> >>> Yes but how many of those downloads turned into someone
> >> actually using
> >>> the software? Probably 5%. How many of those downloads are
> >> from people
> >>> who already use the software? I have downloaded it myself
> >> probably 30
> >>> times this year.
> >>
> >> Definitely concur with that, as I know I've downloaded several
> >> hundred copies between reinstalling computers, upgrading
> postgresql
> >> versions, etc ...
> >
> > But do you download it through the website? AFAIK, direct downloads
> > from the mirrors are *not* included in the count above -
> it's just the
> > clickthrough counts on the website..
>
> How is that applicable? The point is that the 775k # is not
> a "distinct"
> #, and doesn't take into account the multiple downloads that
> happen over the year ... I just happen to get mine directly
> from an ftp server, but I could just have easily gotten my
> multiple downloads by doing a "fetch"
> from the web site itself ...
I was thinking people get the first one from the website, but then if
they fetch more they get it from the ftp server directly.
That of course also means you don't count *at all* in the 775k - the
same for everybody who get their postgresql shipped with their
faviourite linux/bsd/whatever distro.
//Magnus