Web Focus: Report on "Institutional Web Management Next Steps" Workshop

brian kelly

Web Focus: Report on "Institutional Web Management Next Steps" Workshop

Brian Kelly reports on the latest "Institutional Web Management Workshop," this year called: "The Next Steps."

The "Institutional Web Management: The Next Steps" workshop took place at Goldsmiths College, London on 7-9 September 1999. This was the third annual event for institutional web managers which has been organised by UK Web Focus. The first workshop was held over 2 day (16/17 July 1997) at Kings College London. As described in the workshop report published in Ariadne [1] the event attracted a total of 95 participants. The workshop provided a valuable opportunity for web editors to meet their peers at other institutions and compare experiences. Although the event proved popular, a number of delegates commented that future events should last longer.

In response to these comments last year's workshop, which took place at Newcastle University, lasted for 3 days, from 15-17 September 1998. The extra time enabled more speakers to be timetabled, but, more importantly, enabled a half a day to be allocated to discussion group sessions, in which issues such as Web Design, Web Server Management, Metadata and Management Issues could be discussed in more detail. As described in the workshop report published in Ariadne [2] the analysis of the workshop evaluations forms showed that the changes were very appreciated: the overall rating for the event was 4.45 (on a scale of 1 [for poor] to 5 [for excellent]). Interestingly a number of participants requested even more time for participative sessions.

This year's event had the title "Institutional Web Management: The Next Steps" - to emphasise that it is building on the previous events. It had a similar format to last year's workshop (presentations on the first afternoon and the morning of the second day, with parallel sessions on the second afternoon). An innovation this year was the exhibition, which featured commercial vendors, and, from the Higher Education community, national services and projects and displays and demonstrations from the delegates themselves.

This year's event also included a popular social event - a boat trip on the River Thames to view the Millennium Dome. The social event provided an opportunity for the delegates - many of whom have encountered each other electronically on the website-info-mgt Mailbase list [3] - to put a face to the email address.

The Workshop Participants

A total of 124 delegates registered for the event. The majority (almost 100) were from UK Universities and University Colleges. In addition there were ten delegates from JISC services, two from overseas universities, one commercial delegate (one of the speakers) and fourteen from other organisations (including UCAS, CSU, the British Council, the Natural History Museum, the National Maritime Museum, the British Film Institute and LASER).

The Presentations

There were a total of eleven formal presentations during the plenary sessions. A brief summary of each of the talks follows.

1 - Who Will Rid Me of This Troublesome Beast?

John Slater, the Pro Vice Chancellor for Learning and Teaching at the University of Kent at Canterbury gave the opening talk [4]. At previous workshops have made recommendations that senior management needs to be made aware of the strategic importance of the web and the need for appropriate levels of awareness. John Slater's talk responded to this by arguing that senior management have become aware that the web manager is being asked for additional financial resources, but corresponding savings elsewhere are not being offered.

John had been asked to open the workshop by giving a provocative talk. He succeeded in doing this - and helped to ensure that the opening talk provoked many questions from the floor. John argued that:

Web Editors are misfits - but that's not necessarily a bad thing, as misfits are needed at periods of radical changes within organisations.
Web Editors are located in a wide variety of departments within organisational structures. Again, at a period of major changes, this is probably no bad thing.
Wed Editors probably miss-sell themselves if they simply ask for more resources because: they are over-worked; they are stressed; we simply must make use of XML/RDF/etc.; the University of Poppleton provides X (e-commerce, personalised services, etc.) and we have to as well.

After teasing and insulting the audience (especially the librarians) John reminded us that we have been here before. In the 1950s it was felt that the UK only needed three mainframe computers. In the early 1980s some Universities felt that networking was almost "solved" - and so there was no need to employ network staff on permanent contracts. A similar scenario existed for PC support in the late 1980s. Within the web area, we have heard predictions that technologies such as Java will make things better and that web technologies are about to stabilise. History tells us that we should expect to see the position of web editor be treated seriously (i.e. permanent posts on realistic salary scales) within institutions. John concluded by arguing that web editors should be proactive within their institutions. Recent funding initiatives in areas such as teaching and learning provide a useful opportunity for web editors to make a case for additional resources in areas of clear strategic importance to institutions.

2 - Building Relationships Online

Joe Passmore, head of Corporate Marketing at the University of Ulster, built on some of the issues raised in the opening talk [5]. Joe argued that for many university web sites it was unclear what value is added to the university's main activities. Many web sites exist because it can be done or because everyone else has one. We should be building web sites based on customers expectations and requirements, and we should find out what the customers want by asking them.

