🚀 go-pugleaf

In article <1993Jan3.052930.27869@wam.umd.edu> rsrodger@wam.umd.edu (Yamanari) writes:
>  
> 	Can anyone give me a definition of workstation, other than 
> 	cost, that seperates it from modern high-end PC systems (in his
> 	case, one that cost no more than $3500).


This is almost certain to cause a long string of definitions, and
quite possible a flame war or two.


Why?


Well, other than that's what usually happens on Usenet, it's because
there has never been a clear definition of what a workstation is.
What definition there is, has been defined more by the marking folks
than by anyone else.  There have been long discussions about what the
difference between a mainframe and a minicomputer and a microcomputer
are.  What separates a PC from a workstation is much less well
defined.


So, instead of trying to give one exact definition, I am just going to
list a few general items that I have heard people use.  (Chances are,
you won't think that all of these are good qualifications for what a
workstation is, but then, I don't even agree with all of them... :-)



* A workstation must be network ready.
  This doesn't mean that you can't have one workstation sitting
  somewhere, but you must be able to add additional workstations easily.

* A workstation must be multitasking
  But then, I have heard of DOS machines connected via Novell as being
  called "Novell workstations".  I have also heard HP call the
  HP9000/300's running Rocky Mountain Basic as "basic workstations".
  (they were network ready)

* A workstation must run UNIX
  This is a favorite of the unix bigots.  Of course, one of the
  founders of the workstation field was apollo and they didn't run unix.

* A workstation must have a GUI
  Well, the "Novell workstations" and the "basic workstations"
  mentioned above fail this test, but it does seem to be very common
  for workstations to have a GUI.

* Workstations must be multi-user
  Well, this is kind of part of being on a network.  A lot of what
  goes into Novell is adding multi-user functions to DOS.  This also
  makes OS/2 and WNT questionable.  (That may be why I have heard it
  used as a requirement recently :-)

* Workstations must run a Real Operating System
  A great definition which allows you to get into the "What is a Real
  Operating System flame war".

* "Workstation" is a buzz word
  It became a popular buzz word somewhere after "User Friendly" and
  before "Object Oriented"...   :->


I think the more explicitly you try to make the definition, the less
people you will find that agree with that definition.


-wayne