 2.0 are, in my humble opinion, related to an apparent "we know best" attitude by the Working Group that has resulted in the Group trying to drive radical change such as the introduction of "Baseline" out from the core. Needless to say, various people and organisations associated with accessibility at the outer layers of the Pace Laying model felt major change was being foisted on them without proper consideration of their opinions or needs. I'm sorry to say that I suspect many people feel that the current W3C process of providing comments to draft recommendations is not an adequate forum for them to express their ideas.
5.4 Social networking sites
One of the distinctive features of the social networking movement is a redressing of the balance between the providers and consumers of Web content. Unlike the past, when the provision of content was largely in the hands of relatively few, now many millions of Web users are putting their own content on the Web via sites like Flickr and Myspace, sharing their interests through Delicious and Digg or using a Wiki to collaborate on content.
When it comes to the development of social networking sites, it seems to me that most of the action has occurred at the two outer most layers of the suggested Pace Layering model; the layers where users and developers are responding to the demands by many for an improved and more inclusive user experience. And, it appears to me that many of the providers of these social networking services and those who use them are not particularly concerned about either the W3C or their Recommendations.
When preparing this paper, I also used the W3C Markup validator to check the code of the Home page and Internal page of twelve of the most popular social networking sites. On only one of these sites, Wikipedia, the two checked pages were found to be valid, while the validation service reported between 11 and 256 errors for each of the 22 pages from the other sites.
I would have expected that the social networking phenomenon, with all its potential and problems, to have filtered down the Pace Layering model by now, however it appears to be of little interest or concern to the W3C. For example, the Last Call Drafts of WCAG 2.0 do not address the issue of the accessibility of user-generated content on social networking sites. I don't know why this should be the case but perhaps it might be because the current Membership and structure of the W3C, with its bias towards the now-established corporate Web players, does not adequately reflect the interests and aspirations of the users of these new Web services.
6. WHAT FUTURE THE W3C?
Before beginning this concluding section, I would like to clearly state my respect and admiration for the W3C, its Working Groups and staff for the work they have done over the years to shepherd in this great communication revolution. I know in this essay I make some critical comments about the W3C, but please understand my aim is not to mindlessly knock the organisation, rather I just wish to offer a personal, and hopefully constructive, view on what I fear are some emerging problems and make a few suggestions for the road ahead.
The Web was originally the domain of a few mathematicians, scientists and enthusiasts; basically a benign, if rather dull, environment of goodwill and good intentions. Today, it is a dynamic, exciting world used by hundred of millions of people everyday, but not all with the best of intentions.
The W3C, which has historically been at the core of the Web, sees its main role as ensuring the long term growth of the Web through the development of protocols and (technical) standards but is unable or unwilling to take any responsibility for how those standards are adopted or how the Web is used.
I have no doubt that many of the corporate Members of the W3C back in1994, fully supported the notion of the W3C being an overarching organisation that would do the work of developing and promulgating the Web; but an organisation that had no regulatory power and was not financially independent. I am also sure, that "freedom of the press" and "leave it to the market" types of arguments were made around the table in support of this notion.
6.1 Representing the Web Community
As we know, the Web today is very different to the Web of 1994. Since 1994, the number of Web Users has grown from about 13 million to about 1.3 billion, and as a percentage of the world population, from less than a quarter of one percent to 20%. At the same time we have seen a significant shift in the demographic make up of Web users and a huge increase in the commercialisation of the Web. Perhaps a W3C, which was able to reflect the composition and aspirations of the Web in 1994, needs to also evolve if it is to represent the Web in the new environment of 2008 and beyond.
The Pace Layering model of development outlined in this article recognises the friction that is likely to occur between the different layers, largely as a result of the differences in the rate of change. But, for this model to succeed there also has to be the free exchange of ideas between the layers.
Rather than restricting the development of Recommendations to a largely exclusive club, maybe it is time for the W3C to look for ways to open up the process by providing greater incentive and opportunity for participation by members of the general Web community. Maybe it is also time to reconsider the Membership structure of the W3C and the way the Working Groups operate.
In my opinion, any move to improve participation will need to involve a genuine exchange of views. The W3C needs to provide the wide range of people with an interest in the future of the Web and/or the Consortium with more than an invitation to offer comments on draft Recommendations or the opportunity to listen to an occasional lecture or presentation. Maybe the W3C could learn something from the social networking movement in this regard.
6.2 Someone needs to be responsible
The free availability of the protocols and codes that underpin the Web and the unregulated nature of the medium no doubt helped stimulate the astronomical growth of the Web bringing with it the many benefits we enjoy. However, these same conditions have also given us the scammers, spammers, paedophiles and other low life from around the world that seem to increasingly pollute the Internet and Web.
