Collecting information

[Jeff Spaleta]

On 7/5/05, Rahul Sundaram <sundaram at redhat.com> wrote:
> * A tool to collect package stats
> * A tool to collect hardware information
> * A form to fill out for users to send in more information to the
> project with appropriate "incentives" to do it

I'd rather know why we..as fedora..are collecting any of this
information before we collect it.
I want to see concrete plans on how this information is going to be
used before we start collecting.  Who is going to datamine that
package stats information and for what constructive purpose?  Who is
going to datamine that hardware information and for what constructive
purpose?  Who is going to datamine geographic and other userdata and
for what constructive purpose?

I am very much not inclined to building a treasure trove of data...
just to have it around because people are somewhat interested in
knowing things about the userbase.  If there are no concrete plans to
use the data as it comes in, I think having a mountain of data its
only going to prove to be a burden.

Gnome has a stated goal cleverly named "10x10". Getting an estimate of
the number of gnome users is clearly needed to know if they are on
track to reach that specific goal.  Fedora on the other hand, has no
grandiose user adoption targets. I don't see a driving reason to build
this mountain of information up.

I don't see any credible discussion that people agree that the popcon
concept is something Fedora is willing to adopt  or any other need for
use package stats data. I don't see any credible discussion about user
adoption targets that could possible need userdata for feedback. I
don't see any credible discussion about doing something worthwhile
with hardware data if we had it.

All i see are some vague expressions of desire to leisurely mine data
to get some trending statistics. I don't think that is a good enough
reason to start building up the data. I think there needs to be
commitment to actually use and process the data as it rolls in.
Sitting on a mountain of userdata without doing anything with it as it
comes in, is just going to lead to a grotesque mess as people with
specific agendas start mining historic data to try to strong arm
discussion.

-jef