Search
AGOCG are pleased to tell you that Sue Cunningham from the University of Manchester has accepted the post of Multimedia Support Officer and Sue introduces herself in the Newsletter. I am sure that Sue will offer excellent support to the community and act as a focus for information.
This was held recently and a brief report is written in this Newsletter. The full report and other output will be available soon and wilJ be promoted through the Newsletter.
The Support Initiative for Multimedia Applications is funded under the JISC New Technologies Initiative. The Support Officer and the recent Workshops are part of that. In the last Newsletter I reported on the various projects which we have got off the ground. Although material will be made available online, I know that access to the printed versions of reports can be useful to people. I am therefore inviting you to subscribe to the output of the SIMA project. This subscription will cover the period of the project from now until July 1995 and will mean that you will automatically receive reports and other output. The cost of this to UK Higher Education is £50 pounds and an order form is enclosed with this Newsletter. The first reports are expected to be made available in early October. Sign up now!
A new technical report (number 25) is available from AGOCG. This is a survey of the use of Desktop Video carried out by the HUSAT Research Institute for AGOCG. Email J.T.Barradell@lut.ac.uk if you would like a copy.
I have included some new information from CHEST. I intend to make this a regular input to the Newsletter. I also have meetings with CHEST and if you wish me to raise anything with them let me know.
AGOCG has 3 user representatives on the committee which meets 4 times a year and acts as the steering group for the AGOCG activities including the SIMA project. We are looking to bring in one new representative this year and will be bringing in someone new each year. User representatives to date have been Ken Brodlie (Computer Studies, University of Leeds), John Lansdown (Art and Design, University of Middlesex), David Unwin (Geography, Birkbeck College). We are particularly interested in having someone on the group from arts, medicine or the social sciences. If you are interested then please contact me for more details.
Two days ago (when writing this report), 30 people met to discuss the subject of multimedia formats. In my introduction to the workshop I suggested that while many people might consider this to be a "crashingly boring" subject, it was very important. This report overviews the main topics of discussion and the recommendations for future work.
This workshop was part of the JISC New Technologies Initiative's workshop programme hosted by the SIMA (Support Initiative for Multimedia Applications) project which is monitored by AGOCG. A full write up of the report will be available as part of the SIMA report set.
The workshop discussed a number of different types of formats (images, documents, delivery formats) but we kept coming back to the need to recognise some form of taxonomy and to educate people about the issues relating to the selection of a file format. I think we accepted that in general many people using file formats did not want to have to care about what format they saved their work in, but that the selection might well affect what they could do in the future with their courseware, images etc. I think that it is most useful in this brief report to focus first on the issues and then move to the proposed pieces of work.
We very quickly came to the conclusion that the users were not of one type and that we had to consider different requirements. These include courseware authors, archivists (with indexing requirements), courseware readers, support staff concerned with reading and conversion of a range of formats - and no doubt many more. These different needs have to be recognised by any output.
If the data is to be stored and used within a short timescale then format issues are matters of convenience for the user. If, however, the information is to be archived for any length of time and widely used - or has the potential for this - then the selection of the right format is a real issue. It is essential that we can be confident that the format will be readable in the future, thus an "open" standard - and probably a de jure standard is appropriate. It is also important that the information is stored at the highest possible level of quality and that the information is not diminished by the storage method. Archiving images at high resolution, for example, is clearly preferable to storing them using a lossy compression method which will reduce information. The information delivery may be in compressed form, but the archive should maintain data quality.
This is clearly important in Higher Education where we seem to have at least one of any machine that can be named. This should not be ignored when developing courseware, for example, for wide use.
Archived information needs to be searchable and accessed in a range of ways - some of these we may not even have thought about. Some of the simple solutions of scanning text and images and keeping them in raster form does not allow access to as much information as we might like. This is particularly true of "legacy" data which may be available on paper now and which we wish to convert to online form the scanned virtual paper is often high in storage requirements and low on information.
This came up many times at the workshop. We need to get people to distinguish between the components of the application - for example courseware - and the glue that sticks them together. Sometimes the packages we are using do not encourage this way of thinking. Yet, if the end products are to have a long life and to be updateable it is important that this is considered.
