On Sat, 28 Aug 2010 17:16:12 +0000, you wrote:
We have a clear cut example what can happen ( The Oracle and
OpenSolaris
case ) so first and foremost we need to make sure that a single
corporate entity does not have control over the project which means that
no corporate entity may have majority of it's employ in boards, steering
committees, SIG etc. to prevent that corporate entity seizing control
over the projects, steering it's directions to it's will and for us as a
community not being too depended on that corporate entity should that
corporate be bought or otherwise be controlled by individuals that do
not share the same vision as the community.
It's not far from reality that Red Hat will get bought by a company like
Oracle so what's preventing us to get the same treatment as OpenSolaris
got?
1) Given the significant amount of work that Red Hat does in the Linux
ecosystem - kernel, gcc, glibc, xorg, gnome, etc - every Linux
distribution out there whether it be Fedora or Debian or any other is
dependent on Red Hat. There is no simple way of getting around this
reality unless someone has a lot of money to spend hiring developers.
2) The Solaris/OpenSolaris case under Oracle demonstrates the
importance of licence choice and the difficulty of taking a large,
proprietary project open source. This does not apply to Linux. Unlike
the Solaris CDDL the GPL, combined with the variety of contributors,
means it is pretty much impossible to take Linux closed source without
violating the licence. In the case of OpenSolaris they never grew a
community of contributors, the licence allows them to do binary only
distribution, and they never fully opened up Solaris hencing making a
true fork of OpenSolaris difficult because the source is missing from
some key parts.
The first and foremost mission that the project should be doing at
this
point is to make dam sure we are ready for it encase Red Hat will not be
there tomorrow or it's there but not in the same form that we all love.
While some planning will help, if Red Hat goes then Fedora will be the
least of the problems the community faces. As such it is not worth
worrying about.
The first mistake we did was trying to label end user since it's
not up
to the project in whole to decide which end user type it's target.
It's should be up to individual community SIG's to decide what user base
they are targeting and the form they will present that to the end user
in live cd or a predefined installation option be it with the latest and
greatest bits of their product or a not which may or may not be
influenced from feed backs from the micro community they have
established around the product they ship.
Sorry, but won't work. As much as your or I may wish otherwise people
in general won't differentiate between Fedora Gnome/Fedora KDE/Fedora
cutting edge/Fedora stable/ etc. It will all get shortened to Fedora,
and thus impact on the project's reputation as a whole. While the
SIGs are useful to customize Fedora around a particular purpose, all
the SIGs must pull in the same general direction.
The Fedora project in whole should give equal access to those bits and
devote equal amount of marketing resources to promote them.
Fedora, like most (all?) other opensource projects suffers from a lack
of manpower and resources. As long as this is the case there will be
projects that suffer.