The vision for the Fedora Workstation

[Bastien Nocera]

----- Original Message -----
> On Tue, Feb 11, 2014 at 08:26:18AM -0500, Bastien Nocera wrote:
> > We fix bugs upstream so they're not in Fedora, and we fix bugs in Fedora
> > so that they're not in RHEL. So the least amount of time we spend on
> > RHEL bugs means that the community/upstream versions are of better quality.
> > 
> > The RHEL bugs obviously do take precedence, but that doesn't mean it
> > is what we spend most of our time on.
> 
> Which is good overall -- I'm glad Red Hat pays people to work on upstreams.
> 
> It raises a question about this particular resourcing side of the argument,
> though. Given that upstream and RHEL take some of the focus, the number of
> total full-time-equivalents working on Gnome in Fedora is smaller
> than the total number of people employed. It's still a significant number,
> but given that there are multiple desktops with vibrant upstreams, it's not
> really fair to count in any other way. The downstream work is the same way:
> that will happen regardless of what Fedora does. So, counted that way, I
> don't think it's really an overwhelming case.

When we fix a RHEL bug, it doesn't happen in a vacuum either. It's likely a bug
upstream as well, or it's an integration problem. The latter can help create
better documentation and tests, the former means that upstream and Fedora are
also benefiting.

The main difference between GNOME and other desktops when it comes to this support
is that we have people in the Red Hat desktop team that can cover almost all
of the different parts of the system. If I have a problem integrating with a base OS
component, input methods, or device drivers, I know that I can find somebody
that can make it their focus.

> Unless Red Hat desktop team is interested in shifting some of its focus to
> Fedora itself. From what Christian says, I think that actually *is* the case
> -- if Gnome is chosen as the Fedora Workstation desktop, that means more
> people given time to work on that specifically. Right?

I doubt we'd want something in Fedora that's not upstream, and vice-versa. I
can't think of any case where working on something to be ultimately integrated
in Fedora or RHEL (say, Wacom tablet support, better Terminal integration, etc.)
couldn't be worked on upstream.

> Of course Red Hat is also are interested in Fedora as supporting other
> upstream projects Red Hat cares about, but that's a separate thing.