- From: Rohit Khare <khare@pest.w3.org>
- Date: Fri, 16 Feb 96 14:31:59 -0500
- To: www-demographics@w3.org
In our e-commerce work, we outlined a demographics facility that might be
integral to the shopping experience. Here were some of my comments:
> ...
> 2) Automated Form Fill-In
>
> As an aid to e-Commerce and payment-support, we could
> provide automatic hooks in the browser to fill-in name,
> shipping address, etc. We managed to pull this out of
> discussions of payment protocols by citing the approach
> noted by Dan Connolly and myself: designate a set of
> field-names that can be defaulted by configuring a client.
>
> Thus, any FORM with a text field named (say)
> http://pep.w3.org/AutoFill/FName could have a default
> value supplied automatically by the user's browser. This
> can be implemented completely by browsers. We should
> suggest this to the HTML ERB.
>
> 3) Demographic Profiling
>
> XXX, in particular, identifies a need to know a few things
> about the customer almost before shopping even starts,
> in order to filter products not available, calculate
> local currency prices, etc. They are driven by actual
> customer concerns; but the details of what should go into
> this information is unresolved.
>
> Protocol Name: http://pep.w3.org/Profile
>
> Standard Parameters: {age int}
> {residence country-code state-code}
> {delivery country-code state-code}
> etc.
>
> We might consider adding parameters for use when
> requesting a profile:
> {policy
> http://pep.w3.org/Profile/WillNotRedistribute.html}
>
> Rohit is extremely skeptical about creating this protocol;
> it has the "build it and they will come" risk: just using
> it will make it the default demographic-disclosure
> protocol for many other uses. It should be tied to
> developments in the Demographics area. In particular
> advertisers have expressed interest in standardizing
> things like the policy parameter alluded to above.
>
> It seems logical that parameter included here should only
> be those that can't (easily) be used to identify an
> individual. Individual identification is better done by
> the default form filling stuff in 2) above.
> ...
Reactions?
Rohit Khare
Received on Friday, 16 February 1996 14:30:40 UTC