home *** CD-ROM | disk | FTP | other *** search
-
- On Mon, 12 Jul 1993, John Gardiner Myers wrote:
-
- > Mark Crispin <MRC@panda.com> writes:
- > > No, because the ambiguity would still exist at the IMAP level, and the whole
- > > purpose of the ugly syntax is to eliminate the ambiguity.
- >
- > Oh boy, it's the old ambiguous/unambiguous namespace argument again.
- > I mistakenly thought we had gotten this resolved.
-
- I believe the ambig/nonambig name issues *have* finally been resolved in
- the context of a single driver. I don't believe they have been resolved
- in the context of multiple drivers, as we are facing here.
-
- > Exporting an ambiguous/"rubber room" name does not prevent you from
- > also accepting an implementation-dependent unambiguous name, for
- > clients that care.
-
- I think I agree, though I'm not sure who is doing the exporting.
-
- > In the absense of a user who explicitly gives implementation-specific
- > information, clients and servers should communicate with
- > ambiguous/"rubber room" names. Otherwise, clients have to make
- > possibly invalid assumptions about the server implementation.
-
- Agreed.
-
- > If pine assumes that the IMAP server is a c-client imapd, it's likely
- > to get very confused when talking to a non-c-client IMAP server. The
- > CMU IMAP server, for instance, will tell any client trying to use a
- > mailbox or bboard name with a "/" in it to go jump in a lake.
-
- Pine makes no such assumption. In fact, we've worked hard to keep *any*
- notion of filesystem name syntax out of the Pine code. (OK, so there are
- a couple of exceptions for DOS, but not in c-client functions.) However, a
- Pine user can *configure* his/her environment to take advantage of an
- IMAPd that will make an effort to map a user-provided path name into the
- actual filesystem. This is especially handy (indeed, essential) in
- environments such as ours where an IMAP server may also be a timesharing
- host, and it is a requirement to be able to get at the same mailboxes both
- locally and remotely.
-
- -teg
-
-
-
-