mappings-01

[Vint Cerf]

john,

The mappings document outlines what I think is part of a spectrum of
mappings/modifications of user input that may eventually find its way
into a DNS lookup action. Technically, one could refer in the lookup
section of protocol the phrase "implementors MAY apply mappings-01
prior to lookup." it seems to me this is permissive without being
dispositive. I agree that definitions could reference the document
as being a part of the series dealing with lookup pre-processing.

Paul, Peter,

would such a reference do damage to your effort to treat the
matter with delicacy?

vint


On Jul 5, 2009, at 3:02 AM, John C Klensin wrote:

> Vint,
>
> While I agree with all of this, I do need some advice about what
> should be said in the Definitions and Protocol documents about
> mapping.   The former can say something equivalent to "that
> document is part of the series" without any normative language,
> but Protocol has to either invoke mapping as a step (presumably
> with MAY or SHOULD in the sentence, recreating the issue that
> Paul and Pete decided to avoid) or perhaps just mention it in
> the discussion of preprocessing steps.   I will assume that the
> latter is closer to what we want, but would appreciate
> clarification from you and the WG as soon as possible.
>
>     john
>
>
> --On Saturday, July 04, 2009 11:27 -0400 Vint Cerf
> <vint at google.com> wrote:
>
>> Taking advantage of this exchange to underscore for the
>> working group
>> what I consider to be a remarkably perceptive observation:
>>
>> quoting from the authors:
>>
>>>> In the case of mapping user input, we could not give a good
>>>> reason why this is needs to be required. It is clearly a good
>>>> idea because it will prevent user surprise, and we say  so.
>> ...
>