Alberto Mardegan wrote:
Cornelius Hald wrote:
Also using the current download count for the "Popular" section of
Downloads seems a bit unfair then. Shouldn't it be at least
downloads/num_of_releases? Or better
downloads_from_extras/num_of_releases?
I don't think it's possible to know whether it is a new download, or an
update.
No, I don´t think either. But if we get numbers for each release, we
could see something like:
Ver Downl.
0.2 1000
0.3 1200
0.4 1300
0.5 700
0.6 200
This would at least show, that with version 0.5 people have somehow lost
interest. Maybe I´ve made some radical UI changes that most people don´t
like. Or maybe there is another (better) tool for the same task, etc...
But maybe limiting the download count for the packages downloaded from
Extras only makes more sense.
I would think so. I´m often releasing to extras-devel with really
minimal changes - often primary to test something for myself. Other
people who have extras-devel activated, will receive this updates as
well but often it´s just too unimportant to count those as "official"
downloads.
But I like the idea that a package with more frequent releases gets more
visibility than others. :-)
That could still be possible. We could have "Most frequently updated
packages" or something. We could also use the number of releases for the
popularity rating. But I think the numbers first should be separated.
For example:
* AppA 20 releases with 20.000 total downloads. First release 1 year ago.
* AppB 2 releases with 5.000 total downloads. First release 1 month ago.
Which application is more popular? Which has more downloads? This shows,
that is is especially unfair to young applications.
If we would have downloads/releases the situation would look differently:
* AppA: 1.000 average downloads
* AppB: 2.500 average downloads
Also, if we want to use the "number of releases" for popularity or
something similar, we could set them in relation to "time on market".
Having something like "average number of releases in the last month".
I think it would make the statistics between old and new applications
more comparable and it could encourage community members to invest more
in their young project if they can see that they basically have the same
user base as, for example, omweather.
Cheers!
Conny
_______________________________________________
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers
