Thanks for the clarification. I'm envisioning the distro itself being
distributed in two forms:
a) with no games. Upon first boot, the new user is guided on how use the
not-Steam game library client to browse and install games from the repos.
This could done with a pop-up tutorial that can be easily cancelled out of,
and called up again.
b) With a handful of games that show off the best of what libre games (both
code and artwork) offer across a range of genres. Ideally involving a number
of games that can be played over a network, with an updatable list of libre
game servers installed by default. Maybe invites to Libre Game Night
pre-loaded into a desktop calendar too ;) Gaming is a social activity for
many people, and I suspect that being actively invited into communities of
libre gamers would stop some gamers getting bored and giving up.
So the core distro, like default Debian, would be 100% libre (including all
artwork). The games themselves would be packaged in a set of repos,
subdivided in a similar way to the Debian repos into:
* 100% libre games
* games with 100% libre code but nonfree artwork
If making the nonfree repos available by default would keep the distro from
being endorsed by the FSF, that's a shame. My view is that the project would
fail to live up to its potential if it didn't offer users easy access to
every game that runs on an otherwise libre distro. However, getting the GOLD
project off the ground will require a collective effort, and if the majority
of those involved prefer FSF endorsement and a smaller library of games, I'm
happy to abide by the consensus of the working group.
It may be possible to make the nonfree games available as a totally separate
project, that can be plugged into the GOLD distro by users who really want
them (using PPAs, or one of the newer packaging systems like Snap or
Flatpak). I would even consider this for a separate repo for games that
contain nonfree code (eg scripts included with artwork), but can run properly
when the core distro doesn't contain any nonfree code, and whose copyright
owners allow non-commercial distribution. My short-term goal is to build
something that is significantly more libre than SteamOS, and can be a
stepping stone towards a 100% libre gaming experience on Steam Machines (and
ideally other gaming hardware).