This distinction is not always obvious in that some proprietary formats do have publicly available definitions. It is clear that proprietary formats are not necessarily bad and formal standards good. It is important that formats are suitable for the task. Open formats are however ones which people should be looking to use for long term storage. The danger of storing information in a way which depends on the supplier continuing to support a particular format can be fraught. Just as a format which is only supported on one platform can cause problems with the demise of that machine. The workshop agreed that we would keep in touch with the Interactive Multimedia Association (IMA) which is a consortium of major suppliers who look to develop and promote common interchange formats. This link might help us to get the balance right between proprietary and formal standards use.
This is an important area, particularly for support staff who have users with a wide range of formats which need to be converted for use in software and for output. There are good tools around (Utah Raster Toolkit, PBMPLUS, San Diego Supercomputer Center Tools) which can be accessed. We need to inform people about these.
This has come up before at the last AGOCG workshop on multimedia. The new technologies can offer a great deal for people with sight and hearing difficulties and we heard of the use of SGML in structuring information so that it can be presented for people with print disabilities. Yet, the overuse of visual information with no concern to store information in a way it can be represented in a variety of ways, can disadvantage those with sight and hearing difficulties.
It seems clear that we cannot say to people you should use a particular format in al1 circumstances. What o we want to do is to help people to make the right choices in particular circumstances. To this end we intend to recommend the fol1owing to AGOCG:
It is intended that we should start work on the reports very soon.
This report does not do justice to the contributions made to the workshop by participants, though these will be included in the final report. This paragraph gives a very brief overview of the presentations. At the workshop we heard about user needs (e.g. the New Technologies SIRSA project at Imperial Col1ege which will be providing a storage facility for images; the needs of the ESRC Data Archive; courseware developers; online document storage in the Netherlands. We discussed the development of formal standards such as MHEG and HyTime (but still are unclear about how everything fits together and which standards wilJ emerge in the marketplace). SGML was mentioned many times and the need to look at the work of the Text Encoding Initiative was noted. The importance of the internet was accepted and we discussed MIME as a storage method and World Wide Web as a delivery tool. The need to look at image formats and vector formats and to distinguish between these was addressed. The need to consider de facto formats which are wel1 used, such as Microsoft formats was presented. We also asked the question about the importance and role of Adobe's PDF.
In conclusion, this was a useful (and not at al1 boring!) workshop which should result in output which is of wide value to the community.
Sue Cunningham has now been appointed as Support Officer for the Support Initiative for Multimedia Applications (SIMA) and will be based at the Computer Graphics Unit, University of Manchester.
She has been seconded to this post from the Medical School at the University of Manchester, where she has been providing a wide variety of user support, including support for CAL and multimedia projects. As a result she is very familiar with the problems faced by many people trying to use and develop multimedia applications for the first time. With such a range of systems and products where do you start? An important part of her role will be to introduce people to multimedia, showing its potential and providing introductory material to get them started.
There is already a great deal of work being done in the academic community, developing multimedia applications, training materials, evaluating new products and so on. It is important that we do not duplicate work, particularly given the large amounts of time invested in developing multimedia. Sue will keep the community informed of work that is being done and products and multimedia training materials being developed. To do this she wi11 obviously need site contacts, so if you are involved with multimedia, please do get in touch.
This is a support post, so email and phone support is available for any problems with multimedia at the contact numbers listed below. If Sue can't help you, hopeful1y she wi11 know a man who can. She wi11 also be setting up project pages on the CGU web server and the experimental AGOCG WWW server to provide general information, dates of courses and seminars, availability of training materials and any other relevant information. If you would like links to your multimedia projects to be included, please let her know. She wi11 be co-ordinating several courses, the first of which wi11 be an introduction to multimedia, and material from this and material that is already available, wi11 be col1ated to produce a Getting Started in Multimedia pack, in both paper and electronic form. If you have any suitable training material you would be wi11ing to distribute, please get in touch. Sue wi11 also be editing an online newsletter, and again contributions wi11 be most welcome.
