Adam Williamson (awilliam(a)redhat.com) said:
> Benefits:
> - allows roughly 1/6 of the world's population to use Fedora freely
>
> Demerits:
> - requires ongoing maintenance work on some packages
> - may require removing packages that can't comply without being broken
>
> I feel the benefits in this case outweigh the demerits, and the
> amount of work required to be greatly exaggerated. Furthermore,
> making Fedora available for all to use freely is a fundamental
> goal of the project; ensuring the presence of, say, gcompris in
> a form that exactly matches upstream is much lower down the totem
> pole.[2]
> Now, if we can discuss the benefits and demerits without resorting
> to reducuing it to 'aah! slippery slope' or 'I'm offended by
> yellow, take that out too!', it would help, as those are sort
> of missing the point.
I don't think they're missing the point, as they raise valid problems
with your categorization. I don't agree with your 'Benefit', for the
following reasons:
1: a lot of that 1/6th of the world's population does not own a
computer. Or an internet connection. (Or, in many cases, a reliable
electricity supply). Let's not have any illusions about China: it
contains a huge amount of people, but a rather smaller amount of
possible potential Fedora users.
You could say that about many countries. In any case, even if 1/10 of one
percent of those people are viable users... that still dwarfs the affected
packager base by many orders of magnitude.
1a: Fedora is not, in point of practical reality, unavailable to
China
at present. Even if, by official Chinese government policy, Fedora
contains material that should not be distributed in China, it has been
reported that - in practice - it is perfectly possible to download
Fedora in China from many different mirrors, flags and all.
Possible? Sure. It doesn't mean it's necessarily a good idea to
knowingly provide software to people that can get them in trouble
with their local authorities.
2: it has not by any means been established that solely removing
flags
from the distribution would be sufficient for the Chinese government to
be happy with Fedora being actively promoted / distributed within China.
Given that related operating systems with *ONLY* these changes are allowed,
it's a fair assumption to make. Honestly, I think the proof is on others
that it wouldn't be enough, given that existing evidence points to
the positive.
the best way to solve the problem. Other proposals have been made -
such
as delegating the modification work to some kind of SIG, working on a
special spin of Fedora for China - and I haven't seen anyone explain why
that's a worse idea than making all the changes directly in the main
Fedora package repositories.
Proliferation of spins and maintenance for specific geographies is
a waste of space and effort, if it can be avoided. That's why we
have languages included on the Desktop spin, instead of 15 different
localized ones.
Bill