I hope you are aware of the fact Jon did not left kb dev, he just delayed it a bit and the topic was taken out because of some security issues (look at my sig).
And the thing is, you don't really want to update this mod, you'll have to rewrite almost all of it if you really want it working fine and fast.
I know quite well the way it is built for now, and I can tell you, any update at this stage means to change almost all of it (unlsess you want ot keep the outrageous sql queries
It's really cool that you are willing to spend time on this, but I am not sure branching is the best solution.
And for the mx compatibility, I know it may seems odd to some, but it is no big deal to maintain two paths scenari in the code.
Anyway, I have almost fixed all of the current version, as you may know, so what are your plans ?
I think since many people worked on this project and this is not a dead project at all, we should better think about joining our efforts here(we, the mx Team, are running CVS and such on this perticular one, we know it well, and again will continue to dev it).
But as always, you are free to decide, I just think it could en up to be a waste if you just come up with a fixed version (are you really going to find more fixes than I did, hope so, but ...) while we release a totaly new one, optimized and such, and even worst if you really recode it all
I think we should at least talk a bit about this before you spend too much time on this.