Re: Migration of non-free packages to testing
On Thu, Nov 14, 2002 at 06:41:55PM +0000, Andrew Suffield wrote:
> On Thu, Nov 14, 2002 at 10:00:00AM +1000, Anthony Towns wrote:
> > 451 (88.4%) RC bugs in main,
> > 25 (4.9%) RC bugs in non-US/main
> > 16 (3.1%) RC bugs in non-free
> > 13 (2.5%) RC bugs in contrib
> > 1 (0.2%) RC bug in non-US/contrib
> > 4 (0.8%) RC bugs in pseudo packages
> > While those numbers aren't particularly firm and shouldn't be relied
> > on too heavily, they do seem to indicate that non-free isn't a whole
> > lot buggier than main -- based on the proportion of non-free packages
> > we have, you'd expect around 14 RC bugs, to main's 450.
> I postulate that significantly less people use the non-free packages,
> and that this has a significant impact on the number of bugs filed.
> [I think the "X is more buggy than Y" notion is an unfalsifiable
> hypothesis in all cases]
*shrug* I've never seen much point in choosing definitions for terms
that specifically make them unscientific or meaningless.
If you prefer, consider the concept of "adequately maintained" or
"suitable for use". non-free software has some obvious hurdles already
for this (in that it's not legal to use it in some ways, or maintain
it in others, by definition) but once you've ignored those, people who
actually have anything to do with it don't seem to find its brokenness
particularly more alarming than the people who use software in main.
In any event, of the ~10k packages in main, I expect you'll be able to
find a couple of hundred that're used as rarely as most non-free stuff.
Anthony Towns <firstname.lastname@example.org> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
``If you don't do it now, you'll be one year older when you do.''