On Tue, 2003-01-28 at 22:44, Larry Osterman wrote:
> Actually, the implementations that treat "subscribed" as an aspect of a
> mailbox are broken.  The spec explicitly requires that you be able to
> subscribe to non-existant mailboxes, if "subscribed" is an attribute of
> the mailbox, then you can't do that :)

No, it doesn't require being able to subscribe into non-existing
folders. It requires that if subscribed folder is deleted, the
subscription isn't.

Reply via email to