Team Eiffel

The first official meeting of NICE, the Non-profit International Consortium for Eiffel, took place in Paris on Friday, 11th March 1991. It was a small and friendly meeting, involving representatives from Interactive Software Engineering, Tetra Ltd, Enea Data, as well as Dr. Bertrand Meyer. This is not the total membership of the consortium, which at the current time also includes BNR (Canada), and Telecom Australia; however, these companies were not represented at the Paris meeting.

Several key issues were agreed by the members present, and a number of actions tabled for urgent work. These included agreeing to accept the book "Eiffel: the Language (version 3)" as the definitive starting point for the language. The book is currently in final draft, but is expected to be ready for publication in a matter of weeks. ISE also agreed to transfer the ownership of the trademark "Eiffel" to NICE with immediate effect, thereby confirming the commitment to establish Eiffel as a non-proprietary language.

The meeting agreed to accept Bertrand Meyer as a lifetime honorary charter member, in recognition of his work in designing Eiffel and promoting the entire philosophy of software construction which Eiffel embodies. In this capacity he represents himself, not Interactive Software Engineering, his company. ISE is itself represented, and has an honorary membership only for a limited period. This is to compensate for both the costs of launching the consortium, and the transfer of rights to the trademark "Eiffel" and various other copyrights. It would be easy to be cynical, and believe that NICE is just a promotional exercise on the part of ISE. I can state categorically that it is not, that neither Dr. Meyer nor ISE have any special priveleges outside honorary membership, and that the other members are committed to ensuring that NICE is both independent and fair.

On technical issues, a committee was established to examine the problems of interfaces between Eiffel and other languages, e.g. "C". This was as a direct result of matters raised during the Eiffel User Group meeting only a few hours earlier, during which it became clear that the C language interface defined in the current definition of Eiffel is not sufficiently detailed for large-scale software projects. With the imminent prospect of other implementations of Eiffel becoming available, the members of NICE appreciated the urgent need to define an interface to C which is usable for major software projects, portable between implementations, and which does not hinder technical innovation in the development of Eiffel compilers. The technical committee will consult with all interested parties, especially those developing compilers, before making a final recommendation.

Further technical work of a high priority was agreed to be the definition of an external portable object format, allowing the interchange of object structures between different platforms and implementations.

The meeting also agreed on the need to have some form of validation suite in order to be able to certify implementations of Eiffel. This will be solicited via some form of tendering process, since it will be more effective to buy-in the required technology. A more specific invitation to tender will be published in due course; the principle reason for the delay is that the validation suite will need to test conformance to Eiffel version 3, and the definitive book for that version is not yet published.

The constitution of NICE prevented the formal election of a board and officers, since not all members were represented. This will be completed over the next few weeks via post or electronic mail. To enable the consortium's work to continue, temporary acting appointments of Chairman and Treasurer were made.

NICE is an independent organisation, and is determined to be an active and effective one. Although the consortium was created by ISE, it does not represent their interests and now operates independently. This is illustrated by the technical committee on C interfaces, which will not favour details of ISE's implementation over anyone else's. NICE will consult as widely as possible with the Eiffel user community over many issues, especially technical ones. The principle vehicles for announcements from NICE will be the Eiffel User Group newsletter, the independent newsletter Eiffel Outlook, and the USENET newsgroup comp.lang.eiffel. If you use Eiffel, you will want to know what is happening in NICE, and we are always interested in your views and other feedback. If you are serious about Eiffel, either as a committed user organisation, or as a developer of Eiffel compilers, tools, or libraries, you should consider joining the consortium. This will not only give you much closer insight into what is happening with Eiffel, but will also enable you to have a significant influence in shaping the future direction of the language.

I believe that 1991 may well be a watershed for Eiffel, with the much-improved version 3 language definition, competitive implementations across different platforms, and an increasing awareness of the language's ability to deal effectively with difficult problems in object-oriented design and programming. In the midst of this, NICE has the responsibility to act as an independent reference over what can be claimed to be Eiffel, as well as defining progressive enhancements through a consultative process.

The computer business of the 1990s is acutely aware of the need not only for standards, but for standards which do not stifle competition or true technical progress. Many "standards" are in fact proprietary, de-facto standards which give little scope for development by other than one company. The true de-jure public standards have often been the result of painful pulling-together of disparate threads, involving protracted arguments, and frequently resulting in less than ideal compromises. NICE is perhaps unique in this context, being a body with no commercial allegiance, and with the goal of building an evolving public standard before any divergence takes place. The result could be one of the most powerful, portable, and usable programming systems in the history of computing. That is my commitment to NICE.

posted by Roger Browne @ 12:15