[dow2tf] replies to questions submitted to Counsel

[Barbara Roseman <roseman@xxxxxxxxx>]

> The following questions have been submitted to Counsel on behalf of the 
> taskforce:
>
> If a respondent supplies proprietary information to the taskforce 
> regarding their uses of whois, can that information be presented anonymously?

When the DNSO did a Whois survey back in 2001, it carried the following
terms/disclaimer:

        "Your comments will be published for public review, not associated
with your identity in item 21. Your e-mail address (item 21a) will be used
to send a copy of your responses back to you, but will not be provided to
third parties without your permission. Your name and organization (item 21b)
may be published as part of a list of responding parties."
<http://web.archive.org/web/20010812044726/http://www.icann.org/dnso/whois-s
urvey-en-10jun01.htm>

I think that the sort of handling as indicated above would take care of
question #1.


> And, if we are presented with information that indicates someone is using 
> Whois in a manner that is currently prohibited, or violates the AUP of a 
> registrar, will we make use of that information for any other purpose 
> aside from the survey?

On the second question, I'd be inclined to say yes, but subject to the same
provisions as above (i.e. respondents' comments will not be provided to
_third parties_ without permission).  I can see why promising to not go
after abusers might encourage honest reporting of Whois abuses, but I think
in practice the idea would be too complicated and subject to
misinterpretation and dispute.  (Note: of course if a task force's members
really want "anonymous" information, they could just accept anonymous
input.)

> Given that this is a public process, what are the implications of changing 
> the process to allow for anonymity after having already received some 
> responses?

It depends on what you mean by "allow for anonymity".  If survey
responses are handled as indicated in my prior response, I can't think of any
important problems.

> Are there effective ways of presenting the final data that doesn't 
> correlate responses with a given respondent, but allows for confidence 
> that the data is represented accurately?

Yes, publish responses without identifying information, along with a
list of respondents' names/organizations?