Sometimes, it appears that the only major areas of business on the Web that seem to be growing are those associated with crime, either in its commission or from those offering protection. Although a large amount of Internet or Web fraud goes unreported, the 2005 FBI Computer Crime Survey of US businesses found, Over 79% (of major businesses) had been affected by spyware and not surprisingly almost 84% had been affected by a virus attack at least one time within the last 12 months, despite the almost universal usage of Antivirus software
. [8]
The survey also found that only 9.1% of computer based crimes during 2005 were reported to legal authorities. Conservatively, the FBI estimates the total cost in the US to be about $67 billion each year. Now clearly not all this computer crime is directly associated with the Web, but a sizable proportion is; and if it continues to grow you have wonder how long the general Web user or the providers of online goods and services will tolerate it.
At a personal level, I get annoyed at being constantly barraged with requests to visit the site of some bank or other in order to confirm my account details, and emails selling me drugs, telling me I have won a lottery somewhere in Europe or inviting me to help some Nigerian liberate squillions of dollars that they are entitled to (often with a blessing from God). Needless to say all these enticements are ignored. I do however use my credit card on the Web, but I know many people who are now reluctant to do this through fear of credit card or identity fraud.
Thus far, the W3C has basically shunned the notion that it has any direct role to play in determining how the Web is used although it appears to be clearly concerned about the use of the Web for criminal purposes, consulting with different organisations and governments. The global nature of the Web is a key reason why crime has become such a problem. The perpetrators of crimes are usually geographically removed from their victims, often in different countries. While a growing number of countries now have laws relating to different Web crimes, pursuing offenders and enforcing those laws in a number of countries can present jurisdictional problems.
Perhaps it is time the W3C, and the Internet industry as a whole, took greater responsibility for how the Web is used. I am not suggesting we need some sort of Internet United Nations or Web police force for clearly individual countries are responsible for their own laws and how they are enforced. However, I find it incredible that with a global communication infrastructure, which has a finite number of identifiable regional nodes or hubs, it is not possible to track down the source of a phishing expedition or any other form of distributed crime. And when the offenders are located, access to the infrastructure should be denied to them and, just as importantly, to those who facilitate their nefarious activities.
Yes, I know the baddies can use proxies and Trojanized computers in other countries to mask their identity and make detection difficult. But not impossible I would venture to suggest, given the willingness of some entity to step up to the plate and take responsibility, and perhaps step on a few toes in the process.
6.3 The money
It seems to me that the key to the long term survival and relevance of the W3C is money. I have already indicated that I feel the W3C should strive to become an organisation that is more representative of the Web community today and an organisation that is to willing to exert greater control over the Web. No doubt, if these dreams are to be fulfilled it will take money and here is the rub, for I also believe that the W3C needs to become more financially independent of its Members, the major source of funds for the Consortium.
With the benefit of hindsight, perhaps it might have been better if back in 1993 CERN had decided to assign the underlying code of the Web to a responsible body like the W3C rather than making it freely available. If the W3C owned the basic IP rights of the Web it would now be in a much better position to exert some control over the way Web is used and it would also have a potential source of revenue that was independent.
But the past is the past, water under the bridge. In my opinion however, this still leaves us with a need for the W3C to reduce its dependence on the fees of Members for its operations. How is this to be done?
I feel that the most likely source of ongoing funding for the W3C, or any other organisation that is willing to take responsibility for the Web, will be a levy on financial transactions that are conducted over the Web. Credit card transactions would seem to me to be the obvious starting point. Without knowing the details, I would think that only a small fraction of a percent of each credit card transaction would be required to generate sufficient income.
But, what's in it for the credit card companies? Why would they be willing to give up even a very small percentage? In my view, the answer to these questions is directly related to my belief that the W3C has to take a more active role in reflecting the needs of the Web community today and in helping to regulate and control the Web. If the W3C was willing to do this, it would allow merchants, customers and the providers of financial services to conduct transactions over the Web with greater security and allow all of us to use the Web with increasing confidence.
Footnotes
- [1] Brand S., "How Buildings Learn: What Happens after They're Built", Penguin Books, New York, 1994
- [2] Brand S., "Cities and Time", The Long Now Foundation http://blog.longnow.org/2005/04/11/stewart-brand-cities-and-time/
- [3] Brand S., "The Clock of the Long Now", Basic Books, New York, 1999
- [4] Berners-Lee T., "Information Management: A Proposal", CERN, 1989/90 (W3C), http://www.w3.org/History/1989/proposal.html
- [5] Berners-Lee T., "Weaving the Web: The Original Design and Ultimate Destiny of the World Wide Web", Harper Business, New York, 1999
- [6] Ibid
- [7] "About the World Wide Web Consortium", W3C http://www.w3.org/Consortium/
- [8] "2005 FBI Computer Crime Survey", http://www.digitalriver.com/v2.0-img/operations/naievigi/site/media/pdf/FBIccs2005.pdf)