Re: [Gnu-arch-users] extended attributes

[Aaron Bentley]

Tom Lord wrote:

>    > From: Aaron Bentley <address@hidden>
>
>    > >Which points in the general direction of why extended attributes are
>    > >simply an ill-conceived idea.
>
>    > For myself, I'm not sure of this.   
>  
[snip]

>    > If editors were doing what they say they're doing (changing contents), 
>    > any extended attributes would naturally be preserved.
>
> Are you sure?  Extended attributes have no particular semantics.
> There's no rule to say when they should be copied and when not, when
> they should modified and when not.   They are an example of the
> "property lists solve _everything_!" class of design errors.
>  
I'm pretty sure that merely changing contents (e.g. fopen, fwrite, 
fclose) will preserve extented attributes.  I'm not sure whether they're 
a good idea or not.  I certainly liked the Reiser notion of making 
attributes just more files, but the identity vs equality issue is always 
a pain to deal with, and it may be impossible to use them if not 
everyone buys into it.

Ideally, I'd like to see Arch respect an "arch-category" attrib instead 
of having .arch-inventory files in my tree (or ++ in filenames), but 
this isn't an ideal world, and Arch seems designed to work well in the 
real world.

Aaron