Re: [PROPOSAL] Extended attributes for Posix security extensions

[Daniel Phillips]

Anton Altaparmakov wrote:
> I'm not Curtis and don't know about XFS but I will try to answer your
> question anyway. (-;
> If XFS doesn't order the EAs that's fine. - However for FSs which do order
> them the API has to provide the interface.

Does it?  I think you may be jumping to your conclusion.  Here are the
first three ways I thought of for imposing order without providing for
it in the API:

- Return a sequence number with the attribute
- Return a partial orderings 'a before b'.  (Such orderings are flexible
and efficient to evaluate)
- Use a flexible stucture, for example, XML, and return all the
attributes as a single string

My point here is that we may be mixing up low-level fetching/storing
with the more complex behaviour we're looking for.  Our goal seems to be
to come up with the final, be-all-end-all solution that handles the
semantics of all other operating systems - we're probably not going to
get anywhere close unless we start breaking this problem into orthogonal
pieces.

--
Daniel
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]