Part of the EASE programme concerns the installation of Central Servers in the community, to provide support for the single user workstations. At various times, requests have been issued for bids from HEIs to acquire one of these. The current purchasing policy is based on an assessment carried out in 1986, when SUN and Pyramid servers were selected. The market place has changed considerably since then, so the Computing Facilities Committee Technical Advisory Group (CFT AG) asked for a new evaluation, which could be used as a basis for future purchases.
The original intention was to carry out this evaluation during October and November last year, producing a report for the December meeting. Unfortunately, it was not possible to complete the job satisfactorily in the time. Instead, an interim report was produced, and the exercise extended into the first quarter of this year. This article gives a brief report of the work done so far.
A list of 19 potential suppliers was drawn up and these were invited to take part during October and November. Suppliers were sent an Operational Requirement and were asked to provide a machine on loan which would be tested in a network of up to 10 client workstations. With the increased number of Approved machines, and the appearance of dataless machines in the SUN range (ie workstations with a small disc which is used by the system, as opposed to discless ones without any local storage peripheral), it was decided not to insist that the Server be capable of supporting discless clients. However, it was considered essential that each should run Unix, fully support NFS and Yellow Pages, and also be capable of connecting to JANET. This last requirement proved difficult for many of the original set, and some pulled out immediately. Others, also unable to meet all of these conditions at this time, agreed to take part but accepted that their machines could not be selected for purchase until these mandatory requirements had been fully met.
Of the suppliers who replied to the OR, Acorn, Hewlett-Packard, IBM, Meiko, Mistral and Pyramid were unable to provide a machine for testing in the time specified. Apollo, Sequent and Unisys stated that they did not wish to be considered. Data General and Stellar required more time to provide the JANET connection, and decided not to participate this time. This resulted in only four machines being made available:
| Supplier | Model | Processor | Memory | Hard disc |
|---|---|---|---|---|
| DEC 5400 | MIPS | 32Mb | 400Mb | |
| MIPS | M/120-5 | MIPS | 32Mb | 328Mb,663Mb |
| Solboume | 5/801 | SPARC | 32Mb | 830Mb |
| Sun | 4/370 | SPARC | 32Mb | 3*327Mb |
DEC was unable to supply sufficient disc space to test the network to its limit, and could not demonstrate Coloured Book software. MIPS was not able to provide an example of their preferred machine (RC3240) for testing (the M120 supplied being rated at 13 Mips as opposed to 18) and this of course affects the benchmark results. In addition, MIPS could make neither Yellow Pages nor Coloured Books available for demonstration. Solboume have yet to complete the port of Coloured Books to their machine. SUN provided the slower SCSI discs rather than SMD, which also affects the results adversely.
Only SUN was able to demonstrate all the mandatory requirements during the period of the evaluation. Much effort is going on elsewhere to provide software for other machines which meets the specification, but it is not available yet.
With these disappointing results, the CFTAG Meeting in December recommended that the exercise be extended in the hope that some of the above problems can be overcome. It should then be possible to benchmark a more representative range, and some of those who declined first time round may also be persuaded to take part. A full report will be prepared for the April meeting of CFTAG, and a version will be made available to the Community. The information gained will be used to assist CFT AG in deciding the type of server to be offered when the next Request for Bids is made. Since there are a number of suppliers whose software is likely to become available shortly, it is likely that the evaluation will be repeated at the end of the year.
This article was prompted by an EASE panel visit to the University of Manchester revealing that the services and facilities offered through institutional membership of the ECAD Initiative are not widely known among the engineering research community. Indeed, at larger institutions where engineering disciplines may be split and possibly widely separated on a campus, it may well be that only departments of electrical and electronic engineering enjoy the benefits of free access to the broad spectrum of commercially supported electronic system packages provided through membership of the ECAD Initiative.
