NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/security/misc/2348

Internet Message Format | 1992-12-21 | 1.7 KB

Xref: sparky comp.security.misc:2348 comp.org.eff.talk:7849
Newsgroups: comp.security.misc,comp.org.eff.talk
Path: sparky!uunet!ferkel.ucsb.edu!taco!rock!stanford.edu!agate!spool.mu.edu!uwm.edu!psuvax1!news.cc.swarthmore.edu!ralph.cs.haverford.edu!eoliver
From: [e-mail] (Erik Oliver)
Subject: Re: Stupid Licenses (YUCK!)
Message-ID: <[e-mail]>
Sender: [e-mail] (USENET News System)
Nntp-Posting-Host: ralph.cs.haverford.edu
Organization: Haverford College Computer Science Department
References: <bhayden.724690634@teal> <[e-mail]> <bhayden.724865911@teal>
Date: Sun, 20 Dec 1992 18:53:28 GMT
Lines: 21

Proving software correct is extremely difficult to do, and making sure
that there are no "bugs" is even harder.  For example, in the
IBM-Compatible world, it is very hards to anticipate all of the possible
setups that might have an incompatibility with your software.  If my new
product XYZ-Works doesn't work on a SuperDuper486DX/2 is that
necessarily reflective of a bug in my software or a glitch in some part
of the SuperDuper product?

Because of the variety of IBM-Compatible products out there testing for
every configuration is unrealistic, if not impossible.  Even in the Mac
world, there are problems that occur because of interactions between
INITs and CDEVs with applications.  Are these bugs?

I think that if the manufacturers make good faith efforts to keep their
product working and put out bug releases that is about as much as we can
expect.  Personally I think that the software produced by organizations
like the FSF is better than a lot of commerical stuff because they are
not afraid to keep updating their product and release slews of fixes as
necessary.

-Erik