[ih] Internet addressing history section

[Dave Taht]

jnc at mercury.lcs.mit.edu (Noel Chiappa) writes:

>     > From: Paul Ruizendaal
>
>     > There was no broadcast-based discovery of other gateways on the same
>     > local net(s).
>
> There _weren't_ any LANs in the very earliest Internet (although packet radio
> networks were close); that why the earliest IP address had only 8 bits of
> 'network number', to support the small number of WANs.
>
>     > the PUP Gateway Information Protocol, which in turn seems to have been
>     > influenced by the routing protocols used inside IMP's.
>
> The very earliest ARPANET routing protocol, perhaps - both it, and the Xerox
> routing protocols, were Destination Vector. The ARPANET transitioned to a
> Link State design fairly early. I'd have to check the dates for PUP, and that
> transition; my my sense is that the latter happened first.
>
>     > RIP for tcp/ip itself was only codified in an RFC in 1988; not sure why
>     > it took 5 years to do so.
>
> A number of reasons, but part of it was a feeling on some people's part (e.g.
> me) that DV algorithms had been shown by the ARPANet work to have issues. The
> BBN report by McQuillan et al that introduced the idea of LS (might have been
> "ARPANet Routing Algorithm Improvments", No. 3803 - I can't find my copy)
> went into great detail about the problems they were seeing with DV
> prototocols.
>
> In retrospect, the concern over DV was probably excessive; the original
> ARPANET routing was load-sensitive, so the inputs to the path selection were
> orders of magnitude more dynamic, which probably forced a lot of the DV issues
> out into the light.

These days I periodically work on the babeld DV routing protocol. Among
other things it now has an RTT based metric

(https://tools.ietf.org/html/draft-jonglez-babel-rtt-extension-01)

It's generally been my hope, that with the addition of fair queuing
(which didn't arise until nagle in 85 - and the more implementable forms
of SFQ (1990) and DRR (1995), that the early experiences with DV being
problematic were... early bufferbloat.

>
> 	Noel
> _______
> internet-history mailing list
> internet-history at postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-owner at postel.org for assistance.