Do the mail hubs store messages internally as
	MIME
	RFC 822
	X.400
	other

Are there facilities for maintaining a centralised naming scheme that
hides all details of where a user receives/sends their e-mail?
Attach description.

Can the centralised naming scheme handle
	partial name matches
	fuzzy name matches
	"role name" addresses
	returning a list of matches for ambiguous addresses

Can your Message Stores be engineered to still provide access to users'
mailboxes in the face of:
	Disk failures
	CPU failures
	"whole-of-building" failures

Can your mail hubs be engineered to still forward new and previously
enqueued messages in the face of:
	Disk failures
	CPU failures
	"whole-of-building" failures
NB. It is not acceptable to have to manually reconfigure each client
machine or message store to use a different mail hub.

Can the number of messages processed and still remaining to be processed
be monitored via
	SNMP (using MADMAN MIBS)
	Web clients

Is there a flexible and easy to use method for specifying the re-writing
of:
	e-mail addresses
	header lines
	message bodies
	attachments

Are message delivery messages returned in RFC 1891--1894 "NOTARY"
format?  If other, please specify.

Are the DSN ESMTP extensions supported?

Can filters be inserted to 
	convert Word for Windows attachments into plain text
	scan for viruses
	gateway e-mail to FAX modem
	gateway e-mail to pagers 
	other (please specify)


----------

19.9a   - directory lookups to find people

19.17a  Is it possible to remotely monitor the system resources that are
being used processing e-mail?

19.22a  Is it possible to return customised error messages for
particular message delivery problems?

Do the parts of the e-mail service that speak Internet protocols adhere
to to Internet standards, e.g. RFC 1123?

Is your scheme for Mail Hubs and Central Post Office functions scalable?
Please describe.

Can queued messages be categorised in order to apply different
processing strategies?  Please describe.

Are there firm limits on the amount of system resources that can be
consumed processing e-mail?

Are all messaging transactions logged?
- are there tools for searching or summarising the logs?

Are there tools for locating stuck messages and for taking corrective
actions?

Is there support for "group mailboxes"?

Describe how you would support a duplicate data centre at a separate
site?

If you are distributing the Central Post Office functions, please describe 
- how the config files and databases would be replicated
- how multiple mail logs would be managed

Do your message stores route mail to its final destination rather than
just uploading to a smart mail hub?

Do the message stores support strong authentication or encryption:
- Kerberos
- S/key or security calculator
- SSL
- data encryption