[ih] Fwd: nomenclature

[Jack Haverty]

I agree - I also think that the term "host" is older than the Arpanet.
It's important to remember that the Arpanet was the first "packet
network" - not the first computer network.  

IIRC, in the 60s the "host" was the computer at the other end of your
comm link, and it "hosted" the program that you were using from your
terminal.  You would dial up the host to use the program which it
hosted.  The program was a "guest" - it could be kicked out at any
moment to cool its heels on the drum while a VIP (Very Important
Program) was hosted.

I think "host" was used in the context of the various terminal-computer
networks that IBM, CDC, et al supported in the late 60s, but I have no
idea where it originated.  Old Sperry or IBM docs would be a good
source.

/Jack


On Wed, 2009-12-23 at 13:23 -0600, Larry Sheldon wrote:
> On 12/23/2009 11:47 AM, Dave CROCKER wrote:
> 
> > I remember that the term 'host' was one of the bits of mind-share that
> > the Arpanet created.. But I don't recall hearing why that particular
> > term was chosen.
> >
> > Do any of you remember why?
> 
> I'd have to dig out some old USE documentation (or maybe some UNIVAC or 
> Sperry docs if I can find them) but I think we used the term "host" 
> before we had "networks" as the term is now used.
> 
> In the IBM world, what was on the other end of a telephone line from an 
> RJE station?
> 
> It seems to me that one end of a word channel cable was the 
> "controller", the other the "host".