:: Re: [DNG] Why Debian 8 Pinning is (…
Forside
Slet denne besked
Besvar denne besked
Skribent: Jaromil
Dato:  
Til: dng
Emne: Re: [DNG] Why Debian 8 Pinning is (or isn't) pointless

dear Rick I did not expect you to use 'shutdown -h now'!
the real old school thing is 'init 0' (or 'init 6' for reboot) :^)

On Wed, 13 Jul 2016, Rick Moen wrote:

> > In fact, the outcome of Ian Jackson's GR enforces their right to
> > such sabotage.
>
> That is not an accurate representation of the GR outcome


I believe this is the best accurate representation
http://forums.debian.net/viewtopic.php?f=20&t=120652&p=576562#p576502

in that GR the Debian leadership has allegedly allowed the democratic
governance of the Debian project to be boycotted. worthed re-reading
now as dasein just updated it with a more detailed procedural
explanation of his vote analysis.

> I've been explaining to people the essential role of forking in open
> source since at least that 1999 essay that Slashdot picked up.
> http://linuxmafia.com/faq/Licensing_and_Law/forking.html


thanks for this, I did not knew it back then and is a fascinating
read. I believe there is only one thing that you omit in your general
consideration on the faith of forks: that there is also a
*sovereignty* aspect to them.

If the reason for a fork is to acquire sovereignty, it may well be
that it will survive, thrive and grow on its own. If the reason is
simply that to make an interface user friendly or a sourcecode look
better, it will not. Devuan is born out of the need for sovereignty of
definitely more than half of the folks who relied on Debian's
governance and the work that was being done on it by volunteers.

For us, the systemd issue is just the last (big) drop in a full
bucket. We believe that Debian's governance is corrupted and
corruptible.

Therefore, by now, we need ultimate control and outreach to override
any decision in Debian. Be it in 1 year or 10 years from now, we may
go our own way. Even if Debian 9 comes out with openrc, I can
hardly imagine anyone among our developers will ever go back to it.

We need a reliable APT based core distribution to build upon, which is
not out for grabs for the next tribe of pubescent rockstar-coders
wanting to prove themselves worthed of changing history. Let Debian do
the innovation dance, meanwhile we have actual work to do here and
that cannot be stepping back to adjust a core, but eventually work on
what you call 'local implementation of policy', what Katolaz calls
'duck tape', in the most minimal way possible. Because that is the
layer where liability falls on actual humans and operators.

To those who aren't convinced I just say good luck blaming a kid when
the shit will hit the fan in the farm you are liable for.
I believe I do not need to explain you how many mission critical
deployments around the world rely on Debian already. As their workload
in keeping it stable will increase, we can thrive off the workload
we'll peel off for them just by providing a proper base system.

Can one do this by pinning? I doubt. The complexity for 'local
implementation of policy' will just keep growing and growing.

ciao