finger in your face?

John (john@syd.dit.csiro.au)
Thu, 2 May 91 12:39:37 +1000

I wanted to make some comments about Steve's idea. I still may if I get time.
There's something I have to deal with first, though. Philippe, citing reasons
why he doesn't like the idea, wrote:

Lastly, finger follows the Internet protocols philosophy of having
all data be human-readable.

Let's all read that again, slowly and carefully, so there can be no mistake.

[F]inger follows the Internet protocols philosophy of having all
data be human-readable.

Now, there have been little pieces of bait tossed out on the list before.
Tiny flecks here and there. You know who you are -- and what they were.
I resisted them. Manfully, I ignored the provocation. I thought, and
I still think, that the faces list wasn't the place for an ARM versus ISORM
holy war.

But I just cannot allow this one to pass unchallenged.

Let's decompose Philippe's statement. Its main component is:

There exists an ``Internet protocols philosophy'' which
suggests/wants/mandates/insists (he is unclear) that
``all data be human-readable''.

This is out-and-out silliness. The statement is ludicrous on its very face
and betrays woeful and complete ignorance of the ARM. Justify that, you say?
(Thus revealing your own woeful and complete ignorance of the ARM, but
never mind...) Well, last I heard, there was this thing called the File
Transfer Protocol. "ftp". Heard of that one? Sure you have. Indeed,
you've probably used it. Can it transfer binary files? Sure it can. No
problem at all. Are binary files human-readable? Not the last time _I_
tried to read one.

The logic here is airtight, with only one route of escape for Philippe: he
might wish to suggest that FTP does not conform to the alleged ``Internet
protocols philosophy''. I won't dignify that with a refutation.

I should include a small aside here about the meaning of the term
``human-readable''. I am taking it to mean ``directly human-readable'',
that is, put it in front of someone and they can extract its semantic
content without assistance from a machine. It seems quite clear to
me that Philippe is using the term in this sense; after all, an X-Face:
header, while not directly human-readable, is perfectly _indirectly_
human-readable: run it through the right software and you get a face.

Having disposed of Philippe's laughable claim, let's imagine he might
have meant something else. He might have meant (even though this is
quite a leap of the imagination from what he actually _said_) that everything
moved by _SMTP_ is meant to be (directly) human-readable. He was talking
about finger, but then claimed:

In mail, X-Face data is acceptable because it is an "X-" (or
unoffical) field...

thus clearly implying that stuff moved by SMTP is supposed (in his twisted
view of the ARM) to be directly human-readable. I needn't expatiate the
stupidity of this at any length; merely see, for one of a myraid of examples,
RFC1154.

I hear a rumour that, in France, it is illegal to send encrypted mail.
I don't know if this is true. If it is, perhaps it is fascism such as this
that has led Philippe into confusion. If that is the case, Philippe, allow
me to suggest that you ought not to confuse the jackbooted tactics of your
government with the Arpanet Reference Model.

And to all those ISORMites out there -- look, I don't give a good damn _what_
you do in private, as long as you wash your hands afterwards. But DAMN
WELL KEEP YOUR LUDICROUS, STUPID, UNIMPLEMENTABLE, GARBAGE PROTOCOLS

O F F O F T H I S L I S T!!!!!!!

I trust I have made myself as clear as an azure sky of deepest summer.

John.