W3C home > Mailing lists > Public > spec-prod@w3.org > April to June 2014

Re: Informative components of w3c specifications

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Mon, 30 Jun 2014 15:04:38 -0700
Message-ID: <CAAWBYDA-JQvdW0XEDHuXXFjsXCnQrQBH9Qb3S2j0H_GiPRmDdA@mail.gmail.com>
To: "Michael[tm] Smith" <mike@w3.org>
Cc: Shane McCarron <shane@aptest.com>, "spec-prod@w3.org Prod" <spec-prod@w3.org>, Mark Sadecki <mark@w3.org>
On Mon, Jun 30, 2014 at 11:24 AM, Michael[tm] Smith <mike@w3.org> wrote:
> Shane McCarron <shane@aptest.com>, 2014-06-29 17:38 -0500:
>
>> My question is "Are notes *always* informative?"  If not, "Are notes by
>> default informative, and should be annotated if they are normative?"  Or,
>> alternately, "should notes be required to have a class of 'informative' in
>> order to be marked as such?"
>
> For the editors and specs that I work with, the following hold true:
>
> The only sentences in any spec that are normative are sentences that state
> normative requirements. And even if a particular sentence doesn't contain
> MUSTs or other RFC 2119 language, it can still be normative if it's a part
> (subrequirement) of a larger algorithm that's normatively defining a set of
> steps or parts of a larger requirement.
>
> But a note is never normative. A note never states requirements and is
> never a subrequirement of some larger requirement. Instead a note is just
> an informational clarification or aside of some kind. From that it follows
> that a note must never contain normative RFC 2119 language. So, if
> something that's marked as a note contains RFC 2119 language, then that's a
> spec bug that needs to be fixed (that is, it should not have been marked as
> a note to begin with).
>
> So there should never be any need to mark a note either way -- that is,
> e.g., neither note@class=informative nor note@class=normative are needed.

Yes. Notes with normative requirements are spec bugs, and are called
out as such by many people when noticed.

~TJ
Received on Monday, 30 June 2014 22:05:26 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:55:19 UTC