Evolution

[Mario Vukelic]

On Tue, 2008-07-29 at 11:53 -0600, Karl Larsen wrote:
> Here is the problem: Your data says 509363 people Installed evolution.
> This is not accurate. The number comes because every Ubuntu loaded has evolution. 

Yes. Nobody ever suggested  to use the "Installed" number as a
reference.

>     Now 141817 it seems loaded Thunderbird. They had to apt-get the 
> software. But only 45,359 people are using it. Old is just a subtraction.

Well, not exactly, unless I misunderstood what you meant.
509363 - 80793 = 428570, while <old> shows 404082. I am not sure where
this discrepancy comes from, though. Ideas welcome.

>     For some reason the data shows 80,793 people are now using 
> evolution. This is a great difference with both my 100 email samples of 
> this list.

Correct, but it should be noted that "some reason" actually means "the
number of people who (1) enabled popcon data sending in menu System >
Administration > Software Sources > tab Statistics, (2) used Evolution
at least once in the last 30 days, and (3) don't have /usr mounted with
noatime." (2) and (3) are  according to the popcon FAQ whose location I
gave in another mail yesterday.

Unless, of course, you want to argue that the popcon client application
is sending invalid data and/or the popcon server miscalculates the
numbers. If so, then I'm afraid you'll have to show proof for this
assumption.

>     Now my question remains, how were these numbers computed. 

For the second time, the computation method is explained in the
README.gz and FAQ.gz files in /usr/share/doc/popularity-contest

> In particular how was the number of d/l Thunderbird NOT USED counted?

Above, you said with regards to Evolution, that "Old is just a
subtraction". We have seen that this is not exactly the case, but I
guess it is close enough for now.
What I don't understand is why the same method is unclear for you
regarding Thunderbird?