home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!decwrl!infopiz!mccall!ipmdf-newsgate!list
- From: ned@sigurd.innosoft.com (Ned Freed)
- Newsgroups: vmsnet.mail.pmdf
- Subject: RE: vrfy and expn commands
- Message-ID: <01GO9XXMVP4Q9S3Q9U@SIGURD.INNOSOFT.COM>
- Date: 1 Sep 92 22:58:39 GMT
- Organization: The Internet
- Lines: 70
- Return-Path: <epmdf@YMIR.CLAREMONT.EDU>
- Resent-Date: 01 Sep 1992 15:58:39 -0700 (PDT)
- Resent-From: epmdf@YMIR.CLAREMONT.EDU
- Errors-To: epmdf@YMIR.CLAREMONT.EDU
- Resent-Message-ID: <01GO9XYFO8IA95OSJZ@YMIR.CLAREMONT.EDU>
- X-Vms-To: IN%"KLENSIN@INFOODS.MIT.EDU"
- X-Vms-Cc: IPMDF
- Mime-Version: 1.0
- Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
- Content-Transfer-Encoding: 7BIT
-
- John Klensin writes:
-
- > Given that RFC1123 provides for disabling these things for security
- > reasons, would it be sensible to have a per-list attribute that could be
- > set to, e.g., NOEXPN so that one could selectively protect lists? Seems
- > more straightforward than hiding them behind aliases, and it seems to me
- > that a clear "no, you don't get any information" message would have
- > advantages in at least some situations.
-
- In general list access can be controlled via the named parameter mechanisms in
- PMDF. This does not speak to the issue of handling expansion access differently
- than list use access, however.
-
- I have accordingly gone ahead and added (for the next release) a new
- [nonexpandable] attribute for lists. This can be used specifically to block
- expansion via the EXPN verb.
-
- Question: There seems to be only one status code EXPN can return when an
- expansion fails, and that's 550. It would be useful to be able to distinguish
- between "that's not a list" and "it's a list but you cannot expand it". Any
- thoughts?
-
- > This would also presumably disable MAILSERV SEND/LIST if applied to
- > one of those lists.
-
- The mechanisms are somewhat more complex than this and cannot really be
- unified. Think of it this way: A single list can have several "views". A view
- provides a particular means to access and/or modify the list. Access
- restrictions can be applied separately to each view. What we think of as the
- list name in the PMDF alias file is what I'm calling a "view" here for clarity.
-
- For example, it is perfectly permissible to have one list for freshman, one for
- sophomores, one for juniors, one for seniors, one for faculty, and one for
- staff. You could then build a bunch of different "views" of these lists:
- freshman, sophomores, juniors, seniors, underclasspersons, upperclasspersons,
- students, faculty, staff, employees, everybody, etc. The access restrictions
- are likely to be different depending on the "view" you're talking about.
-
- But MAILSERV would see these lists each as a separate entity; it does not have
- to deal with the views. In other words, if you add someone to the freshman
- list, the association of that list with the freshman, underclasspersons, and
- students views is assumed and should be automatic.
-
- MAILSERV also has to deal with write access and list locking (certain types of
- "send me the list" actions are supposed to lock the list until it has been
- updated and replaced). MAILSERV therefore has a bunch of things to worry
- about that a "view" never has to deal with.
-
- As a result the authorization mechanisms are split. I plan to release
- documentation about the latent authorization support in MAILSERV in the near
- future.
-
- It is unfortunate and confusing that we don't have the terminology in place
- to talk about the lists versus views. It makes things rather confusing. But
- the difference is nevertheless very real and quite often is very important
- in being able to implement maintainable lists.
-
- > As I think about this, it may be that the right list analogy to
- > VRFY isn't EXPN but some new command that would take the list name and
- > a mailbox address and return, like VRFY, an indication as to whether or
- > not that address was on the designated list. Slightly more invasive
- > than VRFY, but not nearly as much so as EXPN. Maybe someone should find
- > an SMTP Extensions WG and propose it to them :-).
-
- It is a very interesting idea. The good old "what form is the address in this
- time" problem is there, but after dealing with it in MAILSERV I must say that
- it can be managed. Implementing it in PMDF would be trivial too. If anyone else
- thinks this would be valuable I'd be happy to pursue it further.
-
- Ned
-