We should go through all the kernel config options and decide which are necessary when e.g. both emenlow and atom-pc configure VFAT_FS into the kernel - if that option is needed mainly for boot and would typically be used only rarely after that, it should probably be built as a module. In general, the question of whether to prefer modules or built-in options has come up on several occasions - we should decide on some policy and audit the options in light of that. This bug is mainly a placeholder for that task, so we don't forget.
This is an ongoing thing. We need to break it up into concrete sub-tasks and either get those on the schedule or into individual bugs.
Tom, where would you say we are with this? Sep 6 is the cutoff for Medium importance bugs. I know I haven't refactored atom-pc.
I told Song we've probably done as much as we can for 1.1 and should probably close this out as we said we would and replace it with more specific bugs for 1.2. Unless you can think of something pressing that could be worked on under this bug id for 1.1.
Agreed, let's close it out.
Closed as discussed.