This is a finite post, so Sue sees one of her major roles as setting up a multimedia infrastructure that wi11 be self propelling, providing training materials, training the trainers and putting people in contact with each other, so that support win continue to be available after the end of this project.
The Combined Higher Education Software Team (CHEST) negotiate agreements with suppliers for software and then administrate these agreements handling contracts, invoicing and payments. In the CHEST portfolio are a number of graphical products. In the last Newsletter the PHIGS PLUS Figaro software from TGS was described. This is now available under Windows NT. In this issue we look at some new pricing agreements which have been announced for AVS and for PV-Wave.
CHEST and AVS/Uniras have concluded an Amendment to the CHEST/AVS Agreement which offers a per-platform licence, as an alternative to the multi-platform licence original1y offered. Sites can choose between a site licence or a per platform licence (with AVS Animator as an option in both cases). The site licence being £4800 per year and the first platform being £2150 with subsequent platforms costing £1150. A site licence for AVS Animator is £900 per year with a single platform being £400 and subsequent ones £225. Sites should be aware that once a licence-type has been selected, it cannot be changed. An charges are quoted exclusive of V AT. It is believed that this Amendment wi11 be of benefit to smal1er Institutions which until now have been unable to justify the current annual fees because they did not have the requirement for running the product on several different platforms.
There is also the option to nominate additional technical contacts, for which there wi11 be an annual charge of £250 for each pair of additional contacts nominated.
CHEST have recently reached agreement with Visual Numerics International Limited for the PVWave family of products. This includes the IMSL C and Fortran numerical and graphical libraries.
There are various options on the agreement depending on whether you wish to spread out the payments evenly over a 5 year period or to front-load the payments with a larger amount in the first year. There are also options for sites who have already purchased the software. As a guideline to the prices, the site licence for a new site wishing to spread payments evenly over a 5 year period is £5550.
CHEST have also just announced the renewal of the Cricket licences and details on this and al1 the above can be obtained from CHEST, your local representative or through information available via the NISS service.
One of the most important choices to be made during the process of software development is that of which software tool should be used. For developers in the Windows environment this choice is particularly difficult because of the bewildering array of authoring systems and programming languages on the market. This report aims to be an aid to tool selection by reviewing in depth five Windows authoring packages:
Each package was evaluated by creating a small hypermedia test application. Whilst general aspects of the package are considered, in particular its ease of use, the review concentrates upon its capabilities in the areas of hypermedia, screen design, and accessing Dynamic Link Libraries (DLLs), which are often neglected in mainstream software reviews.
As a technical report dealing in specific details of systems it is primarily aimed at developers and would-be developers of hypermedia courseware who have a good knowledge of the Windows environment from at least a user perspective. However, nontechnical readers will also be able to extract some benefit from the work. It was written under the auspices of the University of Hull ITTI project Multimedia-based IT Training for the Humanities and may be obtained for 10 pounds from Jean Burgan, USDU, Sheffield University.
An updated set of workbooks for Cricket Graph and Cricket Presents (both PC and Mac), and for Micrografx Draw! for PC, are now available. The workbooks for Cricket Graph (PC and Mac) have been revised to incorporate user feedback. All workbooks have also been made available in ZIP format suitable for direct transfer to PCs.
The specific updates are:
| Package | Version | System | Update |
|---|---|---|---|
| Micrografx Draw! | 1.0 | Windows | ZIP format version now available |
| Cricket Graph | 1.3.1 | Windows | Revised text and ZIP format version |
| Cricket Presents | 1.5 | Windows | ZIP format version |
| Cricket Graph | III | Macintosh | Revised text and ZIP format version |
| Cricket Presents | 2.1 | Macintosh | ZIP format version |
The workbooks are designed to provide a step by step introduction to the basic facilities of the packages involved.
The work to produce the updates, which was funded by AGOCG, was carried out by Computing Services at the University of Edinburgh.
Machine readable postscript copies of the workbooks can be obtained via anonymous ftp from unix.hensa.ac.uk. The files are in compressed tar format and are located in the directory /pub/misc/agocg/workbooks. The index file gives details of the contents.
