[OSM-talk] Simple improvement(s) to openstreetmap.org

[Jean-Marc Liotier]

On 01/09/2013 08:49 PM, Jeff Meyer wrote:
> +1 to JM's request for putting together a plan to gather data. Moral
> and ethical questions aside, we don't know the value to OSM unless we
> test.
Actually, moral and ethical problems must be defused preemptively -
especially within a project that upholds exemplary values.

Data retention policy sounds awfully corporate, but such document is the
foundation for not unthinkingly drifting sooner and later into
behaviours that we would not have condoned in the first place. Is anyone
aware of such document somewhere around OSM ?

Of particular importance is the anonymization of user data - an
essential step if we wish to provide useful datasets to a group larger
than a trusted core.

> I believe larger sites do things like offer a different home page to
> different users - e.g. one with these links, and then one without - in
> an attempt to compare two concurrent data sets, but I'm not sure how
> feasible that approach would be.
That is called behavioural targeting. It is rather more advanced than
what is usually needed to begin with and users don't like being aware
that a specially skewed content has been served to them. That said, as
an intermediary step, we might want to differentiate more strongly
between logged users, non-logged users with a specific cookie and new
arrivals.

> The timeframe for the test is another data point - did the Comms WG
> have any thoughts on how to long to run this test? 3 months? 6 months?
We don't even know if they are aware of this thread... I'm cc:ing a few
of them in case they aren't.