bertrand@eiffel.UUCP (Bertrand Meyer) (01/12/91)
This message is intended to correct a widespread but incorrect impression: the belief that the Eiffel language is ``proprietary''. A typical phrasing may be found in the book by Winblad, Edwards, King (O-O Software, Addison-Wesley), which concludes its discussion of Eiffel with: ``Eiffel's robustness as an object-oriented language is offset by its proprietary nature.'' The ``proprietary nature'' part of this statement is incorrect. The non-proprietariness of the Eiffel language design has always been stated clearly, and was reinforced by the recent creation of the Nonprofit International Consortium for Eiffel, which will control the evolution of the language. (We have written to the authors and publisher of the above book to ask them to correct the error.) One aspect that may have contributed to the misunderstanding was the registration of the trademark Eiffel by its original designers (Interactive Software Engineering). However this was never meant to prevent others from implementing Eiffel. The trademark was necessary simply to avoid a situation where others would have trademarked the name and then tried to prevent anyone else from using it - a very real possibility, which in fact did occur in one foreign country. The trademark is being transferred to the Consortium, which will grant its usage to other parties based solely on technical criteria (e.g. conformance to certain rules, validation suites when they exist etc.), without commercial considerations. Of course Interactive's implementation is not in the public domain and remains Interactive's property. In the past this has prompted the somewhat cynical comment ``sure, but this is the only available implementation, so what difference does it make to have the language itself be non-proprietary?''. A big difference. The days of single-implementation Eiffel are numbered. We know of about half a dozen current implementation efforts; some are quite advanced, other still at early stages. (Actually the most recent piece of news came from someone who told me in Seattle, the day before yesterday, that he was writing yet another implementation.) Several of these should yield products in 1991. In particular, a DOS implementation, developed by a European company, will be on exhibit at one of the biggest computer shows in the world, CeBIT, Hannover, in March. (This is NOT the DOS implementation being developed by TS-Controls in collaboration with Interactive.) We have only little information on the other ongoing efforts (after all, they are future competitors), but I am sure more will be announced in due time. I hope this puts the proprietariness myth to rest. -- -- Bertrand Meyer bertrand@eiffel.com