Sorry for the first message. I did not include the revised recommendations. Debi ________________________________________________________ From: Allison Debra on Mon, Jun 27, 1994 10:19 AM Subject: Revised RFQ To: Miami's Electronic Mail Cc: MCIS-Administration You may recall that we sent out a Request for Comments on the committee's recommendations for Miami's email infrastructure and deliverables, on March 30. We asked for your comments by April 8. We received comments/suggestions from one person (Barbara Edwards) and her comments are included at the appropriate points below. Please email me with your comments by Wednesday, July 6. If it appears that we need another meeting to come up with our final recommendations, we will schedule one. I apologize for not getting this out to you sooner. I was consumed by a project to replace Wang equipment for 13 offices; those requisitions are now out, so I can get back to other projects. (48 Macs and 40 IBMs were ordered. The only offices still using Wang equipment that we are aware of are Student Financial Aid and Admission, both of which will have their equipment replaced prior to or at the time of the move to CAB.) Getting the Wang equipment replaced does impact our email project, since it means 13 more offices (3 academic departments; 10 administrative departments) will have broader network service coverage. As another FYI, the Wang replacement project also involved upgrading some central services for purposes of mainframe printing and uploads/downloads -- services that were available from the Wang OIS/VS computers, but not available from a network connection. We have placed on order TCP/IP for MVS, which will allow telnetting directly to MVS. To provide the necessary mainframe printing services, we have on order the Novell SAA Gateway, which will allow MVS and VM to print to network-connected printers which are defined as Novell print queues. I will be sending out an announcement to the MUDIVREP list, since their input was incorporated into these plans. Hope to see you at the 1:00 email audioteleconference today in 360 Gaskill. ----START HERE---- This version of the Email project recommendations includes comments and recommendations added by Barbara Edwards. It is being circulated for your review and comments. Where she makes a recommendation for an addition, it will be added unless there are objections expressed in response to me. Where she asks a question, please email to me your suggestions. Next steps: We need to develop a task list of items to be accomplished, with responsible person listed and timeframes listed. Email Project - Recommendations March 28, 1994 April 13, 1994: Barbara Edwards comments added Subgroup members: Debra Allison, Kent Covert, Barb Edwards, John Harlan, Peter Murray, Rob Pickering A. Recommendations - Qi server/Ph client; Mail Redirection 1. Qi and In-bound Mail Redirection: up in test this summer; in production before Fall 1994 on Rose and/or VMS (Technical Services to decide, in consultation with Client Services (John Harlan) and Instructional Computing (Barb Edwards)). 2. University Directory Survey: (a) ask for preferred email address on form requesting directory updates; (b) include Library question [what does this mean?] or statement on email notices. John Harlan to contact Publications to get this included. Barb's comments: Have we requested that a system be built to maintain the Qi directory? a. Is there a process being defined for taking the information off the directory update forms and updating the Qi server? b. Are we defining a process for coordination between updates of hard copy and availability of softcopy updates? c. Are we defining procedures for coordination of the installation of undergraduate accounts and the updating of the Qi server. d. Will the e-mail addressed be part of the paper directory? e. What is the procedure for documentation and updating of such? f. What about "unlisted numbers"? 3. Kerberos: Investigate using Kerberos authentication as university-wide security scheme. Goal: a university-wide security scheme in place by Summer 1995 (we recommend Kerberos). Steve Moore to look at Kerberos on VM. Barb's concern: There are several issues with regards security left over from old discussions. Has John Andrea been involved? 4. Passwords in Qi database (to allow modifications to data elements in Qi database): we recommend that the database be (a) modifiable and (b) authenticated (to be determined by Client Services, Technical Services, Security, in consultation with Internal Auditing). This is a long-term recommendation, tied to Kerberos implmentation. 5. Queries to the Qi databaseto be provided via: a. Gopher and Web (Fall 1994) b. Ph client (Fall 1994 in test, but unsupported - at Level 3+) c. Mail packages - ASAP; tied to selection of mail packages. Barb's concern: Will all this be added to enduser on VM? Will there be DCL and Scripts written for VMS and UNIX for the clients to use? Is technical services exploring this issue? 