Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!decwrl!mcnc!rti!bcw
From: bcw@rti.rti.org (Bruce Wright)
Newsgroups: comp.windows.ms
Subject: Re: OS/2 2.0 is here!  READ THIS, you'll be impressed
Summary: Not so clear ...
Message-ID: <1991Apr23.160455.3142@rti.rti.org>
Date: 23 Apr 91 16:04:55 GMT
References: <1435@caslon.cs.arizona.edu> <F0wR11w164w@questor.wimsey.bc.ca> <1103@opus.NMSU.Edu>
Organization: Research Triangle Institute, RTP, NC
Lines: 45

In article <1103@opus.NMSU.Edu>, sitze@nmsu.edu (Richard Sitze) writes:
> Yes, but on the OTHER hand I bought into a DOS machine in the
> first place strictly for developing software products targeted
> towards the small business...  And I'd DO ALMOST ANYTHING for
> a decent DOS development platform.  I'm sorry, single DOS windows
> under earlier versions of OS/2 (and UNIX flavors) just don't do
> it for me.  Windows 3.0 is 'close' but to slow for most things.
> 
> I think any system that can provide the developement environment
> will (sooner or later) find it's place in almost any market, take
> a look at UNIX from that viewpoint...

I'm not so sanguine that any OS with a reasonable development
environment will make it.  There's more than that that's
required to make it in the marketplace.  Maybe I've seen
too many "nice development environments" fail miserably
because of one reason or another ... Concurrent CP/M,
Concurrent DOS, CTOS, ... the list goes on and on.  OK, so
they may not be very nice by modern standards, but at the
time they came out they were, by comparison with many of
the alternatives, very attractive systems.  But it was the
alternatives that succeeded - often for reasons that had
relatively little to do with narrow issues of technical
merit.

John Dvorak recently (PC magazine, 30 April 1991, p. 83) 
made the comment that systems that were successful were
easy for the end-user and hard for the developer.  I think
this is looking at things the wrong way:  I think that the
issue, at least for mass-market items, is how easy things 
are for the end-user;  how easy (or hard) they are for the
developer is not very relevant - the market doesn't really
care how easy or hard it is.  If anything, a nicer system
for development can promote nicer end-user products, but
it doesn't guarantee them and it's hardly required for
making them.  There's also no reason that software has to 
be run on the system used to develop it - but whether a 
system is popular with developers is hardly a guarantee 
of success.

But this is just talking about technical issues, which I 
feel are only a part of the equation.

					Bruce C. Wright