So is my problem really just the networking protocol (smb)? Or is it also the filesystem of the server?
It is the networking protocol. This is how the local filesystem is represented "on the wire" (over the network). If there was a network protocol that understood the semantics of HFS+, the filesystem would be better represented over the network and resource forks wouldn't be an issue.
For the time being, HFS+ is shoehorned to work over NFS/SMB and there will be some comprimise.
Quote:
Anyone know of any tutorials on how to configure NFS?
If you have specific questions, I'll be happy to help. In the meantime, NFS has been around since the late 80s, so there is TONS of info on the web.
The common problem is the namespace of the client must match the namespace of the server. If user "jimbo" is UID 1001 on OS X, user "jimbo" must be UID 1001 on the server (common sense, really). Directory services such as NIS, NIS+, and LDAP exist to keep the namespace consistent between NFS clients and servers however they're overkill for the average home network. Just keep the passwd files (specifically the UIDs and GIDs) in sync between client and server.
The other thing to watch are UNIX permissions. If user "jimbo" doesn't have permissions to write in a local directory, he also won't have permissions to write over NFS.
I have worked on NFS as user/administrator and developer for over 10 years now.
Quote:
The .DS_Store files aren't so bad. I just hate to see a ._blahblah file for every single file I copy over from OS X. This eats up more disk space.
If you want to keep the resource fork, there's little you can do.
Comments
Anyone know of any tutorials on how to configure NFS?
The .DS_Store files aren't so bad. I just hate to see a ._blahblah file for every single file I copy over from OS X. This eats up more disk space.
Originally posted by Macvault
So is my problem really just the networking protocol (smb)? Or is it also the filesystem of the server?
It is the networking protocol. This is how the local filesystem is represented "on the wire" (over the network). If there was a network protocol that understood the semantics of HFS+, the filesystem would be better represented over the network and resource forks wouldn't be an issue.
For the time being, HFS+ is shoehorned to work over NFS/SMB and there will be some comprimise.
Anyone know of any tutorials on how to configure NFS?
If you have specific questions, I'll be happy to help. In the meantime, NFS has been around since the late 80s, so there is TONS of info on the web.
The common problem is the namespace of the client must match the namespace of the server. If user "jimbo" is UID 1001 on OS X, user "jimbo" must be UID 1001 on the server (common sense, really). Directory services such as NIS, NIS+, and LDAP exist to keep the namespace consistent between NFS clients and servers however they're overkill for the average home network. Just keep the passwd files (specifically the UIDs and GIDs) in sync between client and server.
The other thing to watch are UNIX permissions. If user "jimbo" doesn't have permissions to write in a local directory, he also won't have permissions to write over NFS.
I have worked on NFS as user/administrator and developer for over 10 years now.
The .DS_Store files aren't so bad. I just hate to see a ._blahblah file for every single file I copy over from OS X. This eats up more disk space.
If you want to keep the resource fork, there's little you can do.
Originally posted by mikef
...If you want to keep the resource fork, there's little you can do. [/B]
Hmm? Then what's the advantage os NFS over SMB?
By the way, thanks to everyone for all the input on this!