Search
Welcome to Graphics Newsletter No 5! If you have not seen this Newsletter before and wish to get on the mailing list. please send your name and address (preferably by e-mail) to me. New readers are always welcome! New contributors are also always welcome!
I am currently working on putting together a set of documented OHPs for graphics standards, GKS and CGM. These have been compiled from material produced primarily by Bob Hopgood, David Duce and myself. These will be added to the PHIGS handbook and accompanying OHPs being produced by Toby Howard of the University of Manchester. AGOCG are also considering a practical guide to using GKS which will incorporate program examples and might be useful for course notes.
I hope that these will be available early in the summer and will be useful for courses being given next academic year. These are not aimed at specialist computer science courses but at short-courses given by Computer Centre or departmental support staff.
The AGOCG request for information on colour hardcopy resulted in a good response. The information which has been gathered will be put together in a technical document once the exercise is complete. The results are being considered and a short list will be drawn up. Major considerations will be ease of use as a central printing service, interfaces to standard formats (CGM, PostScript) and running costs. The short-listed devices will then be field tested to ensure they will interface to a range of software and that they do fulfil our requirements. It is hoped that final negotiations will be conducted during May.
The CGM is beginning to be used more widely in the community as a means of exchanging pictures. To assist in its use a joint effort between the SERC and the Computer Board is being proposed to provide utilities for:
This will use software which is already written in the community where this is available. If you have any views on additional capabilities for this software then I would be interested to hear them.
There is to be a workshop in May to bring together the work on UNIRAS training materials, to test it all, look at slide sets and come up with final versions for the work. All the people who are currently involved with the efforts will be invited. Please let me know if you are interested.
The IGWP met in March, but due to deadlines it is not possible to report progress made at that meeting. However, it can be reported that there have been some changes to the membership: Roger Discombe (Sussex University) and Alan Oborne (UWCOC) have left the IGWP, to be replaced by Yien Kwok (MCC) and possibly others, yet to be confirmed.
Both Roger and Alan have made a great input to the working party, I am very grateful for their valuable contributions. Roger, the last non technical member, will remain in touch as a 'Corresponding' member of the IGWP. Yien will be no stranger to many, I am particularly pleased to welcome him as it will be valuable having someone from a national centre on the IGWP. I would be interested to hear from anyone who feels that they could make a good contribution with respect to PC graphics.
Les Zieleznik has added Lotus Freelance to his table of features of PC interactive graphics software that was published in the February issue of the Newsletter. If anyone requires a full, updated copy, they should email Les.
Precision Visuals (PVI) have mailed the academic community about offers for their PV - WAVE program that runs on V AXNMS and SUN systems. The IGWP will produce a review of this software in due course. The IGWP also hopes to see the results of a review of USEIT shortly, which is being made at the request of the IGWP by SERC.
BCS Meeting, 6-7 December 1989
Huw Jones and Tony Crilly, both Principal Lecturers in the School of Mathematics, Statistics and Computing, Middlesex Polytechnic, organised this 2-day meeting on Fractals and Chaos.
The event brought together speakers from the USA, West Germany, Holland, England, and Wales. Delegates came from as far afield as Italy, Belgium, USA, Sweden, Finland, and Japan, as well as the various parts of the UK. Fortunately for the organisers, the fog at Heathrow was not serious enough to cause problems for the visitors. However, the plane bringing the Keynote Speaker, Prof Michael Bamsley, from the USA had to return to Georgia from the mid-Atlantic due to problems with its navigational systems. This problem was put right and the plane began its journey again. Prof Barnsley arrived in time to speak, but not to have a sleep!
The word fractal was coined by Benoit Mandelbrot in the late 1970s, but objects now defined as fractal in form have been known to artists and mathematicians for centuries. Mandelbrot's definition - a set whose Hausdorff dimension is not an integer is somewhat opaque to the layman. The important concepts are those of self-similarity and sub-divisibility. A fractal object is self-similar in that subsections of the object are similar in some sense to the whole object. No matter how small a subdivision is taken, the subsection contains no less detail than the whole. Examples illustrated are Durer's Pentagons (known to the artist Albrecht Durer in about 1500), a Pythagoras Tree and the Koch "Snowflake Curve (devised by Helge von Koch in 1904). Koch's curve is a mathematical peculiarity which, if continually subdivided, produces a curve of infinite length which encloses a finite area. The distance along the curve between any two points is immeasurable - there is not enough wire in the world to bend into the shape of the Koch curve. These examples have exactly similar subsections, but many fractal objects, particularly those which occur naturally, have statistically similar subsections, so that subsections have similar forms with some variations.
Prof Michael Barnsley began by describing his method of fractal image compression. This was also demonstrated by a live system in a small exhibition held in association with the Conference. Thanks are due to Academic Press and Sun for providing the facilities for the demonstrations.
The remainder of the first day was given over to presentations on the use of fractal techniques in various areas of scientific investigation and modelling. This included tiling structures, sponges, image synthesis, towns and cities, and physics.
Chaos is a topic that has developed with close links to fractals. Chaotic systems have the appearance of unpredictability, but are actually determined by precise mathematical laws, just like many fractal images. Chaotic systems - often referred to as non-linear systems - show major fluctuations in behaviour for apparently minor changes in the parameters which control them. The butterfly effect is used to illustrate the concept of a chaotic system. The breeze caused by the beating of a butterfly's wings may be the initial seed which eventually generates a hurricane. Similarly, minor changes in the parameters controlling the behaviour of a pendulum or the flow of a fluid could lead to major changes in the form of motion produced from smooth to chaotic. Minor changes in parameters thus cause major changes in the behaviour of chaotic systems. Such theories have been applied, for example, to meteorology, irregularities in heart beats, population modelling, quantum mechanics and astronomy.
