Le 05/05/2014 08:47, Jim a écrit :
On 2014-05-03 08:34, David Bruant wrote:
You need to understand the constraints and then apply creative
design within these constraints. For example, a row of attractive
feedback buttons that work with JS disabled, and when JS is enabled
zoom in on these when leaving the page with a request to select one
to leave a feedback selection.
This is an atrocious user experience. When someone wants to leave the
page, they want to... leave the page, not give you feedback.
They may answer negatively just out of this frustration and there
would be no way to distinguish between this frustration and another
frustration. Also, a bunch of buttons is a very small amount of
information. Maybe not even a relevant one.
Good point, don't do it. Spying on people is any even more 'atrocious
user experience' once they realize. Your current design, requiring JS,
is not privacy friendly and it is not consistent with fighting for our
privacy.
so disable JS altogether? Don't visit pages you don't trust?
I'm not sure what's the problem at hand. Mozilla collects information in
a way that is transparent enough that it triggered this thread. Mozilla
uses Google Analytics and explained that this choice fits with a set of
goals regarding user privacy (Mozilla is even in a contractual
relationship with Google Analytics to enforce this).
Mozilla informs its users about its practices on its websites. Gareth
Cull came to this thread to explain these practices. There are clear
opt-opts (like the GA opt-out addon).
What more do you want?
What do you want and can be realistically applied? Don't only consider
your end. Try to understand what's being done now, what's at stake and
provide an alternative where other goals aren't diminished.
Among other things, human beings are both not fully self-aware (far
away from it, myself included) and irrational. The type of feedback
you get from observing how people do behave has nothing to do with
what people would say they do (only a fraction would give feedback, so
there is a "self-selection" bias and this fraction would say what it
thinks, not what it does).
The questions asked on the form also create a bias since they
necessarily direct the answers.
The very position and size of the button would create a bias based on
people screen size for instance.
I may be exagerating, but I feel a feedback button would mean lots of
time processing the answers where a statistical bias (well, several)
making any result questionable.
Listen to yourself. Observing how people behave while they are not
aware!
I admit I haven't checked /privacy, but I can only imagine users are
informed there about this practice.
I'll repeat what I said elsewhere [1]:
"Note that a website is already capable to "track" every single click
(to other pages within the same domain) and know your navigation
patterns within the website. I'll go further by saying that this is
possible even without cookies (look for "web keys" or "capability
URLs". These weren't aimed at tracking, but could be use for that
purpose).
I don't see how recording other clicks is that awful."
Every serious website has some form of HTTP logging (most often for
post-mortem security purposes). That's definitely some form of
"observing how people behave while they are not aware". If you're
uncomfortable with this idea, I'd recommend not going to any website
just to be sure :-/
More realistically, I'd recommend to inform yourself on common
practices of saving informations about users and wonder how much
you're okay with them.
Client side apps are an opportunity to avoid some tracking. Arguing
that just because http server driven apps can log actions that this
should be accepted in privacy friendly client side apps is not
constructive for the privacy cause.
Are you even interested in understanding the issues?
I'm interested in these issues. I'm more interested in how to get to the
subtle balance of getting what you want done (understanding user
behavior) while preserving privacy.
Now, tell us why Facebook and Twitter were associated prominently with
children pleading for their privacy and control on the web?
I don't know. What is this children story? Never heard of it, but I'm
curious.
Did Mozilla receive consideration for these placements? Has Mozilla's
pragmatism sold out the privacy cause yet again?
Where does this "yet again" come from?
How is privacy compromised this time? The best argument I've read was
"despite having signed a contract, we can't trust Google after the Prism
revelations" and I find it weak. YMMV.
These children do not understand what your pragmatism has done, but I
do, and you are using these children, it disgusts me.
I'm not using any children. Nor is Mozilla to the best of my knowledge.
Are you plain trolling right now? This discussion stopped making sense
as soon as you brought "children" up :-/
David
_______________________________________________
governance mailing list
governance@lists.mozilla.org
https://lists.mozilla.org/listinfo/governance
