@Scott:
Ack, the remote use-case is indeed a valid argument. I'll rediscuss that with Jean-Baptiste once he's back mid next-week and we'll fix it (I guess with a separate 'fsimage-layered:', but I'll wait on jibel) and give another round of testing as those layered systems only makes sense for a local filesystem (otherwise, you would just point to one, pre-compiled, full squashfs IMHO).
@Scott:
Ack, the remote use-case is indeed a valid argument. I'll rediscuss that with Jean-Baptiste once he's back mid next-week and we'll fix it (I guess with a separate 'fsimage-layered:', but I'll wait on jibel) and give another round of testing as those layered systems only makes sense for a local filesystem (otherwise, you would just point to one, pre-compiled, full squashfs IMHO).