[postgis-devel] Website analytics

[Greg Troxel]

Sandro Santilli via postgis-devel <postgis-devel at lists.osgeo.org>
writes:

> Recent changes in the PostGIS Manual HTML output (from HTML to XHTML)
> have triggered a problem with the Google Analytics javascript code,
> which for some reason (which I didn't try to understand) seem
> incompatible with the XHTML format.
>
> Given that not all web pages of the PostGIS project are using Google
> Analytics (ie: the single-page HTML manual, the cheatsheets, the
> main page) I'm not sure if anyone in the project management is relying
> on that, thus this mail.

Really, none of them should be be using Google Analytics, and osgeo
should have a policy against it.

> Personally, I'd stop sending our user's IP addresses to Google, using
> some self-hosted solution for stats, which may be provided by OSGeo
> at some point (see https://trac.osgeo.org/osgeo/ticket/2830).

I would rather *not have analytics* than abuse our users.

> What do others think about this ?

> At this very moment the Google Analytics script was removed from the only
> place it was used in (chunked html) as a workaround for the problem
> reported above and in ticket https://trac.osgeo.org/postgis/ticket/5559
>
> In this mail I'd like to ask if it's ok with you to completely stop
> using Google Analytics and you think we still need those analytics
> whether you have a way to take those stats out of Google for reuse
> with Matomo (or whatever other solution we'll end up deploying in OSGeo)

Just stop, please, and if we get self-hosted analytics, that's fine, but
I don't think they should be linked.

I do not recall seeing anything on the list that was useful from this.

I would think that if we need analytics, it would be to understand if a
website design is working, and that's a measurement of now or future,
and the past isn't particularly important.