One objective of the Conference was to examine the relationship between fractals and chaos, and the second day looked at various aspects of this including relativistic particles, physical systems, phase portrait analysis, neural networks, data visualisation, and design.
Some very striking fractal images were presented in the form of slides and video. This added significantly to the enjoyment of the meeting. Although these could not be contained in the proceedings, it is hoped that they will be made available in a subsequent publication.
Fractals is an area of great significance and potential. A new dimension of interest for computer graphics!
A joint workshop between the BSI graphics panel (IST31) and Eurographics UK was held at the end of January to look at the curve specifications which should be added to the graphics standards. In the first instance this is expected to affect the CGM standard (through an amendment) and GKS (which is under review).
The current proposals for extensions have proposed a number of different specifications for curves (in fact just about everything anyone might want to use!). The problem with this is that implementers of software will tend to use the curve specification that most easily matches their internal specification. The result is that supplier A chooses curve specification type X and supplier B chooses Y. The result is that a metafile written from one system may not be able to be read by the other. The inclusion of many forms of specification does not help standard interchange.
The proposals which came out of the workshop will form a BSI input to ISO. The conclusions are that the graphics standards should include a single curve specification, the Non-Uniform Rational B-spline (NURB). Mappings to and from this specification are well known and documented in the literature.
A detailed report of this workshop will be submitted to Computer Graphics Forum.
Copies of the PHIGS and GKS-3D standards are now available as British standards. Their numbers are BS7196 for GKS-3D and BS7217 for PHIGS. These and other standards are available from the BSI at Milton Keynes.
Part 2 (continued from Issue No 4)
The use of SGML enables a document containing markup to be transferred, but sender and receiver have to agree as to the actual tags used and their meaning. The use of the same DTD by sender and receiver is part of this. The representation of the document has to be agreed outside SGML.
SGML does not standardize a mechanism for including graphics in documents. There are many ways in which this can be done; for example, a clear text encoded CGM can be incorporated directly into a document surrounded by appropriate tags. Another method is to use an element with an attribute which gives the name of a file containing the picture.
The requirements of the sites served by AGOCG are seen as an interchange mechanism for documents of a variety of types, which can be tailored and processed locally, all of which may contain text and graphics.
The types of documents currently envisaged are:
The approach being explored is to use SGML for document interchange. This approach is consistent with the open systems philosophy and can potentially be integrated with existing formatters in use. Since the requirement is to interchange documents in terms of their content rather than precise layout, SGML is the appropriate standard to consider.
The following issues then arise:
Two DTDs suitable for general use have been defined, one by the British Library, the other by the Association of American Publishers. DTDs have also been defined by a number of organizations for local use, for example CERN have defined DTDs for reports, letters, memos and viewgraphs. The GMD Research Centre for Open Communications Systems (FOKUS) has defined DTDs for reports (manuals), papers, letters and viewgraphs. There is discussion in the physics community about a DTD for submission of physics articles to publishers. The publishers Addison-Wesley have a DTD for their book series on computer science. SGML has been chosen as the basis for the US Dept of Defense CALS (Computer-aided Acquisition and Logistic Support) programme. The CALS programme is concerned with the control of paperwork associated with the design, development, manufacture, procurement, testing and maintenance of weapons systems. An integrated system in which information is held electronically is to be produced which interfaces with CAD/CAM systems, electronic publishing systems and databases, both within DoD and in the contractors supplying DoD. Defense contractors will be required to use SGML for their documentation, using DTDs defined by the DoD. A DTD for technical manuals conforming to the appropriate Military Specification already exists. Macro packages such as ms and LATEX can be extended by the user. DTDs cannot be extended in this way, rather a new DTD has to be defined, based on an existing type. This means that DTDs have to be chosen carefully, if proliferation is to be avoided. We would like to hear from anyone who has experience of these or other DTDs.
Processors which support SGML are appearing on the market. Two are considered briefly here, MARK-IT produced by Sobemap (a Belgium company, distributed in the UK by Yard Software Systems), and Daphne, produced by the GMD FOKUS group in Berlin with funding from the DFN (Deutsche Forschungsnetz - German Research Network).
The MARK-IT system provides an SGML parser and a translator from SGML markup to troff, TEX and various desktop publishing systems (including Ventura publisher). It also supports the reverse translation process. MARK-IT has an Applications Language in which the translation between SGML entities and formatter entities is defined. The system was developed for the CEC Official Publications Office. Versions exist for PCs, SUN, VAX etc.
The Daphne system (Document Applications Processing in a Heterogeneous Network Environment) provides an SGML parser and translator. Five document classes are defined for use within DFN: report (which covers manuals), subreport (which allows an author to create individual sections of reports), paper, letter and slide set. Translations are defined in the system to TEX, troff and DCF markup for the predefined DTDs.
The system provides a language called LAUREL (Layout Assignment Unified Rewriting Language) in which the translation from SGML markup to the form required for a particular formatter can be described. Laurel enables the context within an SGML document to be determined and the appropriate actions to be indicated.
The system supports the importation of graphical images, either as embedded CGMs in clear text encoding or as bit-mapped images in Tag Image Interchange Format (TIFF). TIFF graphics are stored as separate files.
Implementations of Daphne exist for VAX/VMS, IBM VM and Unix.
We would like to hear from anyone who has experience of using these, or similar systems.
