Summary of the fdo disussion at GCDS

[Jakob Petsovits]

On Thursday 09 July 2009, Sanel Zukan wrote:
> > * if we do take into account other projects, who are already involved
> > here such as XFCE/LXDE/Enlightenment, then we probably need to find a way
> > to weight things so it's still representative (similar to how it's common
> > to have a House of Representatives where each state/province/canton gets
> > a # of seats relative to population size). how it should be weighted is
> > probably a pretty lengthy discussion, and still doesn't take into
> > consideration future changes. in any case, it could be as simple as
> > saying that "XFCE, LXDE, Enlightenment, etc." represent one bloc in
> > equality to KDE and GNOME, and a spec needs 2/3 blocs with no bloc
> > registering an objection. regardless of the actual weighting, the
> > measuring is easy as long as we document acceptance in a machine-readable
> > format. so the weighing would be light weight as a process, but coming up
> > with how to weigh will not be.
>
> IMHO this sucks. E.g. if XFCE, LXDE, Enlightenment and XYDesktop comes
> with needed (and pretty cool, at least for them) spec, KDE says "maybe"
> and GNOME says "no" (due policy or something else), then spec does down.

How about assigning a number of weight - or call it votes if you want - to 
each desktop (e.g. KDE/GNOME have a weight of 2, other desktops are 1), and 
instead of requiring a majority you just require a certain total weight
(e.g. 4) to be reached. Examples:

KDE+GNOME = 2+2 = 4 = pass
GNOME+XFCE = 2+1 = 3 = not yet pass
GNOME+XFCE+Enlightenment = 2+1+1 = 4 = pass
XFCE+LXDE+Enlightenment+XYDesktop = 4*1 = 4 = pass

That's still a significant hurdle for niche-desktop-only specs, but if four 
niche desktops agree on something then there must be something to it either.