rsync - Oct 2006 - Extended attributes in MacOS X

I've written a patch to 2.6.8 which adds extended attribute support  
to rsync. This is not like the plethora of patches which try to  
separate osx meta-data into a  special appledouble file (._filename).  
Instead, it builds on patches/xattr.patch and just adds support for  
the MacOS equivalent.

The good news:
1. It works fine locally and to a remote mac
2. It also preserves the finder info and resource fork (these are  
exposed as EAs com.apple.{Finder,ResourceFork})

The bad news:
1. EAs on MacOS need not begin with 'user.' like they do on linux.  
Therefore, I can't backup from a mac->linux machine. However, linux- 
 >mac->linux works fine since the Mac preserves the 'user.'
namespace.
2. 2.6.9pre2 has some sort of bug where it zeros out the datum_len of  
the last rsync_xa in a file. I suspect a buffer overrun somewhere,  
but couldn't find it, so my patch doesn't work on 2.6.9pre2.

My question: is this useful? should I prefix all MacOS EAs with 'user.'?

On Wed, Oct 25, 2006 at 05:57:38PM +0200, Wesley W. Terpstra
wrote:
> My question: is this useful? should I prefix all MacOS EAs with
'user.'?
I'd love to see such a patch, as this is the direction that I'd like to
see rsync take in handling MacOS extended attributes.

As for prefixing all MacOS xattrs with "user.", we'd need to come
up
with some rules that would preserve the xattrs for all data paths:
Mac -> Mac, Mac -> Posix -> Mac, and Posix -> Mac -> Posix.  So,
a Mac
receiver would need to be able to differentiate what entries needed the
extra prefix stripped, and which did not.  So, perhaps adding something
like "user.N0nP0s1%." to any non "user." data (that's
NonPosix spelled
with 2 zeros, a one, and a percent).  That would make the stripping
rule pretty straight forward on the Mac side.

..wayne..