I think most of it will be done from scratch actually. For a few reasons. One
being consistency, two being quality, and three being I don't think we can
focus on Trisquel initially.
What I'm thinking is a book to help novice users to get up to speed on free
software. Even while these users might be on free software friendly hardware
Trisquel hasn't quite caught up to the usability that is needed for this
base. This is not to say it doesn't work really well. The pieces which are a
problem tend to be beyond the scope of the distribution.
Since we have to start somewhere I'm thinking of doing is sticking to the
current path. Free software friendly hardware → teach the free software
friendly apps (to the extent we can, make it easier to transition users
later) → develop solutions to common non-free problems → actually
transition users.
As I've mentioned before various problems with STS/LTS release model that
isn't so much what I'm talking about here. Mainly things like limited
entertainment/'gaming' options and similar. Trisquel also seems to be buggy
enough still that it would be too significant a deterrent. The last thing I
want is to push a solution that will turn users off of free software in the
future. Something I know a lot of us have done and/or continue to do.