home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!elroy.jpl.nasa.gov!ames!agate!ucbvax!efftoo.boeing.com!crispen
- From: crispen@efftoo.boeing.com (crispen)
- Newsgroups: comp.lang.ada
- Subject: MI last time
- Message-ID: <9208211316.AA12657@efftoo.boeing.com>
- Date: 21 Aug 92 13:16:24 GMT
- Sender: daemon@ucbvax.BERKELEY.EDU
- Distribution: world
- Organization: The Internet
- Lines: 39
-
- wellerd@ajpo.sei.cmu.edu (David Weller) says:
-
- > I believe that MI is useful, but I also
- >believe that the 9X solution is reasonable. As Grady Booch has said,
- >multiple inheritance is like a parachute -- you don't always need it,
- >but when you do, you're quite happy to have it.
-
- And Peter Herman follows up:
-
- >correct. You need a parachute in the worst case only.
-
- What I'm trying to find out is in what domain a parachute is a reasonable
- thing to have. That is, I do accept the need for domain-specific
- architectures. The only question, then, is whether there exists at
- all a problem domain for which MI is a necessary component of a
- sensible architecture.
-
- I have heard from one person who believes that MI helped in the
- windowing systems of Symbolics and LMI Lisp machines, and I certainly
- am not knowledgeable enough about that domain to know if MI was a real
- help or not (I will post a summary of email responses to me, to save
- traffic in this group, BTW). This is the domain I have always seen as
- an example when MI is touted.
-
- So, is MI worthwhile in any other domains? Or was MI even worthwhile
- in that domain?
-
- Dave also asks:
-
- >Have you crossposted to comp.lang.c++ and comp.object?
-
- I have asked one person to crosspost my remarks to comp.lang.c++
- since I only have mailing list access to the Ada newsgroup and don't
- have write access to Usenet news. If it doesn't show up there after
- a while, I'll ask someone else. I would rather not have everybody
- crosspost it for me, though ;-)
-
- Bob Crispen
- crispen@foxy.boeing.com
-