Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!usc!ucsd!hub!eiffel!bertrand From: bertrand@eiffel.UUCP (Bertrand Meyer) Newsgroups: comp.lang.eiffel Subject: Re: Eiffel and the NeXT Keywords: Eiffel, NeXT Message-ID: <221@eiffel.UUCP> Date: 11 Jan 90 03:00:05 GMT References: <6058@mentor.cc.purdue.edu> Organization: Interactive Software Engineering, Santa Barbara CA Lines: 44 In <6058@mentor.cc.purdue.edu>, g2k@mentor.cc.purdue.edu (Frederic Giacometti) raises the question of Eiffel on the NeXT. In , elkins@topaz.rutgers.edu (George Elkins) adds: > [...] I would take NeXT more seriously if an Eiffel > compiler and Eiffel class libraries were > available. Is ISE considering this possibility? Here is a brief statement of the ``party line'' at Interactive Software Engineering. We support about 30 platforms (which, by the way, is driving us crazy, but I'll comment on that some other time). At the moment the NeXT is not one of them. Because we constantly get requests (from either computer vendors or potential users) to port to new machines, we must be careful about which ports we decide to undertake. Three major factors are involved in each decision: A. Technical interest. Some ports are an intellectually rewarding experience, others a pain in the neck. B. Market share. Obviously the number of units of a given platform out there, either current or projected, is a point to consider before we invest in a significant porting effort. Customer pressure also falls under this heading. C. Relationship with the computer vendor. Some vendors understand the value of good software and we have had very fruitful cooperations. I am eager to see Eiffel on the NeXT. Apart from all other good reasons, the presence of Mach, mentioned by Mr. Giacometti, is particularly enticing because it seems to fit well with the recently designed Eiffel mechanism for concurrency (presented in a few talks but not yet published), about which I am quite excited. A straightforward port of the basic facilities is a relatively small job. A version that will take advantage of all the specific NeXT facilities is a bigger task. We do have to consider all of the above criteria, however. If I were to rate the NeXT on a scale of 1 to 10 for each of them, I would give it 10 for A and unknown for B (most people would agree that the jury is still out). By the way, anyone who thinks that Interactive is shamefully dragging its feet on the NeXT is welcome to go ahead and write a compiler. -- Bertrand Meyer bertrand@eiffel.com