Proposal to reduce anti-bundling requirements

[Matthew Miller]

On Wed, Sep 30, 2015 at 04:52:48PM +0200, Ralf Corsepius wrote:
> >people not declaring their bundles and not care about policies did the
> >same before: not declare it and not ask for exceptions - there is a
> >logical flow in "now that i don't need to ask FPC i don't declare it"
> Exactly, that's what I would consider a serious regression.

But re-read what you're just replying to. The thing is, there's all
sorts of bundling going on all throughout Fedora under the radar. It's
like banning alcohol. It might seem like a policy that'd benefit
society overall, but turns out that it doesn't make people stop
drinking — just drives it underground. This approach encourages people
to, uh, drink responsibly. There's still regulation and labeling.

Now, if only we could figure out how to _tax_ bundling, we'd be all
set.


> This proposal effectively is a carte-blanche to bundling and
> carelessness, which I would expect to seriously impact the quality
> of Fedora.

It seems pretty careful to me, and a lot more prescriptive than
"carte-blanche".

> 
> >the opposite is more likely: people trying to avoid the FPC burden now
> >can declare it without fearing somebody takes notice and points out a
> >violation
> If they don't care or are not aware about the consequences of their
> bundling?
> 
> Like I've said many times before, I feel Fedora needs a serious
> vulnerability in a widespread bundled or static library, such that
> people finally comprehend the harm of bundling.

Of course, we've certainly had this happen in the past. I hear you
saying that we're losing that lesson and need to re-learn it. But, I
don't think we've actually forgotten it — I certainly haven't. I just
think we need to try a different approach, because the old one worked
well to a point, but also comes at a big cost.


-- 
Matthew Miller
<mattdm at fedoraproject.org>
Fedora Project Leader