6. Alias structure in Qi data base recommended to be: a. unique b. eight characters c. same as university-wide account ID (Tim and Barb have different recommendations on this -- Tim: 1st 6 characters of last name, first initial, middle initial (to permit sorting by last name); Barb: first initial, middle initial, 1st 6 characters of last name (to correspond to exisiting IDs). We recommend that the Qi alias be: first initial, middle initial, 1st 6 characters of last name, except in the case of a duplicate where a digit is needed to distinguish, e.g. RAPICKER RAPICKE1 RAPICK10 RAPIC100 What will "from" address show to the world? Does it have to be the alias, or do we want it to be? John Harlan to check with Notre Dame, since their "from" addresses are "friendlier." Rob to check if Qi and Kerberos ID name limitations are 8 characters; he will also try to find out about X.500. We recommend Mail.Name@domain for use by sender when email address is unknown. Example: [log in to unmask], [log in to unmask], [log in to unmask] will all cause the message to be sent to Peter E. Murray's preferred email address. If P.Murray is used and there is more than one match, information on all matches will be returned to the sender as bounced mail and sender will choose based upon info in Qi database. Barb's comments: Please add as a recommendation: There will be a record in the directory entry that identifies the real person by full name and Miami Id number (currently social security number is used). This comment should be consistent with the personnel record or the student master file record. For ids not associated with a single person there will be a person identified for that mail box and that person will be recorded in the comment card. This comment and the userid should be in a file or database that is regularly updated and easily accessed by system management. We recommend that the lowest character count be used as mail identifiers. ---end Barb's comments for #6--- 7. Long-term direction: Our stated long-term direction is to replace the Qi server with an X.500 server. Qi evolves nicely into X.500 (Ph client can query X.500 server; Qi server database can be ported to X.500). Barb's comment for #7: Please add to the recommendation: X.500 is a stated goal and will be re-evaluated in 3 years as the technology evolves. IMAP/POP Recommendations 1. Recommendation: IMAP as a replacement for Microsoft Mail and Pegasus Mail,with an alternative (POP) where desirable or required. Barb's comments for #1: Concern: This recommends IMAP as a replacement for Microsoft Mail and Pegasus Mail. What about Open-VMS and VM mail? Currently there are significant numbers of mail users on VM and VMS. Please be aware- CPPO has requested VM accounts for at least sophmores and juniors and maybe first year students by January, 1995 and will continue their request for all seniors for August, 1994. ---end comment--- Timeframe: IMAP - test group of 100 faculty/staff/students July-Dec. 1994. Hope to have Qi server available at start of test. Test participants require: a. Reliable (added by Barb) Network connection to desktop, or at a lab for students (LTCs or divisional labs) b. IMAP/POP-accessible account on one host (Technical Services to decide whether ROSE or OCEAN, in consultation with Client Services (John Harlan) and Instructional Computing (Barb Edwards)). c. Start with core test group and ask them who they email (workgroup commonality); then include them. Barb's comment: Start with a small test group that is computer as well as e-mail literate. Expand this test to workgroups of the initial group. d. We want a variety of clients (MS Mail, PMail, new users (Bill Slover?); high-volume users). e. IMAP/POP software (1/2 using public domain client; 1/2 using commerical client). Debi and Kent to solicit vendors for evaluation/test copies. f. Way to re-route mail to new address for MS Mail, etc. since test participants won't want to tell the email world that they have a new address. We need: a. Solicit eval client software from vendors b. Ask for volunteers. Do they want IMAP or POP client? c. Documentation for test participants on all platforms Barb's comments: Criteria for IMAP or POP selection Software installation procedures Directions on how to use software Directions on how to update records in central databases d. Help Desk involvement in the test (preferably including interaction by e-mail via Target->Hotline, so that we can test the software with Hotline) e. Involvement from Instructional Computing, Networking, Technical Services, Client Services (including Help Desk). Barb's comments: Please add these recommendations: f. Procedures for installing mail boxes for users. g. Licensing and distribution issues to be part of the evaluation process h. Training and documentation for support staff in areas where tests will be conducted. ---end comment--- C. Critical Success Factors - How will we know we've succeeded, and what factors are most likely to determine whether we're successful? 1. Client Services says installation/maintenance concerns are easier with this set of solutions. Barb's comments: Please add these Factors. a. Ease of use to include How quickly the software installed Clients ability to update records Ability to determine another persons address Ability to inform someone else of users address ---end of comment 2. Mail gets delivered reliably (probably the #1 CSF) 3. Survey of client satisfaction. a. Ease of use? b. perception of reliability c. perception of usefulness d. comparison with old sysem e. would client continue to use/recommend this solution? 4. During test, check with clients. Are they still using the new solution, or did they revert to their previous product? 5. System benchmarks - performance monitoring, server sizing. 6. Total cost/scalable cost comparison (added functionality, freed resources?) Barb's comments: 6. Please add this to the evaluation: (... ,installation of volumes of mail boxes, distribution of software and collection of fees)