The term Relationship Marketing has been used to describe this concept. Properly applied relationship marketing can be used together large amounts of data about customers, which can be used to develop personalised web sites. We are familiar with personalised interfaces in many spheres - for example when you visit your bank (especially telephone banks such as First Direct) you aren't always asked if you'd like to open a new account - instead you are told about products you may be interested in based on your profile. Personalised web sites would enable simpler (and appropriate) interfaces to be developed - a confusing array of icons labelled "Click here for New Students"; "Click here for Existing Students", etc. can be avoided.

Examples of personalised web sites exist in the commercial world. Many of these provide not only web interfaces, but also deliver personalised email messages - as users of Amazon are probably aware.

Within the academic world, personalised web interfaces have been developed at UCLA [5], Miami [6] Washington State University [8].

Figure 1: The UCLA Personalised Interface

Although relationship marketing may be expensive or difficult to implement today, we can expect to see a move from the experimental services available today to more mainstream use in the near future. There are many challenges: technical, managerial and legal (such as the Data Protection Act). However we need to have an idea of the direction in which we wish to travel in order to get there.

3 - Beyond Brochureware - Building Functional University Websites

David Christmas and Ian Roddis of the Open University gave the final talk of the first day [9]. It is summarised here as it follows on from Joe Passmore's talk.

The speakers argued that four classes of web sites could be identified:

Indie Sites
Webmaster Sites
Embedded Sites
E-Business Sites

We will all be familiar with indie sites. Although "vanity pages" are typical of indie sites, they can also used to provide useful information. Indie sites are created by individuals. Their target audience may be unclear. Many indie sites make use of a variety of technologies, such as Flash and Shockwave.

Webmaster sites are often developed under the guidance of a committee. They typically provide a corporate service. They tend to avoid use of plugin technologies, and typically consist simply of HTML pages containing images. University corporate home pages are typical of Webmaster sites.

Embedded sites also tend to provide corporate information services. However they make use of backend databases. They are technically more sophisticated than webmaster sites. Figure 2 gives an example of an embedded website.

Figure 2: Open University Web Site

An e-business website uses the web as an essential component of the institution's corporate functions. An organisation has to reengineer its information flow processes in order to implement an e-business web site.

The Open University are moving away from Webmaster web sites and towards embedded and e-business web sites. A key point about this kind of site is that the embedding is not purely a technical issue: it has to be integrated into an organisation in terms of functionality, managerial involvement, aims, objectives and audience. It must also be recognised that developing sites of this kind will change an organisation.

4 - Multimedia and the Corporate Web

The first day of the workshop included two talks on the next generation of web services we can expect to see (personalised web sites and web sites which reflect institutional business requirements). In addition there were two talks which described aspects of multimedia.

Greg Newton-Ingham, UEA gave a presentation of the potential for multimedia on institutional web sites [10]. Greg began his talk with a video clip from the latest Star Wars movie. Greg pointed out that this was the most popular video clip ever downloaded from the web. This is an indication that many people who have an interest in Star Wars (including teenagers who are likely to go to University) probably have PCs capable of displaying this type of multimedia. They will have high expectations when they go to University.

Demand for multimedia on corporate web sites is probably not driven by web editors or their main customers. Instead it is probably led from top-down initiatives such as the JISC Moving Image Pilot Project and the CEI Working Group on Moving Images.

The development of services based on multimedia faces a number of challenges including technical challenges (computational power, disk storage and network bandwidth) and organisational and legal challenges (in particular rights issues).

5 - SMIL and the World Smiles With You

Michael Wilson, Rutherford Appleton Laboratories, followed on from Greg Newton-Ingham's overview of multimedia, to describe one emerging new technology [11]. SMIL (Synchronised Multimedia Interchange Language) is a W3C Recommendation. A SMIL player will enable a variety of components to be played. For example one window may display a video clip. After a specified period of time, a second window may be opened. One of the windows may be accompanied by a synchronised caption (perhaps for the deaf). Each of the windows may have hypertext links associated with areas on the screen. However the link areas may change with time.

A number of companies have already developed SMIL authoring tools and SMIL players - some of which are available free-of-charge.

Unfortunately Microsoft are not supporting the SMIL recommendation. They have proposed a solution known as HTML+Time. HTML+Time is supported in Internet Explorer v5.0.

SMIL is a declarative language, and is easy for novices to master. HTML+Time, by contrast, is procedural and programmers will find that it is extensible.