It is just over three years since the first deliveries of commercial software for electronics computer aided design were made to the 90 higher educational institutions who had agreed to participate in the ECAD Initiative. The scheme was established during 1986 with capital grants from the University Grants Committee and the Department of Trade and Industry to purchase site licences for a range of software to cover the ECAD needs of universities, colleges and polytechnics for both teaching and research. It was the research community's requirements which prompted the early entry of the SERC ECAD Central Facility at RAL to become an active member and contributor to the scheme. The value of membership was immediately recognised in that licences for software packages which previously would have cost from £20,000 to £50,000 per seat. were made available at zero cost for unlimited numbers of seats. It was therefore scarcely surprising that the initial uptake of individual packages was well in excess of what institutions would normally expect to embody in their research or teaching programmes. Nevertheless the demand had to be met but not without considerable strain to the scheme's limited administrative and distribution resources. However, three years on, the users of ECAD software in academia are very much more selective and so the pattern of use which has developed is in consequence more manageable from a support standpoint.
The way the ECAD Initiative operates is as follows. Membership is secured by an institution paying an annual fee of £4381.50, in return for which licences are granted for educational use of a range of commercial software packages for the design of electronic systems. The list of suppliers includes Mentor Graphics, Silvar Lisco, Genrad, Racal Redac, Qudos, MCE, European Silicon Structures, Metasoftware, Integrated Solutions and EEsof. The design needs covered are structural and behavioural system description with linear and digital simulation, test analysis, chip layout and printed circuit board design. With the recent entry of Mentor Graphics as suppliers to the scheme, tools for electronic design documentation and CASE are now also available. The membership fee covers the cost of software maintenance and also support. in some cases directly from the supplier and in others via so-called Lead Sites. A Lead Site is an academic based centre of expertise which acts as an interface between the software supplier and the academic user. Software is distributed through the Manchester Computing Centre which also provides administrative and financial control services for the scheme.
Following the initial capital grants, the ECAD Initiative has now become self-funding with member subscriptions covering the cost of maintenance and administrative support. The software portfolio is however still able to evolve to meet the changing needs of users because software suppliers continue to recognise the commercial value of their products being exposed to successive generations of undergraduates and postgraduate researchers. The uptake of particular packages is thus constantly monitored to ensure that value for money is obtained in the maintenance paid to suppliers. User requirements are regularly canvassed to determine future inclusions for the software portfolio. Indeed the currently stable membership of the scheme is an indication of the members' commitment to continuation, fostered by regular user meetings and the thrice yearly publication of ECAD User Group News. Recently an entry on the NISS Bulletin Board under section Dl2 has been established for members to access the latest ECAD information via JANET. It is hoped to extend the use of the JANET network more widely in the near future for the interchange of software and application data. With this in mind a database is being compiled of Email addresses for sites who wish to participate in a group evaluation of the benefits of network communications using the ECAD Lead Site at Manchester as the coordination centre.
Provided an institution is already a member of the scheme, it is entitled to free licences to run a variety of ECAD packages on any number or combination of hardware platforms at no extra cost. Indeed, one of our members is running HSPICE on a CRAY. More usually however, the software is supplied to run on SUN, V AX and HP-Apollo workstations. Thus if you have not already heard of the ECAD Initiative and wish to learn more, contact me to send you the ECAD brochure and newsletter, also to identify the ECAD representative at your site.
JFIT is the SERC/DTI Joint Framework for Information Technology. One of its activities has been to sponsor a Club open to all those interested in communications between computers and distributing systems and applications across many computers. The main aims are to stimulate contacts between members of the UK R&D community and to act as a point of contact with other organisations - both UK and international - in the same field. The Chairman of the Steering Committee is Professor Doug Shepherd of Lancaster University. He has recently sent out the following note about the Club's programme for 1990.
One of the main aims of the CDS Club is to bring together people working in the fields of Telecommunications and IT. There are now considerable overlaps in areas of interest in the two communities and the amount of software incorporated into the majority of communications products is now very strong. One way to improve this is to provide a forum where members from the two communities can meet informally and discuss topics of mutual interest, and this is the type of atmosphere we hope will be generated in future Club meetings.
