Subject: Re: [netatalk-admins] Netatalk +hfs +ASIP
From: Osma Ahvenlampi (oa@spray.fi)
Date: Thu Mar 19 1998 - 14:23:06 EST
David Perry <deperry@nerosworld.com> writes:
> With the hfs support does netatalk read and write the files in the same
> format as a native mac drive? That is to say, does it still write files
> with the .AppleDouble thing going on or is it all one usable file? Any
> details on the implementation woudl be appreciated.
With hfs support, you can mount a Macintosh drive (move it to the
Linux box - I don't know if you can _create_ Mac filesystems on Linux)
and share the files on it. The hfs filesystem can show the resource
forks to the Linux API as an .AppleDouble directory so that netatalk
finds it as it should - naturally, since the filesystem itself knows
that the data and resource forks belong together, you avoid the
problems of stray files in .AppleDouble when a UNIX or Samba user
moves or deletes files. As far as the filesystem is concerned, there's
one file with two forks, and if you delete the data fork, the resource
fork disappears as well. As far as UNIX API and tools are considered,
they see two files.
On the other hand, hfs is a limited and slow filesystem format
compared to ext2. For a volume shared to MacOS and Windows users, it
might be useful, but if the access to the volume is strictly from Macs
only, you'd be better off using ext2.
-- You can tune a piano, but you can`t tuna fish. Osma Ahvenlampi <oa@spray.fi>
This archive was generated by hypermail 2b28 : Sat Dec 18 1999 - 16:31:52 EST