[193] in Tooltime
Scopus Implementation Meeting Aug 13, 1996
daemon@ATHENA.MIT.EDU (Steven Wade Neiterman)
Wed Aug 14 11:35:34 1996
Date: Wed, 14 Aug 1996 11:39:10 -0400
To: tooltime@MIT.EDU
From: wade@MIT.EDU (Steven Wade Neiterman)
Scopus Implementation Meeting August 13, 1996
Attending: Gougen, Rosen, Saylor, Smyser, Wade
Agenda:
1. Roundtable and review of current activities
2. Demo of Data Probe and decision of use
3. Review of Netinstall comments
4. Next Release
5. Next Meeting
-----
1. Roundtable and review of current activities:
Wade
Notes:
- Machine status: Talked to Roger Roach (8/12), appears to be
problem getting Sun to deliver a Sparc 20 based on the configuration
in our order. At first, problem thought to be backorder of monitor,
now appears to be more serious. Looking at 2-3 weeks for delivery.
- Database: Need to reschedule install date. The database group
confirmed that we can take a "snapshot" (export/import) of the
development database and move it to production, which is preferable
to loading from scratch.
- Mail from Scopus on Aug 7th, engineers still looking at Mac issues
- Call from McGill University, interested in Scopus
Action Items not mentioned during the meeting:
- Database changes: Run Brett's script to capture 2k history log
information for searching in Filemaker legacy data, load software
and OS data, set mit_touched data to zero, enlarge mit_phrase in
mit_lk_abbrevs field, start with sequence number in 100,000 range.
- Update project notebook, work on release schedule
- Clean up help-them machine directory structure
- Write document for Facilities Management on Scopus install
- Work with database group on table sizes and preperation issues
- Continue to work with Facilities Management on SLA issues
-----
Brett
Notes:
- Completed Data Probe investigation, able to demo using MIT Logs
data and fields.
- Researched issues into GPF problem, found to be related to
mit_history_search field. Made change and tested on Solaris.
Need testing on Windows.
- Stored a snapshot of all Tcl code in AFS tooltime locker
- Waiting for Scopus feedback on InformTEAM problem report
- Brett will need to devote time to student arrival issues as part
of athena-rcc team. This begins last two weeks of August.
Action Items:
- Add Tcl code to allow search against mit_history_search field.
Change user interface to allow search from pull down menu. Need
to check with Lynne on changes.
-----
Lynne
Action Items:
- See changes in the netinstall section (item 3 of these notes).
Action Items (from 8/6):
- Reviewing programmers/business rule document from Esther
- Help Esther with training document
- Facilitate with Larry Egan (others?) to test netinstall
- Confirm platforms that can call Crystal reports directly from
within Scopus. [Note: Wade believes that Crystal reports
does not run on the Mac and Unix, which is an advantage for
Brio Query - if this was a requirement]
-------
Esther
Notes (from week 8/6):
- Working with Lynne on programming/business rule document
- Out week Aug 19, out next Wed.
[Note to Esther and those reviewing training issues - feedback notes
from release 1.0b2 and 1.0b3 identified items in a category termed
"training". Someone needs to revisit these notes for possible
incorporation into the document - wade]
Action Items:
- Completed: Send out rough first draft of training template.
- Get feedback from team on template, continue working on guide
(strive for simplicity).
Rob and Lynne have agreed to help.
-----
Peter Lee
Action Items (from 8/6):
- Completed: Send storybook to Esther for review
- Peter to check/test the netinstall Filemaker migration from 8/5.
- Peter will confirm account length needed for netinstall.
-----
Rob (and Peter Ragone)
Notes:
- Need meeting with Rob, Peter and Barbara to review Scopus
architecture, things to learn, explore next steps
Action Items:
- Test 1.0b4 changes prior to general release on 8/16.
- Help coordinate parallel 1.0b4 testing with two pairs of Help Desk
staff. Issues with reduced staff (students leaving this week).
Rob to check with Carol Wood. If preferable, we would like to use
a "buddy" system of two teams to do parallel testing on HelpDesk,
housecall and netinstall.
-----
Saylor
Action Items:
- John will have a student (August Chang) enter data into the
netinstall section to help with testing.
-----
Tooltime team
Action Items:
- Confirm GPF problem is fixed on Windows
- Review and comment on template from Esther
2. Demo of Data Probe and decision of use:
Brett gave an overview of Data Probe and a demo. There are still issues
with searching the mit_history_search field. Brett believes he can
resolve this in an afternoon. Other issues include support for Data
Probe and problems with searching not being case insensitive (although,
with some research, there might be a solution). Testing was done using
Solaris. It's possible that the Windows and Macintosh could result in
further issues.
Based on the demo and discussion, it was decided to not release Data
Probe. The real use of Data Probe would be to allow searches on fields
not on the Help Desk screen. Data Probe would give us more flexibility
in searching and filtering, but this has not been proven to be needed.
It was agreed to create a search on the log history information (first
2,000 characters) by generating an item in a pull down menu. This leaves
the Help Desk screen un-touched and allows searching (and backfill) of
a field deemed important for release 1.0. The advantage of this design
is the simplicity over releasing Data Probe and with greater flexibility
(allowing case insensitive searches).
3. Review of Netinstall comments
Reviewed mail from John Saylor (see discuss transaction [0190] from
August 12.
Made the following decisions:
- Add DCS category to call queue since this field has nothing that
matches complex installs.
- Remove required attribute on the hardware field since complex net
installs would not fill this in. If removing this field causes
data integrity issues, reports can be generated to review the
process of how this field is used.
- Would prefer to either move the position of the status field on
the netinstall screen or at least remove the required attribute. Not
sure the later can happen based on rules coded into Tcl.
General comments on the netinstall were related to the screen being
too complex (e.g. "how will the log text field be used", "many fields
will be left unused"). However, it was agreed that this was not the
right time to revisit major changes to this screen. There were concerns
that the screen looked too much like the Filemaker screens and did not
take into account other needs, such as multiple install drops, etc.
These comments will be saved and revisited in release 2.0.
4. Next Release:
1.0b4 release scheduled for Friday, August 16.
Need to confirm that:
- GPF's eliminated
- history search available
- netinstall changes agreed upon completed
- progress on database changes (see above)
- 1.0b4 tested by Rob/Peter prior to release
- Parallel testing or other testing available
5. Next Meeting and Proposed Agenda:
Friday August 22 (NOTE: No Tuesday meeting next week)
Purpose: Feedback and assessment of 1.0b4 release
Time and place to be announced. Location will probably be in 16-720
since this worked well for the feedback sessions to date.
