Search
Welcome to Graphics Newsletter No 4! If you have not seen this Newsletter before and wish to get on the mailing list, please send your name and address (preferably by email) to me. New readers are always welcome! New contributors are also always welcome!
What's New in Graphics?
I have visited two computer graphics exhibitions recently. One at SIGGRAPH 89 in Boston, and the other at CG 89 at Alexandra Palace. I was impressed with the variety of new hardware and software on display, both for low-end PC's and also high-end Workstations. As low-end PC's become more powerful, and have more memory and disk as standard, they are able to run more sophisticated graphics and other applications packages (e.g. desk-top CAD). Software is proliferating fast in these areas. If any readers have any practical experience of using any of this software for real-life problems, we would be pleased to hear from you. Short articles, notes, contributions are always welcome. At the high-end, Visualisation Software for Workstations is an area of increasing interest. We are planning an article on this for a future edition of the Newsletter, but again - if you have any experience of this software. please send me a short note.
New advances and developments seem to be the name of the game in computer graphics and its applications. This does pose real problems for the provision of standard software and interfaces. which often end up (by definition) supporting a trailing edge of technology that is no longer in real use. People move to XWindows, PostScript. NeWS, etc. once it becomes clear that there are significant benefits in doing so. How does one reach a consensus on what features of the current hardware and software scene are worthy of standardisation (e.g. as ISO or ANSI standards) when the standardisation process takes so long that by the time the standard is approved the technology has moved on significantly? This seems to pose a real problem for standardisation in the area of computer graphics. If anyone has any thoughts on this. or would like to write a short article addressing the issues from either the graphics viewpoint, the standards viewpoint, or the vendor viewpoint - please let me know!
What will be new in the future? Prof Nick Negroponte (of the Media Lab, MIT) in his Keynote Address at the SIGGRAPH 89 Conference painted a scenario of our future use of computers as different from the present as the present is from the punched card and flexowriter. It is hoped to provide a summary of this address in a future edition of the Newsletter.
Just thought I would pick out a few topics that are currently on my mind (and indeed cluttering up my desk!).
A second meeting on this subject was held at Edinburgh University. There was an agreement from the group that they would contribute towards the Training materials and they are going to work on producing a set of exercise sheets for use in courses and in self-help situations. I am looking for offers of help still particularly in the production of facts cards to be used as prompts for users. Anyone interested?
My current plan is to get people together for a brief meeting to review progress at the Eurographics UK meeting in Bath. Following that there will be a workshop to get people together who have been working on the materials to carry out evaluation and testing. This will be in May and out of this will come the final drafts to be produced hopefully in June. This should mean that sites will have the training materials for Version 6 in good time for next academic year. I am hoping that people will pool their efforts rather than developing their own materials.
Look out for the announcement of UNIRAS courses for support staff. I am currently working on setting up courses on Version 6 and also on installing Version 6. I plan to hold these more than once in different locations. Mail me if you want to guarantee getting details.
I hope that we will see a deal on a GKS implementation by the end of March. We will be looking for a 2b level implementation with support for CGM.
I have been collecting information on colour printers and have now got to the stage of evaluating this information. Any experience with these devices would be appreciated. Also I would like to know what you would like from such a deal. How important is the initial cost to your site? How important is the recurrent costs and the cost per copy? Would you be interested in having a printer if a deal was to be got together and under what circumstances? Let me know.
If you have any views on what AGOCG should be looking at in the future then send me some mail, I would be interested to hear from you.
The first issue of this Newsletter announced that we are looking at the Standard Generalized Markup Language (SGML) as a means of exchanging AGOCG documents (such as manuals, reports, viewgraphs and handouts). This article gives some technical background to this.
Documents are currently produced in a wide variety of formats using a range of text processing and word processing systems. These systems commonly fall into two categories: batch formatters and interactive formatters.
In the first category, troff, TEX and Script are in common usage. In the latter category, Microsoft's Word, Interleaf and Framemaker are widespread. Instructions given to formatting programs to control style and layout are termed markup. Formatting commands in batch processors are known as procedural markup, as they are concerned with the appearance of the text at a very fine level of detail and control directly attributes of the text such as font, spacing and page breaks. This is not the level of detail at which authors wish to work, and troff, for example, was soon enhanced by a set of macros, the ms macros. These macros enabled the author to mark the major structural elements of a document; for example, title, section headings, paragraphs etc. The representations, or typographical conventions to be followed in laying out these entities were contained in the macro definitions. This approach ensured a uniformity of appearance in documents produced using a particular macro package.
TEX was similarly enhanced by the LATEX macros, which are in widespread use in the academic community. Other TEX macro packages have also appeared, for example AMS-TEX, a macro package used by the American Mathematical Society for mathematical typesetting.
The troff formatting system has also been enhanced by the addition of specialist preprocessors, eqn for typesetting equations, tbl for tables, pic for diagrams and grap for graphs. Each defines a simple language for describing entities of the type it deals with, which is compiled by the processor either directly into the troff input language or to the language of one of the other preprocessors.
There are two methods commonly used for incorporating graphics in documents. In the first, the formatter is enhanced with graphics capabilities. This is the approach taken by pic and troff. In the second, space is reserved for the diagram by the layout system and a marker is passed to the backend indicating which particular diagram is to be placed where. The backend processor then interprets a description of the diagram to produce a composite page description for the output device.