Over the next year it is intended to run four Club Meetings in the following areas: ATM, Distributed Systems Architectures, Multimedia, and Signal Processing. Unlike the inaugural meeting which was arranged to attract as wide an attendance as possible it is expected that these meetings will be of a more specialised nature with audiences limited to the region of 50 or so. This should allow much more detailed discussion.
The first of these will be held at Kingsgate House on Thursday 8 March 1990 and will address the issue of whether ATM is the solution to the problem of providing a single multi-service network.
The second meeting will address the issue of what is meant by the term architecture when applied to a distributed system. Architecture is one of those terms that means different things to different people and an aim of this meeting would be to emerge with some consensus about its meaning in the context of distributed systems.
The third meeting will be devoted to the important issue of multi-media systems, especially how the new multi-media workstations that will be appearing towards the end of 1990 can be integrated into existing and future distributed systems. The meeting will have to address the issues of how local and wide area networks will carry the new types of traffic that will be generated, how the existing protocols will have to be enhanced in order to provide suitable performance, and what sort of interface should be presented to the applications programmer.
The last meeting of the year will probably appeal to a narrower range of Club members but nevertheless covers an extremely important part of the Club's remit, namely the area of Signal Processing.
Finally, the CDS Club will seek to pursue and encourage joint activities with other clubs set up under the IED programmes either through joint meetings or the formation of joint SIGs. Already a major topic of mutual interest, that of Computer Supported Cooperative Work (CSCW), has been identified between ourselves and the HCI Club and it is hoped to set up a joint SIG in this area.
The Computing Facilities Committee has approved the sponsorship of a small number of places at each of the EASE Education events for the next year. The sponsorship covers travel and subsistence only, except in the case of courses, etc run on behalf of EASE by the Artificial Intelligence Applications Institute, Edinburgh where the appropriate course fees are also included. Details of the Education Programme are published in the Engineering Computing Newsletter.
Council is convinced of the need to encourage, in the research practice of the engineering disciplines, the possibilities arising from Information Technology Research
SERC Corporate Plan 1989
During the decade 1979-89 the SERC Engineering Board has responded wholeheartedly to the challenges and opportunities presented by the rapid development of IT. The Board's own spend on IT has been enhanced by SERC's participation in the Alvey Programme and the present national collaborative JFIT Programmes. The Board's expenditure on IT research grants through its Information Technology Directorate was some £2Om in 1989/90, some 26% of the Board's grant funding.
The Board. and others, were however unclear about the extent to which the enhanced funding had been effective and established a Working Party under Dr Nigel Kingsley of Eurotherm Ltd with the following terms of reference:
The Working Party reported to the Board in July 1989. It reported that while the initial focus for the enhanced funding was, rightly, on the development of pure IT, the time had come for more emphasis to be placed on the application of IT to engineering and other disciplines. There was little sign of this change of emphasis occurring within the IT community which was perceived by the engineering community to be very narrowly focused. The linkage between the IT research community and the IT industry is very strong but there is very little linkage between the IT research community and the wider engineering research community or engineering industry.
Equally the engineering research community and engineering industry appear to lack a full appreciation of the scope of IT. The IT used by engineering researchers does not often appear to benefit from IT research, usually being based upon standard commercial products and not particularly advanced techniques.
The Working Party identified an awareness gap, both ways, between the IT and broader engineering communities regarding their respective activities and needs - and an applicability gap which reflected the inappropriateness of much of IT research to the needs of the broader engineering communities. The recommendations of the Working Party sought to improve the orientation of IT research and increase awareness of the potential of IT in order to lead to:
Central to the recommendations was the concept of a themes approach for collaborative research projects between the IT and engineering communities. Four initial themes have been agreed for the 15 March closing date:
For the first call, a commitment of up to £2m is envisaged, split equally between ITAB and Engineering Board against the initial themes by 15 March. The proposals will be judged by the relevant sub-committees/committees of both ITAB and the Engineering Board. Other recommendations concerning improvements in awareness and technology transfer are being implemented through subject committees/ITAB and the EASE programme.