Copies of the same files are also available in ZIP format so that they are suitable for direct transfer to PC's. These are in the directory /pub/misc/agocg/workbooks/micros. The index file (zipindex) gives details of the contents.
CGM was original1y published in 1987, as IS 8632:1987 by ISO and BS 6945:1988 by BSI. As the technology changes standards need to be enhanced and it is normal practice to review them every five years. CGM has been through this process and a new text has just been published by BSI as EN 28632:1994 (the EN refers to the fact that it is now a European standard, not just a British one). This is identical to the ISO text which was published eighteen months ago as IS 8632:1992.
The new text describes three versions of CGM. Version 1 is the original 1987 specification. Version 2 was the result of an amendment published in 1990 and provides ful1 support for GKS, the ISO standard computer graphics system, and includes segments. Version 3, the result of another amendment in 1992, provides support for enhanced 2D graphics, general curves, and improved font handling. Each version is a superset of its predecessor and so backwards compatibility is maintained.
Version 3 is intended to satisfy the requirements of application areas such as technical publishing, engineering drawing, graphic arts and visualization. It includes the ability to more precisely specify the appearance of the graphics primitives. For example, the dash patterns for lines and the interior fill pattern for filled areas can be completely specified rather than simply chosen from a smal1 set of standard options. Bezier and spline curves are supported and there are enhanced raster facili ties including the Group 3 and Group 4 fax compressed data formats. Text strings can be drawn along arbitrary paths and external symbol libraries can be accessed.
A more detailed description of the facilities available in all three versions can be found either in the standard, available from the BSI, or, in a more accessible form, in The CGM Handbook by Lofton Henderson and Anne Mumford published by Academic Press.
Much of the software currently generating or interpreting CGMs can only handle CGMs complying with version 1. This situation is already changing and in future we can expect to see software upgraded to handle version 3 metafiles.
Queries about CGM can be posted to the CHEST-CGM list at Mailbase.
Earlier this year, the University of Kent embarked on a smal1 project for AGOCG, to see how easy it would be to convert a word processed (WordPerfect) document (the AGOCG Colour Printer Evaluation Technical Report Number 2) to HTML, and then mount the HTML marked-up version on a World Wide Web (WWW) server. This is a summary of that project.
In order to carry out this project, a number of tools and filters were used. These included WordPerfect Intellitag (the prime tool), the Unix PostScript file to image Convertor (pstoppm), the Unix Portable PixMap to GIF Convertor (ppmtogif) and the Unix tool to apply dithering to a Portable PixMap file (ppmdither).
The original AGOCG document was passed through Intellitag, to produce an HTML marked-up form, which could be mounted on a WWW server. The other tools were used to produce sections of the original document that could not easily be converted into HTML.
Intel1itag can either be used to produce an HTML marked-up document from scratch, (i.e. by supplying the text and the appropriate tags), or (as with this exercise), to take a document produced by WordPerfect, and to convert it into an HTML markedup document by inserting the tags into the existing text.
Intellitag must first be given a Document Description from which to work, so an appropriate version of HTML.DTD was pul1ed from the Internet. Before the DTD can be used by Intel1itag, in must be converted by a utility cal1ed DTD2LGC. This utility did not like HTML.DTD! It imposed a number of restrictions on the DTD basical1y in terms of size. This affected the maximum values defined for some constants as wel1 as the number of content tokens in model groups. In al1 instances it was possible to work around the limits, or reduce the maximums defined within the DTD without any (noticeable) problems. Using Intel1itag to convert a document is best achieved if the original document is written in a wel1 structured manner, or if a document is written knowing that it will eventual1y be tagged for HTML (or even SGML).
The conversion can be automated by defining a series of conversion rules. With these, a pattern within the WordPerfect file is searched for, and when found, the appropriate HTML tag or tags are placed in the document. This is why a wel1 structured or planned original document is a distinct advantage. Conversion rules are a very good way to insert tags for section headings, paragraphs, bold and italic fonts and many more examples. If the original document always handles these features in the same way, particularly with the use of style files, then adding the appropriate HTML tags by this automated means is very easy. If however, formatting of the original document is done in a less consistent way (even though the printed version looks consistent), then adding HTML tags automatically by producing conversion rules is less easy, simply because some instances will be missed, whilst inconsistent tags could be added elsewhere. Once the document has been completely tagged, it can be validated (as an HTML document) before being stored in SGML output format.