W3C are currently looking into extensions for SMIL v 1.0. We may find that a future version of SMIL merges the best of SMIL with HTML+Time. However for now the uncertainties will make it difficult for institutions to commit significant resources to the development of synchronised multimedia services based on use of SMIL or HTML+Time.

6 - Browser Management

Brett Burridge, University of Essex, began the second day with a talk on Browser Management [12]. Brett argued that as we have policies on the provision and support for many software packages (e.g. word processors, spreadsheets, etc.) we should introduce similar policies for web browsers. He quoted research from Zona Research showing that increasing numbers of commercial companies have introduced browser policies over the past two years.

Institutions also need to monitor the browsers accessing institutional web sites. This can help to inform browser policies, and also help web developers and designers in choosing appropriate architectures for their services.

An analysis of the Essex website during four days in August 1999 showed that the most popular browser was Internet Explorer (62% of local users and 66% of remote users) and the most popular operating system was Microsoft Windows. These figures were very similar to ones published by commercial companies.

Brett then reviewed the main browsers which are available. Internet Explorer and Netscape will be familiar to most. The Opera browser is small and has good accessibility features. However it is a licensed product and no administration kit is available. Mozilla is based on the Netscape source code. Although it seems to be rich in functionality it is rumoured to be buggy and its release date is uncertain. Other types of browsers include those available on PDAs, mobile phones and WebTV. It is not yet clear whether these are gimmicks or will have an important role to play.

Browser administration kits have a role to play in minimising the support and maintenance requirements. Both Netscape and Internet Explorer provide administration kits which are freely available. Brett argued that if you have a large number of browsers to support you will need a browser administration kit.

7 - Content Management Systems

In contrast to a number of talks, Stephen Emmott (KCL) gave a more theoretical presentation on content management systems [13].

Stephen argued that as web services move away from what David Christmas described as "Brochureware" and provide not only static information, but also multimedia resources and data supplied by backend databases, we need to move away from our concepts of information and to think about content.

Content is (or should be) authored, stored and published in a digital format. It can therefore be replicated, archived, distributed and re-purposed. Content management systems will help to provide these functions.

We need to think about the web site's unit of granularity. A file-based view can place limits on the functionality and maintainability of our services. It may be desirable to adopt an object-oriented view of our web sites.

Stephen concluded by arguing that we need to think deeply about the architectural model for our web sites, the standards we use (and Stephen felt that XML and RDF are probably the key standards) as well as the software products we use.

8 - Experiences with XML: Beyond the Hype

James Currall (Glasgow) described how XML is being used at the University of Glasgow [14]. He described the limitations of HTML and how XML can be used in specialist applications such as Maths. He then went on to describe a more mainstream application - a committee document system.

There is a need to provide a long-term archive of decisions made within the University. XML is the ideal format, as it is open and extensible. An MS Word "wizard" has been developed so that authors will not have to master a new package. The wizard makes use of MS Word styles. These can be converted to XML tags, such as <ACTION&;gt;. The structured information enables new functionality to be developed - such as looking for actions.

In his presentation James demonstrated use of XML in his slide show. He uses a home-grown system (based on XML and XSLT) for storing his slides. These are converted into HTML and CSS. The output was very professional, and the multiple-hierarchical interface provided a useful feature which is not easily achieved in PowerPoint.

James concluded by arguing that XML was an ideal format for use as an information repository; for application to application transfer and for display natively in the browser. Currently only IE 5 has good support for XML, although Netscape 5 is expected to. Until such browsers are widely deployed we can use XML as an information repository, and convert to HTML.

9 - Web Security

Andrew Cormack (UKERNA) reminded the delegates that we still need to think about the security of our web sites [15]. Since October 1997 there have been eight CIAC bulletins about the security of the Solaris operating system and eight about Windows NT. Andrew concluded his talk by reminding us we should that aim to keep the web server host computer simple: don't install software that isn't needed; keep up-to-date with patches and pay attention to log files.

10 - Indexing Your Web Server(s)

Helen Varley Sargan (Cambridge) described approaches to providing local searching across web sites [16]. Her talk was informed by the survey of UK higher educational institutional search engines published elsewhere in Ariadne. She asked if users of Excite (used at 19 institutions) were aware of the Excite security concern and that its development seems to have ceased. Helen also pointed out that the free version of Muscat (used at one institution) is no longer available. She asked us all if we were aware of how our sites would be indexed if we didn't make use of the robots.txt file.

Helen listed some of the requirements to consider when looking for an indexing tool (e.g. platform; numbers of servers to index; number of resources to index; file formats to be indexed; extent of use of dynamic data; type of search interface required; etc.)

