Reasons for hall monitoring

[Patrice Dumas]

On Sun, May 09, 2010 at 10:16:45PM +0100, Adam Williamson wrote:
> 
> I don't agree with that, entirely. Think about it - Red Hat sells big
> enterprise stuff. Mostly servers. Directly, PA and bleeding edge X stuff
> isn't of huge immediate interest to RH. I mean, of course RH is going to
> pay people to work on stuff it thinks will ultimately benefit it, but it
> does take a rather broad and long view of this (think how long it's
> taken for PA to even be in RHEL at all). But we are a _general_
> 'playground' (or rather sandbox), for the development of interesting and
> useful bits of technology. I think you can look at Fedora almost as a

Not so general. For example grub2 or upstart weren't really pushed. 
Or initng didn't got much support. There are also choices regarding how
well the different desktops are taken into account when a feature is deemed
to be integrated enough. My observation is that really new directions 
and most innovative integration, especially at the level of the basic
operating system tasks (kernel, PA, X, policykit, devicekit...), but also 
for important technologies and applications (firefox, java...) is in general 
done by redhat people using redhat developped technology. I don't follow
the Fedora development very closely anymore, but it seems to me that
it was especially true in the past. When it is not an innovation lead
by redhat people, it is quite the struggle to get it in -- though it
remains possible.

Please note that it is not at all a criticism of RedHat or Fedora,
it is just my impression of how things work.

-- 
Pat