Xref: utzoo comp.lang.eiffel:759 comp.software-eng:3108 comp.object:1034 Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!swrinde!ucsd!hub!eiffel!bertrand From: bertrand@eiffel.UUCP (Bertrand Meyer) Newsgroups: comp.lang.eiffel,comp.software-eng,comp.object Subject: The Eiffel Consortium (Part 1: Rationale) Keywords: Towards an industry standard Message-ID: <261@eiffel.UUCP> Date: 8 Mar 90 02:33:38 GMT Organization: Interactive Software Engineering, Santa Barbara CA Lines: 163 THE EIFFEL CONSORTIUM Part 1: Rationale This message and a companion one contain an important announcement on the evolution of Eiffel - and, we think, of the discipline of software engineering. The first message (this one) describes the rationale behind the announcement. The second message gives practical details. BACKGROUND There is a widespread and growing recognition that the Eiffel programming language and environment hold the potential for a dramatic improvement in the quality of software being developed routinely by the industry, and in the productivity of those who develop it. Although Eiffel is being used worldwide by corporations and universities for highly diverse projects, its progress as an industry standard is still subject to concerns about Eiffel's perceived proprietary nature. These concerns stem from Eiffel's origin as a product from one specific company (Interactive Software Engineering). This situation is particularly ironic when assessed against the excitement that Eiffel generates among most people who have looked at it carefully, against the enthusiasm of its users, and against Interactive's own attitude, which long ago put the language in the public domain and has consistently encouraged others to come up with competitive implementations. Our position has been stated loud and clear: although we do expect to profit from the success of Eiffel, our primary interest is in making the technology succeed on a broad scale. We want to be one of the players, but not the only one. The more the merrier. To make this commitment to diversity and competition even more of a reality, a major customer came up with a suggestion: creating an Eiffel Consortium - in the style illustrated by the X Consortium, the Open Software Foundation or X Open. We enthusiastically agreed to participate. Since it was first broached, the idea has been discussed and refined through numerous private e-mail and fax messages and is now long past the ``what if?'' stage. The Consortium is on its way to being formed. This message is aimed at making a wider community aware of what we think will be a milestone in the progress of software engineering, and to enable anyone who has not yet been contacted to join in the initial group of consortium founders. CONSORTIUM ROLE The Eiffel Consortium will be open to companies, research institutes and educational institutions interested in using and advancing Eiffel technology. The Consortium will serve as a catalyst for the widespread acceptability and long-term success of the language, libraries, tools and method. Thanks to the consortium, the progress of Eiffel will no longer be tied to the success of a single company. Another key advantage of the Consortium will be to enable users of Eiffel, especially companies which have made a significant investment in the technology, to exert a significant degree of control over its evolution. Finally, the formation of the Consortium should prove to be the signal that many software developers are expecting to start building what is one of the major aims of Eiffel: an industry of efficient, reliable, reusable software components available from many different sources. This is a unique opportunity both for large corporations and for smaller ventures such as consulting houses. CONSORTIUM ACTIVITIES The precise charter of the Consortium will be finalized during the two planning meetings (see companion message for details). Without anticipating on the result of these meetings, it is not difficult to list a few likely activities. The Consortium may have committees on such subjects as library standardization and language standardization. It will collect dues from its members and will be able to spend the money on projects of general applicability; examples of a project with immediate applicability are a public-domain Eiffel interpreter and a validation suite for conforming implementations. The Consortium could also work with official standards bodies such as ANSI and ISO. Finally, members will be able to make contributions to the Consortium (such as libraries and tools) and get back other contributions made by their fellow members. INTERACTIVE'S CONTRIBUTION. Although, as mentioned above, the idea did not originate with Interactive Software Engineering, we find it sufficiently attractive that: 1 + We are willing to transfer the registered trademark `Eiffel' to the Consortium. 2 + We are also prepared to release full control of the language evolution to the Consortium. This could take effect as early as April 16 (at a time when the book ``Eiffel: The Language'', providing a precise description of the standard language, will be published by Prentice-Hall or in press). 3 + We are considering giving to the Consortium, as part of our initial contribution, some of the current Eiffel Libraries, especially the Data Structure Library, which plays a fundamental role in Eiffel software construction. All this, especially 2, may be viewed as a major risk for Interactive. We have enough confidence, however, in the maturity of the Eiffel user community (as evidenced in particular by the quality of the discussions in comp.lang.eiffel), and in the potential benefits outlined above, to be ready to take this step. Given the enthusiasm with which the draft of this message was received when we circulated it privately, there appears to be no obstacle in moving the idea to reality quickly. In fact, the consortium could be operational by April 15. Organizational details are given in the accompanying message. Any further comment will be welcome. -- Bertrand Meyer bertrand@eiffel.com Note: Having received many interesting replies to earlier e-mail messages, I considered posting a selection. On further reflection, however, it does not appear fair to quote the correspondents anonymously, and the procedure of asking each of them for permission to quote by name is too cumbersome. So if you sent me some initial comments and feel they are of interest to a wider group, please post them. -- Bertrand Meyer bertrand@eiffel.com