[CentOS] OT: RHEL 6.1 is out
R P Herrold
herrold at owlriver.com
Mon May 23 02:14:52 UTC 2011
On Sun, 22 May 2011, Steven Crothers wrote:
> I think you're missing the point, if you read between the lines, the
> complaint I see is that CentOS (Community Enterprise Operating System)
> is not community based whatsoever.
I don't mind-read as to what a third party meant so well as
you, it seems
My intent with cAos (post fedora.us), and with CentOS was to
keep available for the FOSS development community at large,
the fruit of the distribution integration represented in the
'testers-list' non-public beta group for the former RHL, and
the years of work represented there, by people both outside
and inside Red Hat. It initially appeared that there would
not be a binary form integrated distribution in RPM packaged
form. Greg of cAos indeed re-worked a fairly initial
installer called 'cinch'
It was not at all clear that Red Hat would not threaten
litigation to close such efforts down. They had made such
threats previously to one of the other co-founders of the
CentOS sub-project of cAos, as to a RHL rebuild and respin he
had marketed
To suggest that CentOS is 'not community based whatsoever'
will come as a great surprise to the donors of bug triage
effort, of mirroring effort, of wiki authoring, of forum
participation, of live-CD 'mixing', and so forth
But as hughesjr mentioned just last week, letting random
people (seemingly a 'community of random and untrusted
persons') feed content that would end up signed in the CentOS
project's name, is simply not going to happen. CentOS has
never been about that
A 'vetting' and reputation system was proposed in some early
design documents for fedora.us, but that project lacked the
mass to make it work; cAos tried a variation of this, and
encountered a problem with its v.2 when a novice packager
inadvertently introduced a 'one way' library version bump,
impairing the maintainability of that release going forward;
The ATrpms v. DAG archive approach on pushing new versions of
certain core packages shows two approaches, and the DAG
non-invasive approach is clearly the mind-share winner --
We've [the third-party packaging community] (at least, I've
been in projects that have) tried variants of 'anyone's code
is welcome' distribution adjunct preparation before, and it
does not work well
CentOS binaries creation process is by and large is a very
literal and non-creative effort
If people want to start their own rebuild efforts, peace be
with them, and good luck. But fostering spin-off's is not
what CentOS is about -- and people railing to the heavens
about how unfair it is that THEIR false expectations (based on
some amorphous vision of how great something COULD be, if only
... ) are not met by the CentOS core team, are simply making
noise here
-- Russ herrold
More information about the CentOS
mailing list