On Sat, 14.05.11 23:51, Christoph Anton Mitterer ([email protected]) wrote:
> Hi. > > This is not directly file HIERARCHY related, but it might be of interest > for us. > > Many filesystems support XATTRs,... but currently there is not place (at > least none that I'd know) where the namespace is managed. On all real-life fs I am aware of reading xattrs are dead slow which is the primary reasons they are not taken in consideration by gvfs and thelike. As long as this is not fixed (could be fixed relatively easily I think, for example, by including a flag in stat() calls which informs the user whether there are any xattrs on a file, and hence worth reading them) I think there is little point in standardizing them. We don't want people to use them at this point in time. > 1) user. hiarchy: > For general use, a reverse-DNS namespace must be added after "user." and > before any (optional) attributes. > > 2) oid. hierarchy: > Analogous to (1), but using OIDs as namspace, e.g. > "1.3.6.1.4.1.32806.my_personal_attribute" oid? Really? Lennart -- Lennart Poettering - Red Hat, Inc. _______________________________________________ fhs-discuss mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/fhs-discuss
