Herbert M Sauro wrote:
> In article <33872069.41C6 at henri.chem.uleth.ca>, Marc Roussel
> <roussel at henri.chem.uleth.ca> writes
> >Bob Phair wrote:
> >> My point is that everyone will be forced to move to faster machines
> >> and as soon as they do, they will surely move to 32-bit operating
> >> systems. Don't look back! 64-bit systems will probably not be far
> >> behind.
> >
> >I don't know how to break this to you, but 64-bit hardware and
> >operating systems is old news to people using Digital Alpha hardware
> >with Digital Unix. Most of the hardware running Unix these days (if
> >we exclude Linux) is 64-bit hardware. Digital is far out in front of
> >everyone else (except maybe HP; I wouldn't know not having touched
> >one of their machines in years) as far as the OS is concerned, but
> >the other Unix vendors are moving in that direction rapidly.
> >
> What you say is true and exciting but not everyone has access to a 64-
> bit computer (unfortunately), but I think most people have access to a
> PC and most probably a Pentium. [...]
> What I'm looking for is the platform which has the widest possible
> circulation but maximum computer power, and I think these days it's
> probably Win95. High schools and colleges in the UK still have a high
> proportion of Windows 3.1 machines so software for them has to be
> 16-bit [...]
I was just tweaking your noses (and trying to back up Bob's point that
64-bit computing is just around the corner) and here you are taking me
seriously! Very well, let me inject a few serious words into this
conversation. This Win3.1/16bit-Win95/32bit dichotomy is, I believe,
false. (I didn't make this point earlier because I thought someone else
would.) Isn't it possible to run 32 bit software under Win3.1 given the
appropriate hardware and a special DLL? If that's the case, it seems to
me that the thing to do is to develop for 32-bit Win3.1 and to test both
with that platform and with Win95.
> I think writing software for Unix machines must be difficult but I
> don't know for certain. One couldn't just release the executables
> since there are so many different binary incompatible machines out
> there.
That's right. Fortunately, if you're careful, you can write portable
code fairly easily. However, if you want maximum distribution, you
probably want to start with Windows machines. Then Macs. Unix (sadly)
last.
Another option is to buy a cross-platform development toolkit.
However, that might be fairly costly.
Marc R. Roussel
(roussel at henri.chem.uleth.ca)
Department of Chemistry
University of Lethbridge
