|
Sadly the outcome of the process back
then was rather protective. To be concrete two years ago I
requested access to the then defunct UDC data to practice data
mining and never got it because it requested approvals....
On 15/03/2014 3:42 PM, Mike Milinkovich wrote:
On 15/03/2014 3:30 PM, Pascal
Rapicault wrote:
What
needs to be considered when putting together such a solution
(even if it is reduced to just collect error logs) is the amount
of data this represents, and consequently the bandwidth, the
computation and of course the storage required. In addition to
these technical aspects, there is of course the obvious
questions of who is going to take care of the problems coming
in, who is going to have access to the data, and whether the
data will be accessible to everybody who request it since it may
contain product specific information and give hints about the
competition.
Wayne et al dealt with many of these types of issues when
we had the UDC (universal data collector) running. I could imagine
that many of the same policies and approaches could help here.
|
