Re: [eclipse.org-architecture-council] Download stats reliability and SL

[Denis Roy <denis.roy@xxxxxxxxxxxxxxxxxxxxxx>]

On 2020-03-17 4:48 p.m., Mickael Istria wrote:

On Tue, Mar 17, 2020 at 8:12 PM Denis Roy <denis.roy@xxxxxxxxxxxxxxxxxxxxxx> wrote:

The stats are reliable since I adressed the bug.

Great thanks.

And do you know how bots/builds may affect those stats? Are there some kind of filters in place?

There are no filters in place. We're providing raw numbers in the hopes the projects can analyze trends. If there are clear requirements on what to filter out, I'm open; it tends to me a moving target.

The problem is injection points -- we rely on the Find a Mirror script, and p2.statsURI for injecting a stats entry. If a project links to a mirror directly, or download.eclipse.org directly, or worse -- if a user follows a direct link, then no stats are generated for those downloads.

For Eclipse Platform/p2 artifacts, it seems OK as we have the download.stats relatively well defined and easy to define. The cases of direct links are, I imagine, negligible if the goal is to get an approximation of the number of users.

I would think extracting trends data from the numbers is where the value is.  But I'm a server nerd, not a marketer :)

Denis

_______________________________________________
eclipse.org-architecture-council mailing list
eclipse.org-architecture-council@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/eclipse.org-architecture-council

--

Denis Roy

Director, IT Services | Eclipse Foundation, Inc.

Eclipse Foundation: The Platform for Open Innovation and Collaboration

Twitter: @droy_eclipse