Unfortunately, I've only just now had the time to review the threads about metrics that have been floating on this list and fedora-docs-list. I'll now take this opportunity to weigh in on what has been covered so far and add my own thoughts on the matter. == Tracking in Start Page == First, I'd like to point out that this isn't entirely limited to Firefox, as the page we currently use is displayed in some of the other base browsers, too. === Using a Tiny Image === I hate this idea. If we have nothing to hide, don't hide it. We have to remain accountable, no matter what the cost. If we're so worried about letting users know what we're doing, then we shouldn't be doing it. === Using an Optional Link === Perhaps my favorite way to handle this is to present an obvious, well-highlighted link near the top of the page that gives users an opportunity to opt-in by clicking it to get to a survey. Advantages: * Transparency: The user will know what is going on and will be able to decline before anything happens. * Full Survey/More Information: Since this link can take the user to a survey, we can collect more information than a small image could ever give us. * Accuracy: Using this method, we wouldn't have to worry about duplicates and other issues that plague tracking by IP. We could also combine this with other techniques in an attempt to determine what percentage of users are following the link. This could also be implemented with a JavaScript dialog, but we'd have to make sure that the dialog only appears once. == Tracking in firstboot == Ultimately, this is the solution I'd like to see. Presenting a short, opt-in survey during firstboot is probably the best way to handle these types of metrics. == Tracking from yum == We should be collecting traffic data from all of our web-based resources, including the package repositories and websites. This can be used to complement the more comprehensive data collection of a survey. == What We're Already Missing == While we do have logs from the assorted websites, and probably from other resources, we're not using that information. We should be collecting information from those logs for assorted metrics. They may not be extremely accurate ways of determining numbers of users, but it could give us valuable insight into what our users are doing and where we'd do well to allocate more resources. In my opinion, that information would be more valuable than any information about total numbers of users. == Privacy == Whatever we do, we need to pay careful attention to our techniques and policies. We already have a draft privacy policy that only needs review by Legal in order to be made official. Anyone taking part in this discussion should read over it. It lays the basic groundwork for us to follow. http://fedoraproject.org/wiki/Legal/PrivacyPolicy One slip with regard to user privacy could do irreparable damage to the Project. Think before you act. -- Patrick "The N-Man" Barnes nman64@xxxxxxxxx http://www.n-man.com/ LinkedIn: http://www.linkedin.com/in/nman64 Have I been helpful? Rate my assistance! http://rate.affero.net/nman64/ --
Attachment:
pgpfufg0EPCej.pgp
Description: PGP signature
_______________________________________________ fedora-advisory-board mailing list fedora-advisory-board@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board
_______________________________________________ fedora-advisory-board-readonly mailing list fedora-advisory-board-readonly@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board-readonly
