[23] in Tooltime
Scopus mtg 2/14
daemon@ATHENA.MIT.EDU (Barbara Goguen)
Wed Feb 14 16:25:01 1996
Date: Wed, 14 Feb 1996 16:24:02 -0500
To: tooltime@MIT.EDU, cec@MIT.EDU, paheff@MIT.EDU, bdrosen@MIT.EDU
From: goguen@MIT.EDU (Barbara Goguen)
Attending: Bill Papp and Dick Campbell from Scopus; Kim, Carla, Al, Paul
Heffernan, Lynne, and myself from MIT.
Overall, I think the meeting went reasonably well - we've given them plenty
of homework. Following is a list of issues and responses we discussed. If
I've missed anything, hopefully someone else who was present will add it
in.
Issues:
- Installs of client software have not gone well at all; though Bill has
been responsive, he has not had the expertise to help that much. Mac
client is still installed on only one machine, PowerMac 7100, and still
running fairly slowly.
- Poor documentation for installation. No Mac install documentation to
speak of. Only docs are on Mac install diskette, and simply say, drag
Scopus folder onto hard drive.
Bill has brought this issue of bad documentation back to Scopus. Doesn't
sound as though we'll see any improvement in the timeframe we are working
on.
- Worries about Mac support in general. Is Taggert (in California) the
only Mac expertise they have?
Taggert is the head of the Mac development group. All of their Mac guys
are in California as are all of their technical folks for the most part.
- Memory requirements for Mac (and maybe other platforms?) seem extremely
high. Latest word I got from Bill was a recommendation to set memory
allocation to 8MB for Scopus. Many of our Macs only have 16MB total at
this point.
Bill's response was to try the 6mb setting rather than 8. Even at 6mb,
this is a steep requirement. Once again we mentioned that it would be
extremely useful to have a recommended configuration from them for running
the client software - System software version #, installed memory, chip
speed, etc. Also, they keep thinking that some of our install problems
have to do with an init conflict, but they cannot provide us with a list of
known conflicters.
- What about Scopus working with Open Transport?
Bill's question was, "is this a Mac thing"? It's fairly clear that we
need to arrange another conference call with a group of our Mac people and
maybe Taggert in California to figure out what direction they are heading
with their Mac client.
- Concerned about lack of knowledge of running Scopus on Oracle rather than
Sybase.
They didn't seem to have any response to this.
- Concerned that Dan may or may not be the right person to help with olc
implementation.
They both said that we should begin the olc conversation with Joan
Beardsley rather than with Dan Courtney. This is a different story than
they originally told us. Joan is the Director of Professional Services.
They also said that we should have detailed design spec prior to initial
meeting with Joan so that she can do "gap analysis".
- Concerned that I/we need to deal with so many different folks at Scopus
to get answers to our questions rather than having a single or small number
of contacts. Dan was steering us toward Tom Keller in California for Web
info; Ken Erikson in Boston for TCP/IP info. It would be nice to know
that someone at their end would do some of this legwork.
I actually didn't raise this concern, but it continues to grow. Today
they steered us toward Joan Beardsley, Taggert, Dave Gordon/Brenda
O'Malley, and Brenda doesn't even work for them any more - she works for
Charles River Data Systems!
- We need to get a clear definition of concurrent use for client licenses.
Once again, they stressed that there is nothing built into their software
that will cut us off at a certain number - no FlexLM or other such license
server. Their contract apparently includes language that allows them to
audit our installation once a year. But they confirmed what Paul heard
from the President of Scopus that concurrency means everyone running the
software simultaneously on their machines whether or not they are actively
answering questions or querying the database. This means we probably will
need closer to the 125 number than the 50 client licenses. We asked if
maybe we could draft what we think concurrent might mean in our environment
given part-time student staffing, and local expert usage - they agreed that
this might help.
- Pricing still way out of line with what we need it to be.
Dick really doesn't get it on this issue. He wants to treat us like
their standard customers who include software vendors, hardware vendors and
Banking/Finance folks. They only offer tiered pricing on client software
after 100 licenses. Dick feels that their 40% education discount easily
replaces any tiered or volume pricing. We told him loudly and clearly that
given their definition of concurrency, we would likely be looking at the
125 client license rather than 50, and that this puts their pricing out of
our reach at over 300K. In fact, we tried to get them to consider a target
much closer to 125K to include initial product purchase and 1st year
maintenance. Dick instead wants to come back with a site license quote so
that we wouldn't even have the 125 concurrent client limit, and he promises
that the site license pricing will be as aggressive as they can make it.
Other pricing nit-picks: Dick agreed to a consulting pricing of $1250/day
rather than their $185/hour rate; Server license does not limit us to a
single server, but assumes at least two servers 1 development, 1
production; they know we think their Web server pricing is goofy not being
a flat fee; and they heard us say that the on-going maintenance price is
going to have to be better than 15% of list.
Questions:
- Which Unix versions does Scopus run on? Currently we have it running on
Solaris, but we also have deployed quite a few SGI, Dec Ultrix, and IBM AIX
machines in our environment.
Solaris, SunOS, HPUX, RS6000 AIX, SGI will be coming out in March. As of
now, no Digital support.
Follow-up for response:
- Need technical contact at Scopus for technical folks at our end to have
conversation regarding SQL*Net, SNS, and use of Kerberos driver for
authentication. Need contact name immediately.
Dick thought they should be able to arrange a conference call with the
right people at their end as early as later this week or early next week.
Kim will take the lead at this end for coordinating that call.
- Need references, preferably from educational or government/non-profit sites.
Dick asked for a couple of days to get these references to me. He
mentioned Taligent as a definite since they are a large Mac site. They
currently don't have any education sites or anything particulary close to
what we are trying to do.
- Need to gain better understanding of Web interface. Need to be set up
with accounts to use it at their corporate site; Need to get it installed
here for testing/evaluation purposes.
They heard this again. Let's see if they do anything about it.
- Would really like to get our hands on "zephy"-like notification tool that
someone by the name of Brenda demoed here last summer.
They began by pointing us to Brenda O'Malley who no longer works for
them. Then they mentioned Dave Gordon as also knowing about how this was
done. They apparently used a subset of their TeleTeam software to do a
screen pop using a TCP/IP daemon. Apparently this may have additional
licensing/cost issues as we may need to license some or all of TeleTeam to
get this functionality.
Action Items:
Dick Campbell
-------------
Get technical contacts at Scopus and arrange conference call to discuss
security issues.
Get us yet another price quote that comes significantly closer to the $125K
range we are aiming for, with room for 125 client licenses.
Get Taggert clued into the idea of a conference call to discuss Mac
direction and issues.
Get reference names to us.
Get us set up to explore their Web interface.
Bill Papp
---------
Let us know if he happens across any other ideas regarding Mac
installations in general.
Kim
---
Draft agenda and/or list of issues for security conference call and email
them to Dick.
Coordinate the right technical folks at our end for security conference call.
Al and I
--------
Draft agenda and/or list of issues for Mac conference call and email them
to Dick along with some suggested times for this call to take place.
Anyone interested
-----------------
Continue to try to get Mac client to install consistently on different Mac
configurations.
Draft ideas of definition of concurrency that might work in our environment
and send them to me. I'll collect and try to make cohesive.
Carla and I
-----------
Complete draft of olc requirements document in preparation for discussion
with Joan Beardsley.
-barbara
goguen@mit.edu
617-253-6135
