Re: Bits from the Release Team - Kicking off Wheezy
* Lucas Nussbaum <firstname.lastname@example.org> [110429 11:17]:
> On 29/04/11 at 11:00 +0200, Bernhard R. Link wrote:
> > * Lucas Nussbaum <email@example.com> [110429 10:51]:
> > > I think that you misunderstood my data. It's the number of **distinct**
> > > (different) **uploaders**. If you uploaded just one package during the
> > > freeze period, then you are not in the 20%.
> > You did not give numbers of uploaders in the freeze time, but by
> > month. So I still count in the other months. (And with one specific
> > package maintained less, I might have an avarage of less than a package
> > uploaded per freeze).
> I did:
> > One could argue that, even if you are involved in RC bug squashing, you might
> > not be making uploads that frequently, and that would explain why the number of
> > distinct uploaders is going down. So let's take a step backward:
> > Distinct uploaders to unstable from 01/01/2008 to 30/07/2008: 1229
> > Distinct uploaders to unstable from 30/07/2008 to 14/02/2008: 978 (lenny freeze)
> > Distinct uploaders to unstable from 06/02/2010 to 06/08/2010: 1187 (6 months before freeze)
> > Distinct uploaders to unstable from 06/08/2010 to 05/02/2011: 926 (squeeze freeze)
> The 20% was based on those numbers, not on the per-months data. If you
> take the per-month data, it's much worse, e.g: 2010-01: 661 uploaders ;
> 2011-01: 313 uploaders.
Oh, sorry, did miss that part. But as I wrote, without one specific package of
mine I'd guess I had an avarage of half a upload per freeze (and even on
that 'many' release related uploads I'm almost proud) so with 40% of
people in that situation you'd already get those numbers. (I do not
claim there is no such effect you proclaim, just that your numbers in my
eyes mean that it cannot be nearly as strong as you suggest).
Bernhard R. Link