[ih] A paper

[Bill Ricker]

On Sun, Jul 18, 2021 at 8:17 PM Darius Kazemi via Internet-history <
internet-history at elists.isoc.org> wrote:

> Most of you probably don't know me. I'm an engineer by trade and amateur
> historian for kicks. I spend a lot of time acting as a kind of translator
> of technology history for people who work as engineers today, because I
> don't want my colleagues to repeat the mistakes of the past and would
> rather they learn from its successes. I spend a lot of time reading old
> RFCs and other documents from the 60s-80s because I'd rather learn from
> your hard earned wisdom than shoot myself in the foot when I'm working on
> decentralized protocols and applications.


Sensible!

> (I'm currently devouring
> Padlipsky's 1985 "Elements of Networking Style", a political document if
> there ever was one.)
>

Mike might have quibbled that debunking nonsense is political only if the
purveyors of nonsense make it so.
But indeed, that was the case - he (we) knew at the time it was politically
dangerous to discuss the emperor's new fashion sense publicly. He had the
good sense and/or good fortune to be doing so from a place that DoD had
chartered to speak Truth to Power on other weightier matters, which
provided some amount of protection (until they came for the
tobacco-addicts).

I'm glad you're "devouring" and I hope you're enjoying MAP's "Elements".
(I am personally to blame for introducing Mike to the acquiring field
editor; but it's my prior office-mate's lousy timing on his Hawaii office
rotation that initiated the coincidences.)

2) I often describe a protocol as "an agreement between parties about the
> way things are expected to be done". A protocol that no one agrees to use
> is dead in the water, as you all know. I don't believe etymology is
> deterministic but it is interesting to note that the word "protocol"
> literally derives from politics, specifically that of diplomatic protocol!
>

Nicely put!  I feel Mike would have approved. (speaking /ex cathedra/)

> Regardless, it is difficult for me to envision an agreement between parties
> as anything but political. I'll note that I don't mean political in the
> sense of campaigns and government spending. I mean it in the broad sense of
> humans organizing to do things. I think some of the misunderstanding here
> is that we are working off different definitions of "political".
>

Indeed.
In much of engineering usage, "political" is the antonym of "technical" and
generally refers to circuses like Sales and/or Marketing attempting to
blame Engineering for building what was requested not what was wanted, or
the Peter-Principle'd PHBs having turf wars for their own petty purposes.
Using the same word for real-world electoral/diplomatic politics is indeed
confusing to the mind trained over decades to sny away from anything
"political" in a "technical" discussion.

> 3) It seems the paper has been misread ...
>  The absolute meat of the paper is in its last two sentences, which


In which case the most important feedback to the authors is
DO NOT BURY THE LEDE.
If we knew it would end with those sentences, we might have finished
reading it.
Seems like a conclusion that would be worth including in the abstract.

It really feels like some of you saw the word "politics", saw that this was
> written by people who aren't engineers, and assumed they were coming from a
> position antagonistic to yours.


In the current milieu with refrain "Tech is not Neutral", that is an often
correct assumption.

(OTOH I will agree with those making political points e.g. against lack of
diversity in VC-funded development (OxyFlo clip meters) or against Machine
Learning, wherein they are /inter alia/ pointing out that it likely will
learn the human biases in the training set. I've preferred AI with
explanatory mechanisms since the prior AI Scare. Research like "Weapons of
Math Destruction: How Big Data Increases Inequality" is a welcome ally to
my long-held position. It's not a political position to say a rigged demo
that only appears fit for purpose shouldn't be put in production, which i
would claim fits under MAP's Technico-Aesthetic Criticism, but there are
political positions that reach the same conclusion by observing downstream
outcomes.)

> If anything, this paper is highly
> antagonistic to the over-politicization of technology!


That is a pleasant surprise.
Pray, authors, surprise me sooner!

> That you personally
> cannot read the article and parse it isn't anyone's fault,


I probably could if suitably motivated ...
but the abstract and a fast scroll didn't provide that.

> for the same
> reason that you can't expect a nontechnical person to read RFC 114 and make
> heads or tails of it.
>

A number of technical folks have complained of reading Padlipsky, too. :-D

(I for one do like an author who challenges the reader.
But I would, wouldn't I.)

> Anyway it's been pretty frustrating to see everyone talking at cross
> purposes


One might suggest that talking at cross purposes was the original purpose
of the Internet,
and Cat pictures only a fortuitous later discovery. (Citation: SF-LOVERS,
the entire run.)

Bill Ricker
The Literary & Spiritous Estate of Michael A Padlipsky
https://n1vux.github.io/articles/MAP/