[linux-elitists] Why haven't you switched to MacOS X yet?
Sun Jan 12 05:00:01 PST 2003
Quoting Karsten M. Self (email@example.com):
> Rick may have some cautionary suggestions for mixing GNU/Linux and OS X
> use of common partitions owing to peculiarities of OS X's (mis)use of
> the filesystem and resultant fragility which may result. I've witnessed
> several such discussions, can't say I've firsthand knowledge.
Hmm. OS X can deal with its version of UFS, HFS+, and HFS. Last I
heard, nobody had functional Linux support for HFS+. One might want to
create an old-style HFS partition using MacOS version-something to serve
as common ground, but you'd have to go out of your way to create it.
That leaves OS X / Apple Darwin's version of FFS/UFS. Linux kernels
have included UFS filesystem support for a long time, but the fly in the
ointment is that UFS implementations are pretty varied among BSD-ish
'nixen -- which Linux attempts to account for using a "ufstype" mount
option (nextstep, openstep, 44bsd, sun, etc.).
My recollection is that you do "mount -r -t ufs -o ufstype=openstep ...",
to mount an OS X "UFS" partition within Linux. And I'm pretty sure it
works just fine.
What Karsten probably has in mind is OS X's storage of old-style MacOS
"resource forks" as separate files in the same directory as the related
data-fork portion of the composite file, with the same name except with
"._" prepended to the filename.
To explain: Old-style MacOS files could have contents stored in either
or both of two extents, the data fork and the resource fork. In order
to usefully host such a file on UFS (which otherwise has no provision
for handling multiple branches of a file's data), OS X uses the "._"
prefix trick to emulate the desired semantics using two physically
distinct UFS files to hold that data.
Apple's trying to phase out resource forks, so the issue will eventually
If you use tools that copy or move the data fork without the resource
fork or vice-versa (e.g., GNU tar, GNU cpio, rsync) to or from such a
UFS partition, the destination won't receive the entire file.
(Obviously, this is a problem only for files with both forks.) And, of
course, watch out for handling of such files on and off HFS: I can't
even remember how Linux sees them.)
Cheers, "My file system's got no nodes!"
Rick Moen "How does it shell?"
More information about the linux-elitists