[ih] IP/TCP to TCP/IP

David P. Reed dpreed at reed.com
Wed Oct 24 17:35:11 PDT 2001

Well, when we split TCP into TCP and IP, we thought of them as distinct.

I've always thought of TCP/IP as an odd moniker.  If it really meant "TCP 
over IP" it would not include IP alone, or UDP or RTP.  And if it meant 
"TCP or IP" (like he/she) then it wouldn't be parallel - you'd expect 
TCP/UDP/RTP/ICMP to be the name.

And if you wanted an all-encompassing name, you'd use "Internet Protocols".

My bet is some marketing dude for operating systems at some point started 
including TCP and IP on checklists of features supported.  It would be 
natural at that point to use an acronym he/she saw on a lot of whiteboards 
without spending the time to understand semantic distinctions.

At 05:21 PM 10/24/2001 -0500, John Kristoff wrote:
>Alan, I hope you don't mind me following up to the list.
>I noticed it today when I was reviewing an older paper:
>Development of the Domain Name System
>Paul Mockapetris, Kevin J. Dunlap, ACM SIGCOMM 1988
>When I saw it, I was curious, because I know I've seen it before and
>assumed that was how it was originally written.  Perhaps its not as
>prevelant as I thought.  I did a quick search and although TCP/IP seems
>to be at least as common if not more so even early on, it does turn up
>RFC 896, Congestion Control in IP/TCP Networks
>J. Nagle, 1984
>NSL Technical Note TN-2
>Using screend to Implement IP/TCP Security Policies
>J. Mogul, 1991
>And most interestingly:
>I wonder if it was a 'purist' thing not unlike the proper pronouniation
>of router.
>"Alan J. Maitland" wrote:
> > I suppose the first time I even heard the term TCP/IP was around 1980, then
> > represented as an up and coming replacement for ARPANet.  I don't believe
> > that I ever heard it expressed as IP/TCP.  By the time I received any
> > formal training on networking, it was 1984/1985 and TCP/IP was expressed in
> > that way then.  Which docs are you referring to?
> >
> > Perhaps it was just for emphasis of the Transport Control Protocol part in
> > anticipation that other protocols might fall on the other side of the '/'.
> >
> > I believe that I had read that some of the early developers of the DNS
> > service bemoaned the decision to go x.y.com vs com.y.x as the standard for
> > name hierarchy.  Of course, my company uses the hierarchy to attempt to
> > logically order its portals, so I suppose that I am grateful for the
> > current system - I don't think that Net.Clothing.Mens look as good as
> > Mens.Clothing.Net ;-)
> >
> > On the other hand, sometimes things just happen.  The decision to name the
> > company Hewlett-Packard verses Packard-Hewlett was decided on the toss of a
> > quarter by its founders!
> >
> > Best,
> >
> > Alan Maitland
> > The Commerce Company - Connect Your Dot(sm)
> > http://WWW.Commerco.Com/
> >
> > At 12:59 PM 10/24/01 -0500, you wrote:
> > >Everyone now knows the suite of Internet protocols as TCP/IP, but a few
> > >of the early Internet documents I've seen have them documented as
> > >IP/TCP.  Did they purposely get reversed?  If so, why?  Was it just a
> > >matter of personal preference and TCP/IP won out?
> > >
> > >John

More information about the internet-history mailing list