Helen pointed out that possible solutions included free indexing services, free indexing software and commercial indexing software. She then gave three case studies:

Essex: Uses Microsoft indexing software on Windows NT platform. Indexes 16 servers and about 11,500 resources (including MS Office files and HTML). Incremental updating is possible. Free with Windows NT. Requires NT expertise to manage.
Oxford: Uses ht://Dig indexing software on Unix platform. Indexes 131 servers and about 43,500 resources (including MS Office and HTML files). Incremental updating is possible. Free. Requires Unix expertise to set up and run effectively.
Cambridge: Uses Ultraseek indexing software on Unix platform. Indexes 232 servers and about 188,000 resources (including PDF and HTML files). Incremental updating is possible. Relatively expensive. Very easy to install and maintain.

11 - Getting Your Web Site Listed

Danny Sullivan (SearchEngineWatch) gave the final timetabled presentation [17]. Danny outlined techniques for getting your web site listed by search engines such as AltaVista and directory services such as Yahoo.

One simple technique for ensuring that a web page is indexed by a search engine is to use the service's "Submit A URL" page. Typically the page should be indexed from within a few days to two months. A number of search engines allow multiple URLs to be submitted using an email interface. Go allows hundreds of URLs to be submitted, whereas HotBot allows up to 50 submissions per day. AltaVista allows 5 submissions per day.

Within a University context, however, it is probably unlikely that such submission services will be used to any significant extent. We will normally rely on a search engine to crawl our web site. Search engines normally index about 500 pages from any one site, so it is desirable to split up a web site if the site contains more resources than this.

A simple but effective way to split sites is to make use of subdomains. For example instead of using <www.bath.ac.uk/maths/>, <www.bath.ac.uk/english/> etc. it would be better to use <maths.bath.ac.uk/> and <english.bath.ac.uk/>.

Frames should be avoided if you want your web site to be indexed, as many search engines cannot get past frame interfaces.

Dynamic web pages which include a ? in the URL (such as <http://www.nike.com/ObjectBuilder/ObjectBuilder.iwx?ProcessName=IndexPage&Section_Id=11200&NewApplication=t>) also act as barriers to search engines. You should look for workarounds (such as the Apache rewrite module).

Danny concluded by suggesting that interested delegates should visit his SearchEngineWatch web site.

The Parallel Sessions

A total of seven parallel sessions were held. A brief summary of the sessions is given below.

1. - Intranets and Extranets

This session was over-subscribed, indicating a clear interest in this topic area. The session was led by James Currall (Glasgow) who also gave a report on the session [18].

The session provided a useful opportunity for the 24 participants to share experiences. It was found, however, to be difficult to define the term Intranet, which added complexity to the session, as not everyone was talking about the same thing.

The session included a report on a MIMEO JTAP project [19].

The session identified that authentication was a big issues, and that intranets were more complex than traditional web sites. The recommendations of the session were:

Need for clarification of the term Intranet
Need for more events on this topic
Need to promote example of best practice
Need for guidance on authentication issues including CAs, digital certificates, smart cards, LDAP, etc.

2. - Web Site Navigation

Participants completed several exercises including an analysis of (a) UK HEI search engines and (b) 404 error pages (which were based on article published in Ariadne issue 20 [20] and 21 [21]).

The recommendations of the session were that participants would like:

One stop web site of useful information and briefing sheets
Rerun of session, possibly on regional basis

3. - Metadata

The session entitled Metadata: Has The Time Arrived? used the HE Mall (which will be a portal to UK Higher Educational institutions) as an example of a metadata-driven service to address several issues:

The need for the HE Mall.
The difficulties in providing structured information for application-application use.
Technical solutions.

The main recommendations from the session were:

HE Mall developers need to be aware of potential difficulties within institutions.
Institutions would like software to help manage their metadata.
Institutions would like case studies illustrating how institutions approach participation in the HE Mall.

4. - Design, Access and Accessibility

This session addressed the design of institutional web sites, and how design could help or hinder the accessibility of the web site. The main recommendations of this session were:

The W3C Quicktips card [22] proved very useful for authors and for training. The community should consider bulk distribution and printing distribution of these (centrally or within institutions).
Training for authors is important. It should not be treated as a separate issue to general design, but an integral part of the process. The RNIB/WAI video [23] is very helpful and could be used as a training resource.
New material/web interfaces should be designed with access in mind, and checked with the Bobby tool.
Training materials that raise access issues is vital.
Policy makers should provide funding so that accessible web material can be developed.
Web community needs advice on where funding may be available to redress poor design / access.
It is important for web managers to know what their institutions disability policy is, and who can help them.

