Re: Statistics on non-free usage
On Fri, Jan 09, 2004 at 05:35:42PM +0100, Sven Luther wrote:
> On Fri, Jan 09, 2004 at 10:17:50AM -0600, John Goerzen wrote:
> Taken from the data you quoted :
>
> > Package Vote Old Rcnt Unknown
> > ocaml-book-en 0 0 0 19
> ocaml-book-en non-free/doc 0 0 0 1 1
> > ocaml-book-fr 0 0 0 10
> ocaml-book-fr non-free/doc 0 0 0 0 0
> > ocaml-doc 0 0 0 54
> ocaml-doc non-free/doc 0 0 0 3 3
> > unicorn 1 0 3 0
> unicorn non-free/net 0 0 0 0 0
> > unicorn-source 0 0 0 7
> unicorn-source non-free/net 0 0 0 0 0
>
> So, 3 out of 5 were indeed marked as not being used in the data you
> presented as argument. This means that it seems that the data you
> provided were false and not can not be used for this kind of discussion.
No it does not. You claimed that the packages did not show up as being
installed in popcon. That is not true.
Only one package above actually has binaries; the usage statistics are
not collected for docs, but as you can plainly see, the installation
statistics are.
> > The results do show that these packages are installed. In fact, from my
> > own results, I showed that 3% of people had ocaml-doc installed. These
> > figures support that.
>
> Well, from your own mail.
>
> > I'm not sure where you got that idea that popcon showed that nobody had
> > installed these packages.
>
> I hope it is clear now.
No, it's not. I show that 3% installed ocaml-doc. You have even quoted
that figure. My data was based on percentage and decimal places
truncated; therefore, for the packages that had less than 1% use, it
will rightly show up as 0%; not that the difference was significant, and
you did have packages with greater than 1% use.
-- John
Reply to: