>> I wonder if we might consider some kind of "user survey" option
I really like the idea of a survey - at least a lot more than that of a usage data collector. And I think it is actually more suited for the questions asked here than an ongoing data collection. I'd give it some time - say two weeks to a month after first launch of the installation (or workspace, not sure at what level this might make the most sense without being too obtrusive). Then pop up a dialog _once_, asking to take part in the survey and optionally including a snapshot of the relevant settings. The idea is that the user's typical setup including all configuration options that are interesting for us in the scope of this discussion will have mostly stabilized by then and a single snapshot would be enough to get comparable value as an ongoing data collection. > A user survey is better than nothing, but we are definitely going to get far > fewer survey responses than opt-ins for automated usage data collection, > which makes the answers less valuable. I think if you ask nicely for participation _once_, you'll get at least as many participants for the survey as you'd get for the data collection, if not more because it's less big-brother-ish. And if it optionally includes a snapshot of the preference data, it'd be of at least comparable value (my bet would still be it's of more value). _______________________________________________ cross-project-issues-dev mailing list cross-project-issues-dev@eclipse.org https://dev.eclipse.org/mailman/listinfo/cross-project-issues-dev