search

w3c / dnt

Archive of DNT deliverables
https://www.w3.org/TR/tracking-dnt/
Other
12 stars 19 forks source link

Purposes Extension: Generalize interaction between Tracking Exceptions and Browser Features #92

Open jasonanovak opened 5 years ago

jasonanovak commented 5 years ago

In the July 23, 2018 Editor's Draft of Tracking Preference Expression (DNT)- Purposes Extension Addendum, the Introduction says:

Browsers that wished to actively help to protect users by restricting the capability of servers to track them, such as Firefox's Tracking Protection feature, or Safari's Intelligent Tracking Prevention [ITP] MAY use this extra information, i.e. that users have expressed their right to object, or have not given their explcit consent, to restrict the lifting of these protection procedures to particular domains. For example, if a first-party site specifies a DNT:1 Tracking Exception, this could be taken as a command to purge all cookies. The opposite DNT:0 resulting from a site-specific Tracking Exception could then be used to allow "partitioned" cookies to be sent to the relevent third-party domains (and made available via document.cookie to the associated browsing contexts).

Firefox Tracking Protection and Safari ITP are both browser specific implementations of anti-tracking technology that have different underlying models of identifying tracking domains; in addition, Brave has a third implementation and ethos. Moreover, Safari ITP does more than block cookies but also, with ITP 2.0, truncates referer information sent to domains classified as having the ability to track. As a result, it seems premature to suggest specific interactions between specific browser features and TPE. Instead, perhaps it would be best to replace this section with a statement like “How TPE interacts with browser specific anti-tracking features is left to browser vendors.”

rvaneijk commented 5 years ago

As this paragraph is included in the introduction, I think the comment by @jasonanovak can be handled as an editorial change.

Furthermore, I suggest s/MAY/can/, since the introduction should not contain normative text.

michael-oneill commented 5 years ago

This paragraph is redundant anyway , so I removed it. I added Jason's suggested sentence to the paragraph about an indicator in the chrome.