Re: [abnf-discuss] [art] [Technical Errata Reported] RFC7601 (5435)

[Dave Crocker <[e-mail]>]

On 7/23/2018 7:49 AM, John R. Levine wrote:
> More to the point, the IETF has used ABNF in its current form for over 
> 20 years,


41 years.

The changes made in the document since ABNF's inception in 1977 -- as 
part of RFC 733 -- are helpful but not fundamental.

And I'll take this opportunity to cite an error in credit I've been 
making over the years...

Within the set of RFC 733 authors, I was the newbie with no formal 
training and wound up as editor of the document.  I spent an afternoon 
in a technical bookstore, getting a little background on parsers.  There 
were lots of specifications being written in those days that did 
variants of BNF; ours was merely one more.

I remember being concerned that we hit a good balance between features 
(complexity) and utility (usability).  And I remember worrying quite a 
bit about the formatting of the ABNF, since that affects comprehension. 
(I was, at the time, in a nascent Human-Computer Interaction graduate 
program.)

What I didn't remember until reviewing the Introduction to RFC 733, 
recently, is that the core credit for creating ABNF goes to Ken Harrenstein.

d/

-- 
Dave Crocker
Brandenburg InternetWorking
bbiw.net