> There is an introduction to Coda that covers these issues. It looks > promising:
Sorry just a final comment about this for anyone that's interested. 1. Unofficial Debian packages that you can add to sources.list are available from here: http://www.coda.cs.cmu.edu/mirrors.html Debian APT package installer sources.list line Upgrading an already running Coda client typically fails somewhere, so be careful when adding this. deb http://www.coda.cs.cmu.edu/pub/coda/linux debian/binary-$(ARCH)/ 2. Coda requires a minimum of two raw partitions to provide log and data storage (although loopback file access is possible--and not recommended). 3. This doesn't give access to all the files on the server computer. 4. There is a distinct separation of clients and servers. 2, 3 & 4 mean Coda is not suitable for what I want to achieve. I'll just have to accept the limitations of NFS and replicate all the partition mounts at the client end. Regards, Adam