The Colour Printer Evaluation Report had been written in a way which was usually consistent, so it was possible to tag the vast majority of the text by using Intellitag's conversion rules, so most of the problems referred to above were avoided. To tag the rest, the document had to be run through Intellitag with the human user selecting the appropriate tag. This was less time-consuming than might be thought. With the majority of the tags already in place, Intellitag could be configured to tell the user what the valid tags were for insertion at any point in the document. It then simply became a case of selecting the appropriate one from a context sensitive menu.
The major problem to be encountered was the fact that the original printed document used features for which there was no equivalent in HTML. The document was not rich in such features, and it did not really do anything which was fancy, but none the less, there were problems. These included multi-column output, tables, and the sterling symbol. HTML can just about handle 2-column output, where (say), the first column contains a term to be defined, and the second column contains the definition. The resulting output is acceptable, but not as neat as the printed form. For such instances, the appropriate conversion/tagging for HTML was used. Where 3 or more columns appeared in the original text, maybe as a very simple table, the solution adopted for all other tables (see below) was used.
HTML has no concept of tables at all, let alone tables where text may be rotated through 90 degrees, as was the case with the AGOCG report. The result after tagging has taken place is for these tables to be treated as raw text, so that when the HTML marked-up version is displayed on a WWW server, the actual tables become unintelligible. The solution was to copy the WordPerfect information for the table, plus the table data, into a separate self-contained WordPerfect file, and then to produce a PostScript file from this. This PostScript file was transferred to a Unix system, where the PostScript was first converted to a portable PixMap image, and from that to a GIF file. The resulting GIF file was either inserted in-line in the original document at the point where the table appeared in the printed form, or it formed an HTML image, which could be clicked on in the main document, and then displayed in a new window.
The problem with the sterling symbol is that it does not appear amongst the special symbols defined within the HTML DTD. As a result, amounts preceded by that symbol in the printed original had to be postfixed by pounds in the HTML marked-up version. The only exception is where the symbol appeared as part of a table. There, because the table was turned into a image file, the sterling symbol remained.
Once the very short period of learning about Intellitag was completed, the conversion proved to be very straight forward, and the whole exercise was completed in just over six weeks. (This was elapsed time with many other items calling on the author's time.)
Providing the original document is formatted in a consistent way, and complicated pieces of text (eg. tables) can be converted easily to image files, then the author believes that it should be possible to convert other AGOCG documents (or any document) with the minimum of effort. The fact that other documents have been produced by other word or text processors should not prove to be a major problem. On Unix, there are already convertors/ filters to enable WordPerfect, Latex/Tex and groff documents to be converted to HTML. Intellitag via its ConvertPerfect utility will convert many word processor formats to WordPerfect 5.1 format, thus allowing the use of Intellitag itself.
This exercise used Intellitag as a tool to achieve a specific goal. In that respect, it performed very well. The package does have lots of other features which were not tried. Even the conversion rules which were tried, could be made much more complex than were actually needed for this exercise. Overall, it would seem that Intellitag is a very useful SGML tool.
The final HTML marked-up document and the report produced for AGOCG can be found on: http://www.hensa.ac.uk/
I have always been intrigued by the large Victorian college I pass on one of my routes into Brookes. From the top of the bus one can catch a tantalising glimpse of well laid out gardens. This turns out to be St.Hugh's, the venue for the Eurographics UK Chapter 1994 conference. However it was only on the last day that the weather was good enough to stroll round the grounds. The main building comprises nearly the whole of a block between the two main roads north out of Oxford, the buildings facing the interconnecting road in one long line. The dark oak panelling, staircases and long corridors will be evocative to anyone who has been to boarding school, in particular the Mordan Hall with a minstrels gallery, where the plenary sessions and one of the parallel streams were held.
