Re: Statistics on non-free usage
On Mon, Jan 12, 2004 at 11:26:59AM +1100, Craig Sanders wrote:
> > Oh c'mon. Just because I made a mistake doesn't mean that I'm
> > dishonest. After all, you are the one that said your package has "0
> > entries in popcon", then tried to change it to "used" once I had
> > shown you to be incorrect (knowing full well that "used" is a different
> > category in popcon), then said that I gave the "impression" that nobody
> > was installing your packages, even when my own figures showed some of
> > them were installed on 3% of the machines. This is being honest how.
> well, frankly, your use of percentages was a little dishonest to say the least,
> as it let you round-off many packages to '0'.
I had no idea what the results would be before running the program, and
did not alter it to adjust those results later. You have no basis to
know whether I was being dishonest or not.
Moreover, right from the start I said that they were percentages. I
also published the code used to generate this for public inspection and
analysis, with a complete change history. It includes the specific
popcon file I used (so that the results can be validated even after the
next popcon update, and so that they can be compared to popcon before
the next update).
> you should have given actual numbers, and let people calculate the percentages
The actual numbers are already in popcon. Implementing "cat" would
probably not have helped much here.
> partly your fault for not being more clear or providing column headings that
> indicated percentages). i guess other readers might have missed that too.
Quite possible; though using this as a basis for assuming I am being
dishonest is quite a leap indeed.
> in any case, you have shown yourself to be dishonest on numerous occasions in
> this long and tortuous argument. you have no care for truth, or honour - you
> will utter any lie in the name of your cause.
Which is interesting statement, since you have yet to describe even one