No flaming here buddy...

But, I can see some inherent problems with putting what I'll term "Meta-Data" about the Primary "Data" in a Directory service or "Alternate Data Function"... Okay, now for it to actually work properly with the WAY filesystem currently work, you'd have to put a "FAM" or File Alteration Monitor in place, or an Abstraction Layer called virtually which actually does the I/O.

Performance hit. Possibility for "Interception/Alteration" is far greater. If you were to slop this on top of EXT2/3 and Make it EXT4... Well... I can just imagine... It'd be GREAT to support all of this. But I just can;t see the current Format or Structure of the resaonable filesystems supporting it well.

Now there are only couple of filesystems I see capable of transforming into a "Multi-Stream Journaling File-System" currently, one of course ReiserFS. It already is a form of Database under the traditional inode system. I am sure others will disagree with me... BUT...

adding a "directory" ontop of an existing filesystem will always give you fragility and slower performance {hence NTFS fragmentation performance sucks) once the filesystem ages. Also, this would apply more to MAC HFS+ also. I am guessing why we haven't really seen it, is just that we've never really seen a NON HFS filesystem on MacOS. OSX... now we can see it. I'd like to know how much difference we really see in speed from HFS+ to UFS. IF any.

Blah, Blah, Blah... Yackety... Schmackety... I am just rambling now.