Dr Kingsley is expected to attend the EASE Education and Awareness Programme Symposium at UMIST on 18-20 April to present the findings and recommendations within the Kingsley Report. Copies of the Kingsley Report and details of the four initial themes can be obtained from me.
The Leicester Image Processing Suite (LIPS) is a series of modules developed with the specific aim of teaching the principles of digital image processing. Developed at the University of Leicester by a multi-disciplinary team from the Geography, Geology and Physics departments, it has been funded by the Computers in Teaching Initiative (CTI) and is available free (including source code) to all UK universities. Student documentation and work sheets are also available in an ASCII format.
So far, seven modules are available covering the following topics:
Two further modules covering Multi-Spectral Classification, and Image Degradation and Restoration, are under final development.
The package is written in FORTRAN77 using GKS as the graphical interface. The system is currently in use with Sigmex 6100 series graphics terminals and their proprietary GKS (called WKS), and will port immediately to any site with this kit. Interested parties with GKS facilities on other hardware should contact us first outlining their specifications, as we have previously experienced problems mainly due to the high specification needed in respect of the colour table.
Introduction to Object Oriented Programming Concepts March 26th
UNIX Introduction March 27th-29th
Introduction to Object Oriented Programming using Smalltalk-80 March 27th-29th
Sun System Administration March 29th
The C Programming Language April 3rd-5th
C++ Programming April 3rd-5th
Advanced Course in Smalltalk-80 April 9th-10th
An Overview of Pure Functional Programming April 9th
An Overview of Semi-Formal Methods April 10th
An Overview of Artificial Intelligence and Knowledge Engineering April 11th
From A to Z : An Introduction to Formal Specification April l8th-20th
Introduction to Formal Methods using VDM April 24th-26th
UNIX Application Builders May 29th-3lst
The AI Support for Engineers project is funded by Engineering Board to provide support for SERC-funded engineers who wish to investigate the application and applicability of AI to their work. The support services are provided by the Artificial Intelligence Applications Institute (AIAI) at the act as a technology transfer unit, facilitating the use of AI tools and techniques on real problems by communities that are not AI specialists. This support covers a wide range of activities and here is a summary of some of the services that are available.
Getting started with AI tools and techniques can be a major problem. Although a large selection of products now exists, making an appropriate choice is difficult Facilities are available at AIAI for visitors to come to evaluate, before purchasing, many of the most useful tools under the guidance of project staff. More intensive support is available through the Visitor Scheme. This enables researchers to work in collaboration with staff at AIAI for a period of several weeks, with access to computer facilities, our AI library, and technical support.
Workshops and seminars are organised periodically so that engineers can meet to discuss their common requirements and share their experiences. Our next one day seminar, which will be on Integrating Knowledge Based and Conventional Systems, is scheduled for Friday 11th May 1990. This one day seminar looks at the interface facilities provided by some of the current AI tools and languages. The systems covered will include Crystal, Goldworks, KEE, Leonardo, Edinburgh Prolog and Xi Plus. AutoCAD, a popular CAD package, with its own embedded version of Lisp - AutoLisp - will also be considered. Demonstration of some of these systems will be provided.
Following the format of our previous seminars, we would like to limit the number of attendees to the Computing Facilities Committee of the around thirty. If you are interested in attending, please let us know as soon as possible by filling in and sending back the registration form distributed together with this issue of the Newsletter.
