GrADS download statistics at sf.net

Arlindo da Silva arlindo.dasilva at GMAIL.COM
Mon Mar 30 10:05:42 EDT 2009


On Mon, Mar 30, 2009 at 8:29 AM, Paolo Ciraci <info at paolociraci.it> wrote:
> Hi All,
>
> A little issue about usage statistics based on browser/user-agent:
> someone may download a package for a certain SO using a browser (or
> other tool) on a different SO. For instance I always download the
> freebsd or linux builds from windows so the windows stats are somehow
> "higher"... to the detriment of freebsd or linux.
>

Sourceforge keeps a counter of how many times a given tarball was
downloaded, regardless of the OS/browser used for download; this is
what I used for the pie charts attached to the e-mail. The other pie
chart on http://marketshare.hitslink.com/ indeed used the originating
browser/OS for their statistics --- that pie chart should be used only
as a rough broad brush estimate for OS usage for the community at
large (not only the grads community).

Having said that, you make a very important point regarding these
statistics --- they should not be taken too seriously.  What really
constitute grads usage? Is it the number of plots you make? Is the
number of distinct users running grads? For example, many times a
single copy of grads is installed on a Linux server and dozens of
people use it daily, while people tend to download the windows version
for each desktop. It took me about 15 minutes to create those pie
charts on excel reading the numbers off sf.net. Any serious attempt to
collect these statistics would require designing a survey and making
sure we have a representative sample answering the survey, etc. I'd
rather spend the time completing the documentation...

  Cheers!

    Arlindo

--
Arlindo da Silva
dasilva at alum.mit.edu



More information about the gradsusr mailing list