[ih] Why did TCP win? [Re: Internet-history Digest, Vol 63, Issue 3

[Dave Crocker]

> I think the political/management lesson of all this is: without 
> someone to say No and make it stick, design efforts always devolve to 
> the lowest common denominator. All the dumb ideas get folded in.
> OSI was a committee without a BDFL. That's why they tried to satisfy 
> everyone by having different "profiles."
>
> TCP didn't have a single BDFL but it had a group of wise people, most 
> of whom are still with us, who could say No. In my own WG (RFC 1697) 
> we had Marshall Rose, whom I remember saying, "What part of 'no' don't 
> you understand? The 'n' or the 'ol'? "


hmmm.  I'm pretty sure my comments in this space have not cited an 
ultimate authority, for breaking deadlocks, as required. Certainly there 
are examples of its being quite helpful, in the Internet's history, but 
I don't think it was a primary reason for success.  (Early central 
funding, from ARPA and later NSFNet was certainly important, but that's 
different from the technical decision processes.)

And, as I do make a point of citing, both communities had bright, 
well-intentioned folk forming their communities. And both had 
administrative hierarchies.


An essential difference that surfaced back when I was writing about this 
was that one community required unanimity, and thus tended towards 
solutions that were the union of everybody's wishlists.  The other 
wanted something working yesterday and therefore tended to 'settle' for 
the intersection of the many lists.

The difference between leaner, quicker specs that got plenty of field 
experience, versus bloated, slow specifications with essentially no 
serious field experience, is what swamped the OSI boat.

d/

--
Dave Crocker

Brandenburg InternetWorking
bbiw.net
bluesky: @dcrocker.bsky.social
mast: @dcrocker at mastodon.social