Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!caen.engin.umich.edu!ejd
From: ejd@caen.engin.umich.edu (Edward J Driscoll)
Newsgroups: comp.lang.c
Subject: Re: Portability and the Ivory Tower (was Re: Book on Microsoft C)
Message-ID: <425bf40d.b11a@falcon.engin.umich.edu>
Date: 31 Mar 89 21:41:00 GMT
References: <754@oravax.UUCP> <225800146@uxe.cso.uiuc.edu> <9937@smoke.BRL.MIL> <424ce87c.b11a@falcon.engin.umich.edu> <28587@ucbvax.BERKELEY.EDU> <4255794d.b11a@falcon.engin.umich.edu> <8079@chinet.chi.il.us>
Reply-To: ejd@caen.engin.umich.edu (Edward J Driscoll)
Organization: caen
Lines: 36

In article <8079@chinet.chi.il.us> les@chinet.chi.il.us (Leslie Mikesell) writes:
>In article <4255794d.b11a@falcon.engin.umich.edu> ejd@caen.engin.umich.edu (Edward J Driscoll) writes:
>
>>[my defense of machine-dependant code, you've seen it]
>
>Are you prepared to either abandon these applications when the machine
>becomes obsolete or be forced to purchase something that offers
>backwards compatibility?  If you can afford to abandon it, then the
>application (and whatever advantage it gains from machine-specific
>code) must not be very important.  I recently read that computers
>have an average life of about 5 years but the same applications
>are maintained for 10-15 years.  
>
>Les Mikesell

Your point is certainly worth taking into consideration.  My
reply question would be:  Are you willing to stick with
teletype compatibility forever?  In all honesty, if the
application is that valuable then the odds are good that I 
would be willing to hold out for backward-compatible hardware.
You also left out the option of changing the code to
exploit the new hardware.  If applications do not take
advantage of the features of the hardware, then why pay
for that hardware?  You seem to be claiming that it is
important for software to be backwards-compatible, but
not for hardware.  If the software outlasts the hardware
by a factor of three, I would think the reverse would be
true.

[PS:  We seem to be drifting a little from C.  Should
we move this over to the software engineering newsgroup?]

-- 
Ed Driscoll
The University of Michigan
ejd@caen.engin.umich.edu