Thanks for all the suggestion. The sad news is that this costs time and
energy, and I'm not sure there is actually people looking at these
stats, thus my suggestion to possibly remove the panel altogether.
is there someone willing to implement these more complex rules?
cheers.
Il 14/11/19 15:47, M
Question regarding download count: if I merely upgrade a plugin, does the
download to upgrade the plugin counts as an overall download?
We could have a refined popularity count if we'd find a way to come up with
an active users count (which shouldn't be that difficult to implement, info
could be p
What Paolo said seems a suggestion to modify the existing query. Beside
popularity (downloads), the query should filter out the plugins that
provide functionality already in core (by version). This means having a
stronger control of the plugins and features to decide what is in and what
is out for
It appears to me that popularity should be what is currently popular. If
you are only keeping track of the total downloads and not when they are
downloaded, then I am not sure that category has that much meaning. Adding
in the ratings probably adds a little, but as I have looked at some of the
star
Can't remember exactly why that formula, but it was carefully discussed at
the time (7 years ago):
https://github.com/qgis/QGIS-Django/blob/master/qgis-app/plugins/models.py#L113
On Thu, Nov 14, 2019 at 2:58 PM C Hamilton wrote:
> What kind of stats are kept for the plugins? Popular plugins
What kind of stats are kept for the plugins? Popular plugins could be those
with the most downloads during the past year, but I don't know if you keep
track of when the downloads occurred.
Calvin
On Wed, Nov 13, 2019 at 11:50 PM Paolo Cavallini
wrote:
> Hi all,
> I noticed that the Popular plug
Hi all,
I noticed that the Popular plugins sections on our plugin web app is
somewhat misleading, listing mostly plugin with a high number of
downloads and good ratings in the past, not necessarily still very
useful, often replaced by better alternatives in core.
I'd suggest either finding a better