The conference theme of Multi-media was exemplified by the first tutorial which I attended and was an introduchon to multimedia by Manlyn Deegan, dIrector of Oxford University's Centre for Humanities Computing, operating an impressive array of hardware and demonstrating some of the commercially available educational CD-Roms. While not only demonstrating the professional expertise and breadth of knowledge needed to compile these discs, they also demonstrated the way this media can make a specialist subject accessible to others in a way which is enhanced through the varied presentation and interactive choice of paths through the material. World War One poetry is always moving, even when delivered by a computer. The HyperCard stack of Isaac Rosenberg's Break of Day in the trenches supported by movie extracts with sounds and images of contemporary material was especially so. While the existence of several packages makes creation of multimedia discs accessible to non computer specialists, it evidently needs a sense of vision comparable to directing a film to produce one that does justice to the media and the subject.
I was glad of the diversity offered by the main program and the well timed papers within the sessions, together with the close proximity of the two lecture rooms, which facilitated my session hopping. I was particularly in terested in topics on the second conference theme at the other end of the graphics spectrum to multimedia; John Patterson and Stephen Todd's paper on The application of dot modelling techniques to half toning offered an intriguing approach to optimum resolution for grey shading which is somehow un-photocopiable! I found the papers from the BCS colour group on colour standardisation of similar interest. I am indebted to Anne Mumford for giving me insight into the jungle of the current set of higher education research funding initiatives. But it was the paper presented by Ralph Highnam and Mike Brady of Oxford University which won the award for the best paper. It was an excellent presentation of socially sensitive applications of computer technology.
The first evening social was a sherry reception in Blackwell's bookshop. This, their main shop, boasts the Norrington room; a subterranean chamber with descending terraces reminiscent of les Hailes in Paris, underneath Trinity college with 3 miles of bookshelves on which to browse. This was followed by a buffet in the dining hall of Hertford College just across Broad Street. Definitely the best way to view an exhibition is with a glass of sherry in your hand. Good job this prevented me from getting at my credit cards. Having not seen Silicon Graphics Indigo machines before I was impressed not only by their power but by the flexibility with which they could be configured to run the Falcon flight simulator to form a central screen and side views to give the peripheral vision not available on a single screen. Their teleconferencing capability is not to be sniffed at either! Pleasant preparation too for the conference dinner in the college dining hall. I came to the conference thinking that multimedia was still just a combination of existing techniques and needed some fresh insight to progress. For me the Wednesday afternoon panel session was a turning point. The panel members were all at the forefront of the publishing industry and their individual views were fascinating. The ensuing discussion convinced me that this insight has arrived and that multimedia is on the verge of becoming an exciting and spectacular new vehicle.
Anyone who reads the Sunday quality papers, the Guardian on Thursday or watches The Net on BBC 2 will know that 1994 is the year that the man in the street became aware of computer networking. Interest in computer networks, such as the Internet, is not confined to the chattenng classes. MTV regularly alludes to computing and cyberspace and the Yorkshire Evening Post ran a feature on networks. in June 1994: Brian Kelly, the Principal Information Officer in the Computing Service, University of Leeds. attended two international networking conferences recently. He reports on the conferences.
The world's first World Wide Web (WWW) conference was held at CERN, Geneva (the birthplace of WWW) from 25-27 May 1994. As a long-standing Webmaster Leeds University has been running a WWW service since January 1993) I was looking forward to meeting software developers and other Webmasters whom I had previously only met on the 'Net. The conference also provided me with the opportunity to renew my acquaintance with Robert Cailliau, who, together with Tim Berners-Lee founded WWW. Robert, whose wife is from Leeds, made contact with the Computing Service on a visit to her family in March 1993. This chance contact convinced me of the importance of WWW, especially after Robert had reassured me that I should not be too concerned about the small size of the development team at CERN (2 people!) - CERN had established links with an organisation known as NCSA. Robert felt that NCSA would develop software which would help popularise the Web. This was clearly something of an understatement!
