🚀 go-pugleaf

In article <1i86t8INNc5b@cs.widener.edu> Rob Young,
YOUNG@tattoo.cs.widener.edu writes:
>I agree with wayne's assertion that it is a marketing thing.  Leads me
to the
>question:  If you have a *PC* that outperforms most workstations (and
also
>can run Unix and jump through all your hoops above), does that make it a 
>workstation?  Couldn't a company call something a PC when it is really a 
>workstation in PC clothing?

What about a PC that outperforms a mainframe (or an AS/400)?

It's a question of the use to which it's put and how it's configured.

Other than that it still needs a lot of work to set it up because
it's still going through debugging etc., Linux installed on a 486
with a decent graphics card and ethernet would meet most of the
definitions posted on this thread. Not bad for free software and
pretty cheap hardware.

Specifics like "real operating system" (= UNIX?) etc. are not important.

When you use a workstation, most of the computational resource is
dedicated to giving you good response time and you are not inhibited in
accessing networked resources (files, computation, etc.). The key is
the user-centredness: minimum inhibition in getting work done. Detail
such as TCP/IP, user interface, big screen etc. is devoted to this
goal. If you have all these things and the system is badly designed
so you can't get effective work done, it's not a workstation. If some
detail is wrong but it still lets you get work done fast, maybe it
still is one.

The whole thing is so subjective that no one is every going to define
the concept precisely. But don't worry. The current divisions between
classes of computer are crumbling so fast that this whole discussion
will soon be irrelevant.
--
Philip Machanick                   philip@concave.cs.wits.ac.za
Department of Computer Science, University of the Witwatersrand
2050 Wits, South Africa
phone 27(11)716-3759 fax 27(11)339-7965