5. - Legal, Decent, Honest and True

Session addressed three topics:

1. AUPs: You need one
If it's based on JANET AUP, remember it has changed
Think about your concerns (being sued, bad publicity, etc.)
2. Data Protection: Likely to be part of Web Editor's job; New act comes into force in March 2000
3. Copyright: New EU directive out soon; You should assert your copyright; You should trademark your logo; You should probably trademark name variants (e.g. oxford-university.com is up for grabs)

6. - Web Tools

This session looked at two topics: (1) Browsers and Browser Administration Kits and (2) Middleware.

Browsers and Browser Administration Kits:

Opera is lightweight and has good support for accessibility, but it's licensed, has no development kit and its Java support is poor.
Web pages may look OK in WebTV browsers under certain circumstances. If WebTV takes of it may have many implications for us.
Agreement that browser administration software is desirable, but there may be (political?) difficulties in implementation.

Middleware:

Zope is an open source middleware system which is used at Newcastle and Bristol [24]. Paul Browning gave a brief presentation about Zope [25].
Lycee is a commercial content management system [26]. It claims to release webmaster from mundane chores. It aims to enable non-technical people to publish. Lychee is under trial at Kings College London and Liverpool Universities, with the possibility of a CHEST deal.
Lotus Notes is used at City University Business School.

7. - Web Editor

This session was led by Miles Banberry. A report on the session was given by Damon Querry [27].

The session addressed four main topics: (1) Getting A Job; (2) Managing The Job; (3) being Supported and (4) Being A Professional.

The "Getting A Job" session was based on an analysis of jobs which had been advertsied within the UK HE sector over the past few years. It was found that there was a wide range of job titles and wide pay scales. This led to the conclusion that employers did not know what they wanted.

The "Managing The Job" session looked at the location of the web editor post within the organisation, and attempted to define the ideal web management team.

Having recognised that web editors were located in a range of depertments within instituional organisational structured, it was recognised that it was important to consider support infrastructures. The "Being Supported" session felt that there was a need for national accreditation, and for support to be provided through regional groups and mailing lists.

The "Being A Professional" session looked into national and international professional bodies which can help set standards for the profession. These included bodies such as the British Computer Society (BCS) and the Institute of Information Scientists (IIS).

The Exhibition

An exhibition was available on the third day of the workshop. The exhibition included commercial vendors of web software, JISC services and JTAP projects, together with contributions from the workshop delegates.

The commercial vendors demonstrated products and services which were relevant to many of the themes discussed at the workshop. They included Netcentric Solutions [27] (who demonstrated their Lychee content management system), Redleaf [28] (who demonstrated Netobjects and Colfusion web development products), Open Objects [29] (who demonstrated the Ultraseek indexing software), Highlander [30] (who demonstrated the Cold Fusion), Softquad [31] (who demonstrated the HoTMetal and XMetal authoring tools) and Thames Digital Media, a website design and production company.

In addition to the commercial vendors, the Institute For Information Scientists also provided a stand [32].

Over a dozen displays and demonstration from national services and projects, and institutional activities with the higher education community were provided. Further details are available [33].

Feedback

At the time of writing the workshop evaluation forms are still being analysed. The initial findings indicate that, on a scale of 1 (poor) to 5 (excellent) the overall content of the workshop was given a rating of 4.08 and the organisation was given a rating of 4.35. Seventeen people gave the content a top rating of 5, with 32 people rating the content at 4 and 12 at 3. The ratings for the organisation were even more impressive, with no fewer than 30 scores of 5, 24 of 4 and 8 scores of 3. There were no rating of less than 3 for the overall content or organisation.

The organising committee is very pleased with this feedback, which reflects the hard work and dedication< of the members of the organising committee, the speakers and parallel session leaders, and, of course, the enthusiasm which was shown by the conference delegates.

A more detailed report on the evaluation will be published shortly [34].

Conclusions

The workshop clearly succeeded inits aim of addressing a range of new issues of concern and interest to the web management community. Plans are already in progress for repeating the workshop next year. The recommendations from the parallel sessions and other comments included on the evaluation forms will be analysed in order to identify and prioritise further activities. Many of the topics raised are likely to be discussed on the website-info-mgt Mailbase list [35] which provides the main mailing list forum for the UK HE web management community.

References

Author Details

Picture of Brian Kelly Brian Kelly
UK Web Focus
UKOLN
University of Bath
Bath
BA2 7AY

Email: b.kelly@ukoln.ac.uk

Brian Kelly is UK Web Focus. He works for UKOLN, which is based at the University of Bath