The WWW '94 conference attracted over 380 delegates from around the world. Participants from computer companies (DEe, Cray, Novell, IXI, Oracle, SCO), publishers and Government funded organisations (Science Policy Support Group, World Meteorological Organisation) attended as well as Thursday. or watches The Net on many from academic institutes. For me the most important announcement came from Robert Cailliau in the opening session. Robert informed us that a World Wide Web Organisation (WWWO) is to be set up shortly. WWWO will have its European centre at CERN and a US centre at MIT. WWWO will concern itself with the development of standards, validation test suites etc. This consortium which has been compared with the X Consortium, will provide the stability needed by software developers and large organisations wishing to use WWW. Another important non- technical area was addressed by Tim Berners-Lee in the opening session. Tim emphasised the importance of a bill of rights for cyberspace and to avoid disenfranchising groups, such as developing countries with poor network connectivity.
The technical sessions I attended included Dave Raggett's review of HTML+, and sessions on server technology and HTML authoring tools. Dave Raggett (of HP Laboratories, Bristol) described how HTML+ will in future be referred to as HTML 3, HTML 4, etc. HTML 1 was the initial specification of the Hypertext Markup Language, as defined by Tim Berners-Lee. HTML 2 is the de facto implementation which is supported by the NCSA Mosaic client, which includes support for forms. HTML 3 will include support for tables and provide increased control over formatting. Future developments of HTML will include the provision of a safe, platform-independent scripting language. This is an exciting development, which will enable WWW to be used for the development of interactive distance teaching and learning services.
The sessions on server technologies included. presentations on the caching and mirroring technologies which are now available. The CERN http server, for example, provides caching facilities, which mean that regularly accessed files can be held locally, improving the access time and reducing network traffic. The main problem with caching is ensuring that the cache is updated when the file on the remote server is updated. Work is in hand to deal with this.
A number of presentations of HTML authoring tools were given. Unfortunately, from my point of view, the talks described the use of Unix tools. The only aspect of the conference which disappointed me was the emphasis on small-scale developments by research workers. I hope that WWW' 95.to be held in Frankfurt, will address issues associated with running a large-scale WWW service.
The INET '94/JENC 5 conference was held in Prague on 15-17 June 1994. This conference, jointly organised by the Internet Society and RARE, attracted over 1,100 delegates. The week before the conference a network training workshop was held for delegates from technologically emerging countries. Delegates from over 100 countries attended this workshop which provided the opportunity to install network services, such as WWW and Gopher servers.
The INET '94/JENC5 conference was split into six tracks, covering User Support and Training, Distributed Applications, Policy Issues, Regional Issues, Network Engineering and Network Technology. I attended sessions covering User Support and Training issues and Distributed Applications. In addition I presented a paper on Becoming An Information Provider on The World Wide Web.
One of the most interesting talks I attended was Dave Raggett's presentation on Extending WWW To Support Platform Independent Virtual Reality. Dave described VRML, the Virtual Reality Markup Language, which is currently being developed. VRML will extend the capabilities of WWW from a flat 2D graphical environment to include distributed virtual reality. Anyone interested in finding out more should access the Virtual Reality WWW service which is available on Wired's WWW server at the URL http://www.wired.com/vrml/.
I also attended sessions on Electronic Documents and Issues In Building The Virtual Library. In talks at both of these sessions WWW was mentioned frequently.
It is clear to me that the World Wide Web will have a profound effect, not only for academics and students in their study and research, but for society as a whole. Once cable technology becomes widely available access to Internet service from the home will become feasable. Sporting information, such as the World Cup server at: http://www.sun.com/ and home shopping services: http://www.internet.net/ are already available. The most popular services, however, may well be pornography servers (and yes some already exist). Universities should currently be considering how the global virtual campus will affect them. More importantly society as a whole needs to define Tim Berners-Lee's bill of rights for cyberspace. Technological-push should not take precedent over society's needs.
This conference is the main international event for all users and developers using the Application Visualization System (AVS). The attendees numbered over 300 and included participants from the UK, Japan, Germany and France as well as the US. The attendees were a mixture of both academic and industrial members and came from a wide-range of backgrounds. The format of the conference included a number of parallel streams consisting of assorted lectures, tutorials, panels and workshops with a large exhibition area for attendees to visit during lunch and breaks. The conference had a number of application specific streams but overall covered a broad range of applications and disciplines with the common theme that they were all using AVS.
