[ih] Fwd: When did "32" bits for IP register as "not enough"?

[Brian E Carpenter]

On 2019-02-19 16:35, Bill Ricker wrote:
> Vint says
>> not so richard. By 1983, DOD had officially endorsed OSI.
> 
> Indeed.
> DoD formal endorsement of OSI was Mike Padlipsky's cross to bear.
> His portion of our Department at MITRE was on the DODIIS procurement.
> (DODIIS was the compartmented secure fork of the MILNET fork of the DARPA
> Internet for "the community".)
> "They" were very unhappy with his pointed commentary on ISO OSI not being
> ready for real use, let alone being "bad art".
> Luckily his pointedest papers were approved for public release in 1982.
> 
> I erred in replying off-list over the weekend -- remedied below.

And an updated version of my off-list reply is inserted below...

> -- 
> Bill Ricker
> The Literary Estate of M.A.P.
> <bill.n1vux at gmail.com>
> 
> 
> ---------- Forwarded message ---------
> From: Bill Ricker <bill.n1vux at gmail.com>
> Date: Sat, Feb 16, 2019 at 3:50 PM
> Subject: Re: [ih] When did "32" bits for IP register as "not enough"?
> To: Brian E Carpenter <brian.e.carpenter at gmail.com>
> 
> 
> 
> fwiw, it was by then two years after the rant** which marked my own
>> epiphany. [...]
>> ** B.E. Carpenter, Is OSI Too Late?, RARE Networkshop 1989, Computer
>> Networks and ISDN Systems, 17 (1989) 284-286, DOI
>> 10.1016/0169-7552(89)90040-8
>>
> 
> Now that the words "rant" and "epiphany" have been mentioned, I should
> mention that such epiphanies were foretold by a ranting prophet, not
> respected in his own bureaucracy (a metaphoric country, as is traditional
> since Cassandra):
> 
> M.A.Padlipsky's *Tea-Bag Papers* of 1982.

They still make fun reads. Two points:

1) Most of his arguments equate OSI with X.25/X.75 at layer 3. Of course,
those of us actually trying to adopt OSI for research networking mainly
rejected that approach and were only interested in CLNP. ISO/IEC 8473:1986
was the first formal CLNP standard, but it was surely in discussion by 1982.
I implemented a toy version of ISO DIS 8473 in early 1984, where the tricky
bit was fragmentation and reassembly, since our in-house network at CERN
had a criminally small MTU size. Unfortunately I have no archives from
that period.

A little bit of virtual dumpster diving reveals that the ballot version of
DIS 8473 was also numbered SC6/N1534, but I haven't found exactly when it
was posted for the committee ballot. However, being in Geneva we had no
great trouble getting hold of early drafts.

2) The analysis in RFC873 "THE ILLUSION OF VENDOR SUPPORT" was spot on.
Many of us were in fact fooled into believing that the products would
come at a reasonable cost. 

   Brian
> 
> (The original private-circulation edition of these technical papers came
> with a coverpage with photocopied Salada tea-bag-tag-lines, hence the
> sobriquet.)
> 
>    - RFC871: <https://tools.ietf.org/html/rfc871> A Perspective on the
>    ARPANet Reference Model
>    - RFC872: <https://tools.ietf.org/html/rfc872> TCP-on-a-LAN
>    - RFC873: <https://tools.ietf.org/html/rfc873> Illusion of vendor support
>    - RFC874: <https://tools.ietf.org/html/rfc874> Critique of X.25
>    - RFC875: <https://tools.ietf.org/html/rfc875> Gateways, architectures,
>    and heffalumps
> 
> which are summarized by his pastiche of a Tuna advertisement of the day
> which made it into one of the Quotes files:
> 
> On Networking Architecture
> "Do you want protocols that look nice or protocols that work nice?''
> Mike Padlipsky, internet architect -- Bitmover Quote file
> <https://web.archive.org/web/20050308174523/http://www.bitmover.com/lm/quotes.html>
> 
> These papers were the "existing draft" that Mike showed to the
> Prentice-Hall field editor through whom I'd ordered another colleague's
> book as alternate text for a class -- Unix for Secretaries -- who'd asked
> me if I'd like to write a technical book, and I said no, but my colleague
> Mike has the beginnings of one, and thus became the core of  *The Elements
> of Networking Style (& Other Essays & Animadversions of the Art of
> Intercomputer Networking*), 1985. (So yes, I'm implicated marginally, which
> is why I'm here.) Alas the Literary Estate of A.A.Milne refused to grant
> mechanical copyright to include the actual Heffalump quotation in the book.
> (Possibly they feeling burned by "*The Tao of Poo*"?)
> 
> IIRC the GOSIPing ISORMites in DODIIS management - his nominal customer at
> MITRE while continuing to do WG like things - were not most pleased, and
> his later commentary on ISO OSI RM vis-a-vis TCP/IP had to be
> "handkerchiefed" to hide the "constructive snottiness" (and/or privately
> circulated).