Word processing and desktop publishing systems also include markup, but this is more hidden. Systems such as Word, tag each paragraph (a document is a sequence of paragraphs) with a code which determines the style in which it is to be laid out. The layout is described in style sheets. There is a correspondence between style sheets and macro definitions, and between formatting commands and paragraph tags.
Using Word on the Apple Macintosh as an example, diagrams are prepared using a drawing package such as MacDraw, and are pasted into the document as images. The diagram is treated by the Desktop Publishing System as a single paragraph entity with particular dimensions. More recently Desktop Publishing Systems have appeared which are capable of accepting diagrams described using CGM. SGML, Standard Generalized Markup Language, was published as an ISO Standard (ISO 8879) in 1986 and amended in 1988. It is important to understand exactly what is standardized by SGML. It is described in detail in the very readable book by Bryan, SGML: An Author's Guide, Addison-Wesley (1988). Barron in his excellent article (Why Use SGML?, Electronic Publishing 2 (1) pp. 3-24, 1989) puts it very clearly: SGML is not a standard markup language, and it is not a formatter: it is something altogether different SGML is a meta-language that defines only the syntax of a standard generalized markup language, i.e. it prescribes how we should specify markup, but not what that markup is, nor what it means. SGML defines an abstract syntax for a markup language, and so provides a standard mechanism for generating a family of descriptive markup languages which can be used to describe the structure of a document, but nothing else. A descriptive markup language generated by SGML is normally used as a front end to a program that will perform subsequent processing of the document this may be a formatter if we are producing printed documents, but may just as well be a database management system or a hypertext stack generator.
SGML defines a standard way of specifying a document type definition (DTD), which defines the logical structure of a document in terms of the elements from which it is constructed (for example, paragraphs and headings) and their relationships (for example the constraint that a second level heading can only occur within the scope of a first level heading). A generic identifier is associated with each element, defining tags that will be used for the descriptive markup. The DTD can specify that a tag has mandatory or optional attributes. There are mechanisms for minimizing the amount of markup an author needs to give; the positions of missing tags can be inferred.
An SGML parser (which is really a parser generator or compiler-compiler) takes the DID and generates a parser for documents of that type. This program checks that a particular document conforms to the specification given by the DID, and inserts tags whose presence is implied. The output is then passed to another program which maps the SGML tags onto the command set of the underlying text formatter. This mapping has nothing to do with SGML and is totally outside the scope of the standard. (This article will be concluded in the next issue of the Newsletter.)
Seminar to be held 28 February 1990
The BCS Computer Graphics and Displays Group, in conjunction with Eurographics UK, is holding a state of the art seminar on the theme of Systems Integration and Data Exchange.
Integration is a topic which many people talk about but do not always demand when buying products. The seminar brings together leading experts in the field of using standards for integration who will show that standards are available today to bring closer together the design, documentation and artwork involved in many applications. The isolation of systems and applications within and between sites should not be a feature of today's computing. Distributed computing across networks is a fact of life. This seminar will look at the ways in which applications can also be more closely coupled.
The case for standards will be introduced by the day's Chair, Dr Anne Mumford from Loughborough University. Dr Mumford has been an advocate of the use of graphics standards and organised the recent demonstration of the use of the Computer Graphics Metafile standard at the Computer Graphics Show at Alexandra Palace. She will be suggesting that it is practical and realistic to demand standards for interchange when purchasing software and hardware.
The integration and exchange of text and graphics is of concern to most people in the industry. Many people wish to combine graphics drawn on one system with text on another and to exchange complete and draft documents. The issues surrounding document interchange will be addressed by Heather Brown and Paul Ellison. Alan Francis will then address the confusion which often seems to be present when talking about the choice between CGM and PostScript, Alan will suggest that it is a case of horses for courses.
The exchange of application data for engineering designs is also an important subject and will be addressed by Dr Jon Owen of the CAD-CAM Data Exchange Centre at Leeds. The exchange of engineering designs and product information between dissimilar CAD-CAM packages is a complex problem and one which only internationally accepted standards can solve.
The US Department of Defence see the use of interchange standards as a key issue for their work, much of which is subcontracted to a wide range of companies. Their CALS (Computer-aided Acquisition and Logistic Support) initiative is demanding the use of interchange standard when contracts are issued. Product data, graphics and document standards are all included and already we are seeing this having an impact on the market place. This is inevitably going to affect the European market. Norman Harris from Procad will describe the CALS work.
The European Commission are also aware of the need for interchange standards and have funded an Esprit project looking at the integration of graphics and communications. This is the ARGOSI project which will be presented by Prof David Duce.
Testing of implementations of standards is something which is essential if interchange is to be a reality. This topic will be referred to by all the speakers either directly or indirectly. It will be the focus of a presentation by Sheila Lewis of the National Computing Centre who will round up what looks to be an exciting and informative day of presentations.
Looking through the report of the CAD/CAM Working Party of the IUSC I saw that the issue of PostScript from Pafec was being addressed. I have a utility which will convert GINO PSEUDOCODE into PostScript. This copes well with the files generated from Gino directly and worked with the one DOGS file which I tested (lack of a suitable terminal makes the use of DOGS difficult) was satisfactory.
