- From: Poul-Henning Kamp <phk@phk.freebsd.dk>
- Date: Sat, 15 Oct 2016 08:58:30 +0000
- To: Kari Hurtta <hurtta-ietf@elmme-mailer.org>
- cc: HTTP working group mailing list <ietf-http-wg@w3.org>, Martin Thomson <martin.thomson@gmail.com>
--------
In message <20161015054336.F1EDC16A9A@welho-filter4.welho.com>, Kari Hurtta wri
tes:
>If I guess correctly Martin Thomson's format gives
>
>| common-structure = 1* ( (identifier dictionary) / anonymous-dictionary )
>|
>| dictionary = * ( identifier value )
>|
>| anonymous-dictionary = 1* ( identifier non-empty-value )
>|
>| non-empty-value = <value; empty value not allowed>
>
>
>and
>
>| h1_common_structure = h1_element * ("," h1_element)
>|
>| h1_element = ( identifier * (";" identifier ["=" h1_value]) ) /
>| ( 1* ( identifier "=" h1_value) )
>
>
>This looks serializable, but is it parsable?
First, I think Martins format is wrong, with or without Common
Structure, and the "CE-param:" header is a better and more general idea.
Second, I think what you propose is parsable, but Getty's 3rd rule
("The only thing worse than generalizing from one example is
generalizing from no examples at all.") and 1st rule ("Do not add
new functionality unless an implementor cannot complete a real
application without it.") tells me to not add it.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Received on Saturday, 15 October 2016 08:58:57 UTC