- From: Rigo Wenning <rigo@w3.org>
- Date: Tue, 16 Jul 2013 18:24:54 +0200
- To: David Singer <singer@apple.com>
- Cc: Alan Chapell <achapell@chapellassociates.com>, public-tracking@w3.org, Sid Stamm <sid@mozilla.com>, Justin Brookman <jbrookman@cdt.org>
David, the issue is that you don't want the browser or one of the search bars or tools to phone home. This was pretty clear in HTML 4. In HTML5 the distinction between browser-search-bar phoning home and widget ABC phoning home and a web page tracking starts to blur. But what they all do is sending browser history information or uniqueID. All other stuff is mainly harmless. What Issue is this attached to and if there is none, we should create one. --Rigo On Tuesday 16 July 2013 18:06:33 David Singer wrote: > >> "A user agent MUST NOT share information related to the network > >> interaction with any party other than the user without consent." > > > > > > > > This would mean that the user agent can not load a thing without > > consent. Because the user agent must share IP address and other > > things with a lot of parties other than the user to obtain the > > content and just that (not even tracking). > > > > > > > > I know what you mean, the wording still doesn't do the trick. The > > problem here is not the consensus, but the wording… > > whoops, maybe you are right. let's keep thinking. > > but, let's say I visit a site that needs a plug-in. somehow I trigger > you into loading a page that loads that plug-in. haven't you just > visited that other site? > > OK, let's imagine a browser that has a table, "if a page needs X, then > load Y form site Z". Now I visit Q, which needs X loaded (a font, a > plug-in, whatever). The browser detects this and asks "do you want > to load Y from Z, it's needed for this page?". The user says > yes; the browser then visits Z to load Y, but there is no reason for > it to mention Q (or for any data to flow between or about Q and Y) is > there?
Received on Tuesday, 16 July 2013 16:25:29 UTC