- From: Dave Kristol <dmk@allegra.att.com>
- Date: Thu, 1 Aug 96 10:01:54 EDT
- To: sommerfeld%apollo.hp.com@hplb.hpl.hp.com
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com, snowhare@netimages.com
Bill Sommerfeld <sommerfeld@apollo.hp.com> writes: > The "security considerations" section of the draft does not include > any text regarding privacy concerns. That's hardly true. Section 7 is entitled PRIVACY; its first subsection, 7.1 is entitled "User Agent Control". > > Here's some suggested text: > > PRIVACY CONCERNS: > > The protocol described in this draft can be used to keep track of the > browsing habits of a user without the user's knowledge or permission. > Many people consider this to be an unethical invasion of privacy. > > Any HTTP client implementing this protocol MUST provide at least three > options for the user: > 1) disable cookies entirely. > 2) ask the user before setting a cookie. > 3) set cookies without asking the user. These are enumerated, in slightly different terms, in 7.1. > > The default "out of the box" behavior of the client MUST NOT be #3. > > Any HTTP client should provide a way for the user to know which > cookies are associated with a given page. The thrust of the privacy considerations throughout the document is to give the user control. But I have to agree with Benjamin Franz (<snowhare@netimages.com>) that user agent behavior is outside the IETF's purview. IETF can specify protocols, and they/we do that to ensure interoperation. I believe user agent behavior can be suggested (we do), and then we have to rely on public (and, dare I say it, marketplace) pressure to shine a spotlight on implementations that pose a risk to users' privacy. Dave Kristol
Received on Thursday, 1 August 1996 07:06:34 UTC