The following short courses on AI are provided University of Edinburgh. The purpose of AIAI is to free of charge to suitable applicants who are working on projects funded by SERC grants. The course schedule for 1990 is:
Another presentation of the course for new users of EASE Sun Workstations is planned for Wednesday April 4 and Thursday April 5 1990 at Sheffield Polytechnic. This is in the same location as the EASE Education Programme seminar Computer Algebra for Engineers to be held on Tuesday April 3.
The first day offers a basic introduction to using SunOS for the benefit of users with limited experience of the UNIX operating system. The principal aim of the second day is to introduce the concepts and skills involved in working with window based software. It offers attenders an opportunity to acquire and practise these skills through the use of some general purpose tools.
EE Computing & Control Division Colloquium:Cybernetics Today: 6 March 1990, IEE, London
UK IT 1990 Conference 19-22 March 1990, University of Southampton
ALP-UK90 Conference, UK Association for Logic Programming 28-30 March 1990, University of Bristol
On the first day there will be two tutorial sessions in parallel, which can be attended separately from the main Conference. The rest of the Conference will consist of single sessions. Invited speakers will give survey talks on subjects of topical interest and the remainder of the Conference will be devoted to contributed papers. Tutorial speakers: Richard O'Keefe, Royal Melbourne Inst of Tech, Keith Clarke, Imperial College. Invited Speakers: Chris Mellish, University of Edinburgh, David Warren, University of Bristol, Bob Kowalski, Imperial College, Pascal Vanhentenryck, ECRC, Munich
The 1st International Conference Sponsored by the British Computer Society: The Many Faces of Design: Computing as an aid to innovation and design" 9-12 October 1990, Nottingham Conference
Themes include:
Prof P M Dew, University of Leeds, P Cornwell, Meiko and Prof M J Wozny. RPI will all be making presentations in this session.
Presenters: Dr M Susan Bloor, University of Leeds, Mr J Wix, Wix McLelland Ltd, Dr J Fowler, CADDETC, Dr A Watson, University of Leeds, Prof K MacCallum, University of Strathclyde
The role of product data in engineering infonnation management is increasingly recognised.
The evolving understanding of what is product data is coming from the STEP activity, which aims at data exchange between dissimilar systems, from research projects investigating integration mechanisms and from company needs for configuration management.
Using examples from ABC, the speakers will present
Presenters: Dr C P Wadsworth and Mr D J Johnston Rutherford Appleton Laboratory
Presenter: Dr K Robinson, Rutherford Appleton Laboratory
This tutorial aims to give the non-expert an overview of the X Window System, its capabilities and drawbacks, and where to go for further information. Participants will be assumed to have had some exposure to a window system, though not necessarily to have programmed in such an environment.
Presenters: Mr J Gallop, Miss J Haswell and Mrs R Popovic, Rutherford Appleton Laboratory
Presenters: Dr P Chung and Dr Terri Lydiard, Artificial Intelligence Applications Institute, University of Edinburgh
In recent years, as computers have become ever more breathtakingly powerful it has been realised that many computer tools and techniques employed by researchers in artificial intelligence (AI) could have great practical utility. Conventional computing techniques have been developed mainly for implementing algorithms for manipulating large quantities of numerically oriented data, and data structures like arrays are good for representing matrices and structures of fixed sizes and types. AI workers, however, have been more concerned with representing descriptions of things - their attributes and the relationships between them. They have investigated areas for which algorithmic solutions are not know, and have had to formulate ways of dealing with inexact knowledge and rules of thumb. These tools and techniques, in particular knowledge based tools and techniques, developed by AI workers are very relevant to engineers because, in engineering, there are a lot of problems that do not fit into the strict rigid framework of algorithmic solution. In practice, engineers are always having to make decisions that are based on their accumulated knowledge and experience. They need to weigh up various pieces of information, consider the constraints that are influencing the current problem, and then decide on a particular action. This decision making process occurs in the design, construction and maintenance of all kinds of engineering products in all areas of engineering.
This tutorial will introduce participants to various practical aspects of building knowledge based systems. The topics covered will include:
