[e2e] TCP design descisions

Jon Crowcroft Jon.Crowcroft at cl.cam.ac.uk
Fri Jul 25 23:46:46 PDT 2003

the point of the final resolution in rashomon is that you cannot trust anyone AT ALL to
even interpret events in the past in a vaguely similar way, not
that there are slightly different viewpoints based on different emphasis.

i'd be more interested in more hard to find design debates like 
i) who was involved in the ISO
OSI Teransport Protocol (e.g. class 4 was quite nice -we did an implementation which
outperformed tcp:)

ii) where did XTP end up

iii) wasnt there a multi-link X25? did anyone ever build it and if so how well did it work?

iv) what other nameservices apart from dns and grapevine were built ? is there equivelent
documentation to the grapevine experiences papres out of PARC for them?

v) what other routing systems were built (e..g the french research net i cannot recal that
had load sensitive routing that thrashed - what was that and is there a decent paper on
what went wrong - also the Netcomm dynamic routing for 2Mbps x.25 nets? was that ever
written up - ? has anyone got an open source pnni ? etc etc)

the hidden fortress is a good laugh if you like your Kurosawa lighter...

In missive < at>, "David P. Reed" typed:

 >>IEN's, personal archives, and personal memories are what we have.   Many 
 >>design decisions were argued out in groups in person and online and 
 >>transcribed after the fact into IEN's.    I agree with Craig that the best 
 >>way is to focus on specific questions.   There are many folks who were not 
 >>involved at the time who have opinions about what happened and why.   Those 
 >>of us who were there are a bit more reliable, but memories vary, and 
 >>there's a Rashomon effect depending on what mattered to each participant most.
 >>At 01:08 PM 7/25/2003 -0400, Craig Partridge wrote:
 >>>In message < at pop.interlap.com.ar>, 
 >>>Fernando Go
 >>>nt writes:
 >>> >I'm looking for some paper about TCP design decisions.
 >>>My guess is that what exists is in the IEN series (Dave Mills used to
 >>>keep them on-line.. hoping he still does).
 >>>For later stuff, such as the decisions behind the mods to TCP in the late
 >>>1980s and early 1990s, some is captured in the End-to-End Interest list
 >>>discussions and the rest is probably in the heads of the people on the
 >>>list.  So if you've got specific questions, fire away.



More information about the end2end-interest mailing list