Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!decwrl!labrea!agate!pasteur!ames!lll-lcc!well!shf From: shf@well.UUCP (Stuart H. Ferguson) Newsgroups: comp.sys.amiga Subject: Re: IPC - Generality, etc. Message-ID: <5704@well.UUCP> Date: 15 Apr 88 22:53:28 GMT References: <5693@well.UUCP> <674@imagine.PAWL.RPI.EDU> Reply-To: shf@well.UUCP (Stuart H. Ferguson) Organization: The Blue Planet Lines: 67 Keywords: Object-Oriented, IPC, Standard, yeahbut >>I hope that this clears up some of the confusion about the >>object-oriented approach I'm pushing for. It's not a single-minded and >>specific idea, but rather it is a very general model which provides a >>wide range of interpretations and uses. The object-oriented approach >>will require different kinds of implementations than other methods, but >>it is at least as general, if not more general, than anything else >>proposed so far. > I think this is a very nice idea, and with a good set of servers >turns the amiga into a real nice object-oriented system. Sort of like >a Super-ARexx. (I'm sure Arexx could be made to work well in such an envirion- >ment also.) Yes, there are a couple of ways I can think of right off the top of my head to incorporate ARexx compatable programs into an object-oriented IPC standard. A quick and dirty way would be to have ARexx compatable programs run inside the main system as REXX class objects with the main ARexx program running as the object-class server. Programs using the standard (object-oriented) IPC would send commands to these REXX class objects (which are actually each a program with an ARexx interface) and the commands would be serviced by the ARexx program and passed on to the programs referenced by the objects. A better way would be to enhance the ARexx language to allow users to register the services of their programs and scripts directly into the global service tables. ARexx would then be a scripting language inside the object-oriented IPC scheme. Either way, the ARexx program would have to be modified. I have to argue, however, that an object-oriented IPC such as I'm proposing does not make the Amiga into a real object-oriented system. I've just borrowed some of the concepts and ignored others in order to make the idea work on the Amiga as it is today. Important concepts such as data-hiding, for example. Since objects may be implemented by several servers possibily written by several different authors, the object description must be public, and cannot be allowed to change radically from implementation to implementation. Class heirarchies will have to be done in a somewhat strange way as well for the same reasons. You also wouldn't want to write object-oriented programs using this scheme either, since the task-switching overhead would probably kill you. But for a number of medium to large sized programs chattering among themselves, the object-oriented approach gives them a simple, powerful common language to use which is abstract enough that anyone can break into the conversation, or can use parts of another system for something completely different. > Now all we need to do is make Workbench a set of servers and a user- >interface. Then it would be really object oriented, and could be extended >at will! An idea after my own heart! In fact, the whole IPC idea came about as a result of my design for a user-extensible Workbench. I know that my ideas are getting across when other people start to invent them on their own :-). Needless to say, I have plenty of ideas about how to fit this scheme into a new Workbench. > Will you be at the Devcon? No, probably not. I've already shot the travel budget this year. > // Randell Jesup Lunge Software Development > // Dedicated Amiga Programmer 13 Frear Ave, Troy, NY 12180 > \\// beowulf!lunge!jesup@steinmetz.UUCP (518) 272-2942 > \/ (uunet!steinmetz!beowulf!lunge!jesup) BIX: rjesup >(-: The Few, The Proud, The Architects of the RPM40 40MIPS CMOS Micro :-) -- Stuart Ferguson (shf@well.UUCP) Action by HAVOC (shf@Solar.Stanford.EDU)