This year the conference coincided with the release of details for the next major release of AVS and a whole track within the conference was dedicated to technical overviews of the new architecture. The following sections provide additional information on the new developments.
Over the past 18 months Advanced Visual Systems Inc. have been working on the next major release of AVS, called AVS6. Prior to the release of AVS6 a new product, AVS/Express, will be released in June 1994. AVS/Express is designed for developers who are creating technical applications for distribution and resale to external customers and is the replacement for the AVS Developers Kit.
AVS6 is the next release of AVS, replacing AVS5, and is what UK academic sites who have signed up for the CHEST deal will receive. The planned release of AVS6 is early 1995. Both AVS6 and AVS/Express share the same common architecture and so new features detailed below will be reflected in the release of AVS6.
An extensible data model
In AVS5 different data types were required for different classes of data (ie 2/3D data, finite elements, geometric). AVS6 now has one unified data type which can be used to represent these classes of data within the same data structure. This removes the need for different visualization modules to perform the same function on different types of data eg there is now only one isosurface module for both field and unstructured cell data.
Another improvement is the facility to extend this base data structure with user defined fields. For example you could add patient attribute information to image scans and the modules within AVS would still recognise and process this extended data type as an image.
AVS5 has different viewers for image, graph and geometric data making it hard to combine geometric and volume information within the same view. AVS6 will have just one viewer for all types of graphical data. There have also been many improvements to increase the performance of the viewers and one of these is the facility for direct data rendering. This would be ideal for large datasets as it allows large geometric scenes to be produced directly from the data without the need for intermediate generation and storage of geometric information.
AVS5 and earlier versions were based on the dataflow paradigm which meant generating multiple copies of the data as it passed through the visual data analysis pipeline. Improvements such as the use of shared memory segments and multiple modules within a single process were used to eliminate some of these overheads. The new architecture is based around the concept of a single datapool and modules use data-references in most cases. Data is only copied when mandatory and then it is just the portion being modified.
The visual programming paradigm was really the concept of connecting together modules in a dataflow network to construct an application. The modules in this sense were at the level of functions or subroutines. In AVS/Express this concept has been extended where modules, or objects, themselves can be constructed from other objects within the system using the visual programming paradigm. This was illustrated in a plenary session were Dave Kamins, AVS Inc, constructed a module, similar in functionality to the AVS5 geometry viewer, from smaller base objects. To support this new functionality the V language has been designed as a replacement for Command Line Interface (CLI) and can be used to represent these inter-relationships and connections.
AVS5 was based upon their own graphical user interface, LUI, which as a user you either loved or hated. AVS6 will now have a Motif user interface on the workstation platforms and a Windows interface on the PC.
AVS6 is being planned for release onto PC platforms with the first release based on Windows NT and later versions for Windows 4.0 and DOS.
Other features include support for cylindrical, polar and spherical coordinate systems and the provision for users to specify NULL (undefined) data values. The visualization techniques have also been extended to operate on any primitive data types (byte, integer, float) and handle 1D, 2D and 3D grids for scalar and vector data where applicable. New processing techniques include a math expression evaluator that has been added to the AVS6 command line language and new visualization techniques include glyph which produces a user-defined, scaled and coloured geometric object at data values within a dataset.
Since AVS have acquired UNIRAS they also plan to include some of the UNIRAS technology into the release of AVS6.
At this point some existing AVS5 users may be wondering if there are any compatibility issues. The release of AVS6 will have a very high level of compatibility with AVS5 which includes:
The only real area of change is the user interface aspects due to the change from LUI to Motif and for this reason an applications appearance on the display will be quite different in Motif versus LUI.
While users are awaiting the release of AVS6 the new graphing and plotting module for AVS5 will be released in Q3 94. It will be freely available to existing AVS5 users as part of the maintenance agreement and will therefore include sites signed to the AVS CHEST deal. The module makes use of assorted Uniras agX/Toolmaster routines to provide a variety of plot styles including line graphs, scatter marker plots, multi-component stacked bar charts, contour plot, etc.
