[Gnu-arch-users] Re: extended attributes

[Tom Lord]

    > From: Miles Bader <address@hidden>

    > Aaron Bentley <address@hidden> writes:
    > > While the link-breaking behavior is useful to Arch, I'm tempted to
    > > call it a bug.  My expectation of an editor is that it doesn't change
    > > the properties of the file (e.g ctime, permissions, # of links) -- it
    > > just changes its contents (and atime/mtime).

    > This `bug' is so ingrained into common tools, 

Basically because it has a sound basis in physical reality.

Classical unix filesystem semantics manage to surface a pretty
abstractly clean model for a physically constrained process
(filesystem implementation on a very broad class of tertiary storage
devices).  The next step up, to a fully transactional file system, is
a _huge_ step.

It's not very surprising -- the semantics come out of an
implementation process which amounts to "Ok, like multics, except with
a relentless focus on `what is that we can quickly (and actually)
implement?'."


Yes, it's vague -- but it's also true.
-t