[rescue] "Workstation" [was Re: What do people think of a Sun Fire V210^W^W Ultra 45]

[der Mouse]

[Jerry Kemp]
> I don't think that the PCI bus is inherently bad, but I do agree that
> the PCI bus brought a lot of bad stuff with it.

That's basically my take on it, yes.  What I know of it makes me think
it's actually a reasonably good bus design.

> My definition of a workstation is:

> * a powerful system (at least for its time of release), that is not a
>   server

> * the main/core CPU is a RISC processor

> * runs a real Unix, not a Unix clone

Hmm.  What's a "server", then?  As a real example, Sun sold
"SPARCstation N" and "SPARCserver N" for some value of N (1? 5? 20? I
forget) where the only hardware differences were a factory-installed
graphics head and the printing on the front of the case.

As for RISC processors, well, that means there aren't any peecee
workstations, which isn't in very good line with how the word is
commonly used.  (I'm also not sure it's a useful basis to draw a
distinction on.)  It also means things like the VAXstations and the
Sun-3 line weren't/aren't workstations.

Real Unix...there aren't very many of those.  No Linux, no Haiku, no
Smalltalk, no Plan 9, no NeXT, no VMS.  This is sounding less and less
like anyone else's definition of "workstation"...and less and less like
a useful definition, actually.

> IMHO,  m$ was the first to muddy the classic workstation definition
> when they released a product that they called windows nt workstation.

I dunno.  DEC and Sun (to name just two I'm relatively familiar with)
were selling things marketed as workstations that didn't fit your
definition long before NT was more than a gleam in anyone's eye.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse at rodents-montreal.org
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B