Atomozero I fully share Kevin's concerns regarding fragmentation. Objectively speaking, I believe a true fork of Haiku is humanly impossible right now: we simply lack the human and technical resources to sustain it, and we would risk burning out and seeing the project vanish after a few months of initial enthusiasm.
Agreed 100%
Atomozero However, I see a substantial difference between a 'fork' and a distribution based on Haiku that adheres to certain standards of 'Historical Elegance.
I wouldn't even use the word "distribution".
I have long argued that there is space for "flavours" of Haiku. A flavour for gaming, one for graphics, another for productivity. Not an incompatible mess, just an ISO pre-loaded with that kind of software, all on the same base. The graphics flavour would come with GIMP and Inkscape preinstalled. The Productivity flavour comes with LibreOffice, and so on. In theory, you could install the gaming flavour, change your mind, and then, if you sit there giving pkgman commands for half an hour, change it to the productivity one.
There could even be a minimalist flavour that comes with WebPositive, Terminal, one text editor (not three) and nothing else, not even gcc. Want gcc? Install it yourself.
Of course, I was arguing in the context of a future post-beta situation, but the idea could be brought forward.
So, what would be on the DoF "flavour"? Gentlemen, start your keyboards.