Re: [abnf-discuss] Deviant form of ABNF in RFC 989/1049/1341/1521/2045/2397

[Paul Overell <[e-mail]>]

Hi,

I remember raising this issue when we were working on RFC2234, =, := and 
::= were discussed.

= from RFC822, RFC733 also used by Wirth (1977) and the BSI (now ISO) 
meta language.

::= from the Algol 60 report, AFAIK the original BNF

:= in use in some other RFCs, perhaps this was misremembered Algol 60, 
confusing ::= with its assignment operator :=

We settled on =


Regards

On 23/04/2023 23:06, Dave Crocker wrote:
> On 4/23/2023 2:42 PM, Carsten Bormann wrote:
>> Does anyone remember the history that made this alternative form 
>> appear, teeter on for a while, and then pretty much go away?
>> Was there any other difference apart from “:=“?
>
> I don't know anything about this variant.  For those who need a wider 
> context:  Through the 1970s and 1980s, there were lots of variants of 
> BNF.  We did one for RFC733, continued into RFC822, and it happened to 
> become popular with other folk.  So, RFC822 got incorporated into all 
> sorts of non-email specifications, to get it's form of ABNF.
>
> The later splitting out of the ABNF into a separate RFC was, really, 
> just to make specific citation easier, rather than for technical 
> reasons.  But of course, since we had the hood up,...
>
> d/
>
-- 
Paul Overell