W3C home > Mailing lists > Public > public-tracking@w3.org > July 2012

Re: ISSUE-4 and clarity regarding browser defaults

From: Matthias Schunter <mts-std@schunter.org>
Date: Sat, 07 Jul 2012 16:02:36 +0200
Message-ID: <4FF8417C.3080207@schunter.org>
To: Tamir Israel <tisrael@cippic.ca>
CC: rob@blaeu.com, "Roy T. Fielding" <fielding@gbiv.com>, Rigo Wenning <rigo@w3.org>, "publ >> \"public-tracking@w3.org\"" <public-tracking@w3.org>
Hi Tamir,

the goals are
 a) the DNT MUST reflect user preference
 b) We want to allow innovation for user agents and do not be too
prescriptive. This is implemented by MAYs: 

As a consequence, there are many suggested ways how this can be
    The browser MAY ask the user at first use.
    The browser MAY ask the user at install-time.
    The browser MAY ask the user once a day.
    The user MAY choose to install a "maximise your privacy" suite
(e.g., the TOR anonimizer) that includes DNT as a tool.
    ... ... .... [add any other suggestion how to reliably collect user


On 21/06/2012 06:50, Tamir Israel wrote:
> Yes, but the sense I'm getting is that the current spec sacrifices a
> great deal in certainty in ways that can get very simply taken care of
> by requiring UAs to get a user election.
> That wouldn't really be overly prescriptive. UAs, plugins, etc., will
> have a great deal of leeway in how to go about doing this, when, etc.
> On 6/21/2012 12:45 AM, Rob van Eijk wrote:
>> Because an overly prescriptive objective will not leave room for
>> other ways to accomplish the same.
>> In order to have a meaningful global DNT standard, it is important to
>> have technical building blocks built in.
>> On 20-6-2012 20:34, Tamir Israel wrote:
>>> If  the WG's objective is to ensure user preferences are expressed,
>>> why don't we simply make this a MUST?
>>> On 6/20/2012 10:15 PM, Rob van Eijk wrote:
>>>> Tnx Roy,
>>>> Normative: "... users MAY be given a choice during installation,
>>>> update or first startup."
>>>> Non-normative:
>>>> There are use cases, where a choice given on first startup would be
>>>> the preferred choice mechanism.
>>>> For example,
>>>> - a device can have multiple user profiles per installation;
>>>> - in cases where browsers are not installed by the user.
>>>> Rob
>>>> On 20-6-2012 18:22, Roy T. Fielding wrote:
>>>>> On Jun 20, 2012, at 5:31 PM, Rob van Eijk wrote:
>>>>>> "... users may be given a choice during installation, update or
>>>>>> first startup."
>>>>> Not during installation or update -- only during first use.
>>>>> The preference should be stored in the user's browser config (along
>>>>> with their other personal preferences); a device might want
>>>>> to have multiple such user profiles per installation.
>>>>> In normal enterprise environments and the vast majority of purchased
>>>>> computers, browsers are not installed by the user.  They are
>>>>> installed
>>>>> by sysadmins, vendors, etc.  Likewise, updates should not modify
>>>>> user preferences.
>>>>> ....Roy
Received on Saturday, 7 July 2012 14:03:03 UTC

This archive was generated by hypermail 2.3.1 : Friday, 3 November 2017 21:44:52 UTC