Comments on RFC 580: Machine readable protocols
A short objection to requiring protocol documents to be written as NLS files.
Clements, R., "Comments on RFC 580: Machine readable protocols", RFC 582, DOI 10.17487/RFC0582, November 1973, <https://www.rfc-editor.org/info/rfc582>.
@misc{rfc582,
series = {Request for Comments},
number = 582,
howpublished = {RFC 582},
publisher = {RFC Editor},
doi = {10.17487/RFC0582},
url = {https://www.rfc-editor.org/info/rfc582},
author = {R. Clements},
title = {{Comments on RFC 580: Machine readable protocols}},
pagetotal = 1,
year = 1973,
month = nov,
} RFC text
Network Working Group R. Clements
Request for Comments: 582 BBN-TENEX
NIC: 19962 5 November 1973
Comments on RFC 580 -
Machine Readable Protocols
I fully support the requirement for machine-readable protocol
documents. In my situation, the line-printer is a much more reliable
device than the copying machine.
However, I object to the phrase "preferably as nls files" in RFC 580.
My objection is based on the lack of conversion mechanisms INTO NLS,
not to the retrieval process or NLS itself.
Most sites have their own text editors and RUNOFF's (or their
equivalents). Most large protocol documents are prepared at least
partially by secretarial help. Those persons should be able to
prepare the documents in the home machine (or wherever) in languages
with which they are familiar. There should be a general program
(preferably clever, but at least generally available and predictable)
for converting nicely formatted text to NLS files.
Perhaps the program which receives mail for the journal will do the
trick; if so it needs further documentation beyond the mail-oriented
RFC 543, and its existence and usage need to be publicised.
RECEIVED AT NIC NOVEMBER 14, 1973.
[ This RFC was put into machine readable form for entry ]
[ into the online RFC archives by Lorrie Shiota 1/02 ]
AI ELI5
On this page
§1Support With One Objection
R. Clements of BBN-TENEX writes in reply to RFC 580. He fully supports the requirement for machine-readable protocol documents, noting that in his situation the line-printer is a much more reliable device than the copying machine.
His objection is to a single phrase in RFC 580: "preferably as nls files". He is clear that the problem is the lack of conversion mechanisms into NLS, not the retrieval process or NLS itself.
He likes the idea of sharing documents as computer files. He just does not like being told to use one special program that few people can feed.
§2Why Sites Cannot Easily Use NLS
Most sites have their own text editors and RUNOFF programs (or equivalents). Most large protocol documents are prepared at least partly by secretarial help, and those people should be able to work on the home machine in languages they already know.
People write in the tools they know. Making them learn another tool just for sharing is a big ask.
§3The Requested Converter
The author wants a general program for converting nicely formatted text to NLS files. It should preferably be clever, but at the least generally available and predictable.
He suggests that the program which receives mail for the journal might do the job. If so, it needs documentation beyond the mail-oriented RFC 543, and its existence and usage need to be publicised.
He asks for a helper machine that turns ordinary writing into the special format, and for someone to tell everyone it exists.