[ih] The web as wind and whirlwind? (was Re: History from 1960s to 2025)

[Bill Nowicki]

I agree with Dave Crocker, albeit my perspective might be skewed since I worked in commercial networking since 1985.
For example, Sun Microsystems had TCP/IP included in each product, and any kind of restriction on TCP/IP protocol stacks (as opposed to, say, Arpanet) would be virtually impossible to enforce. Many customers were in a gray area, since they might have commercial customers as well as a government contract or two. This resulted in many islands of TCP/IP on LANs, which created a huge demand for commercial TCP/IP WAN routers to connect the islands.
Sun itself had some government customers but as far as I know no government contracts. We connected ourselves to the Arpanet to better support the customers who did, and policing the traffic was not practical. My feeling was that the majority of traffic we exchanged was in fact Usenet and email lists. Besides, as soon as possible we switched to the Bay Area Regional Network (BARRNET) which became self-supporting fairly quickly. The Clinton/Gore/Magaziner commercialization was really just admitting reality that was already starting to happen, instead of spending resources trying to squish it. The legacy telephone carriers tried to get into the game with technologies like ATM, but the path of least resistance was clearly doing a native TCP/IP backbone, if the US Government liked it or not.
Bill    On Thursday, December 18, 2025 at 10:30:26 PM PST, Brian E Carpenter via Internet-history <internet-history at elists.isoc.org> wrote:  
 
 On 19-Dec-25 17:44, Dave Crocker via Internet-history wrote:
> On 12/18/2025 6:52 AM, Andrew Sullivan via Internet-history wrote:
>> 1. The introduction of URLs/URIs made the identity of a site (the host
>> part of an http URL) really important and encouraged the
>> identification with trademarks.
> 
> My impression was that, since the issue is with domain names' ability to
> have real-world semantic, the trademark concern surfaces with /any/ use
> of domain names.  The web certainly exacerbated concerns, but it didn't
> create them.
> 
> 
> On 12/18/2025 12:16 PM, Brian E Carpenter via Internet-history wrote:
>> Here's a counterfactual question: what would have happened if the
>> whole Clinton/Gore/Magaziner commercialization project had never
>> taken place?
> 
> Commercial use of the Internet was already a serious issue by the late
> 1980s.  Before the Web was invented.
> 
> NSFNet had funding but was still ramping up.  So, again, the NSFNet,
> etc. effort pushed growth, and it pushed some organizational and
> operational choices, but I do not believe it created the inevitability
> of a commercial Internet.(*)
> 
> So, no, I think ISDN was not the likely alternative.  More likely was a
> version of the Internet, albeit with less operational and/or
> administrative flexibility.

Yes, it's important to recall that when TimBL invented HTTP, he
could perfectly well have decided to implement it over OSI (we had
enough OSI running at CERN for that to have been technically plausible)
but he chose TCP/IP precisely because of the Internet** (including the
Cornell-CERN link that meant we were directly peering with NSFnet).
TCP/IP had already won before the web and long before Magaziner.

** I haven't yet read his new book, but he said that explicitly in
his 1999 book "Weaving the Web".

    Brian


> 
> d/
> 
> 
> (*) In the late 1980s, I was managing development efforts for TCP/IP and
> OSI stacks on several platforms.  We went to a number of customers --
> mostly commercial organizations -- to find out their requirements for
> moving from TCP/IP to OSI.  Without exception they said they had no
> interest in that capability.  And, in fact, they were eager for
> transition tools from OSI to TCP/IP. Again, this was before the Web was
> invented.
> 
-- 
Internet-history mailing list
Internet-history at elists.isoc.org
https://elists.isoc.org/mailman/listinfo/internet-history
-
Unsubscribe: https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history