Re: [radext] [dhcwg] draft-ietf-dhc-dhcpv6-radius-opt-10

[Ted Lemon <[e-mail]>]

On Apr 4, 2013, at 4:57 PM, Alan DeKok <[e-mail]> wrote:
>  Positive statements are usually clearer to understand.

Yes, and when I read your statement, I understood it to mean the exact opposite of what you intended.   Easier to understand doesn't help if the statement is ambiguous.

>  The intention is for the option to carry RADIUS attributes.  Making
> that a requirement rather than a suggestion is a good idea.

This isn't what the text says.   It says that the option must only carry a subset of RADIUS attributes; those listed in a special registry.

If you don't like the double negative, here's a precise way to say it that doesn't contain a double negative:

	This option MUST NOT carry any RADIUS attribute unless it is listed in the
	IANA Registry of 'RADIUS attributes permitted in DHCPv6 RADIUS option'.

But what the text that means what you said in the second quote above would read like this:

	This option MUST NOT carry any RADIUS attribute unless it is listed in the
	IANA Registry Radius Types registry in the section titled 'Radius Attribute Types'.