>Matthias Urlichs wrote:
[...]
>> IMHO, we should start with specifying what the solution should look like.
>>
>> - access to normal files must be identical whether or not the file has
>> á attributes or whatever (i.e. no AppleSingle solution where the metadata
>> á for a file reside at the beginning of said file, no "the file's data are
>> á really in the file 'filename/data'"),
>> - a normal recursive directory copy (via NFS, or whatever) will also copy
>> á the metadata
>> - there should be no visual clutter (i.e., no AppleDouble solution, where
>> á the metadata for "foo" reside in "%foo" or "foo.res")
>> - no kernel hack, because the solution needs to work on systems where we
>> á cannot _do_ any kernel hacks.
>>
>> The only solution I can think of which supports all of these is to use some
>> sort of dot.directory.
>>
>> --
>> Matthias Urlichsááááá |ááááááá noris network GmbHááááá |áááááá smurf@noris.de
>Ummmm....á All of the above sound fine except the last bullet and the
>conclusion.á As far as I can see, the absolute only way to handle this is
>kernel support.á Hopefully, someone comes up with a cross-platform
>metadata-interface standard.á Until then, the big thing is to support the
>metadata in a nice way.á Hopefully, utilities and programs will be
>modified to make use of the feature.á None of them should depend on it.
I agree. It is hard to make all unix vendors implement metadata at the
same time, someone will probably have to be first. This could be linux.
An implementation where the metadata works but nothing depends on it looks
like a good idea to me. Utilities will simply loose none-essential
functionality when accessing files over NFS. Such as showing generic
icons instead of file-specific, editors not remembering the last position
in that file and so on. NFS etc. will support metadata with time, if
metadata becomes common.
Using a dot.directory is visual clutter to me. I use ls -a often enough,
and wouldn't like to see lots and lots of extra directories
(one per file with metadata?) showing up.
Helge Hafting
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/faq.html