The big picture is - If you continue to have (how many distros are there now 300+) lots of
distributions then all efforts are duplicated, triplicated etc etc, resulting in a lot of
wasted effort. (If the kernel development was to split into 15 different seperate kernels
would that still work as efficiently as it does now. probably not.)
Redhat, suse, ibm, HP, ubuntu etc will still contribute because like you say we all benefit
from the community around open source software. The kernel, samba, apache, that kind of
enterprise software will continue because they all contribute to it and need it for their
business to survive.
What does linux run on now, mobile phones, super duper multi 57 processor super computers,
500+ network clusters but if I upgrade mplayer to convert a video to an ipod playable file
then it can take me 4 days to find out that the latest release has broken ipod/ffmpeg
compatibility, but it was spotted in ubuntu and that was fixed but not in fedora and only in
debian experimental. That is helpful to me.
The desktop is still a problem area, and there again must be millions of people trying linux
and being put off by the desktop usibility (for the want of a better phrase). I dont think
all home/multimedia users expect to get all there apps for free, but how do they financially
contribute. Do they contribute to each and every piece of software that they use or do they
donate to gnu or what. A single place that pushes the desktop/multimedia as hard as the
commercial distros push the serverside might address the balance.
Now if everybody contributed to one distro. (To be fair it doesnt have to be debian) and make
that the best stable system then the commercial businesses can work from that. In the end it
means even less work for them because it will have potentially millions of users reporting