To: cease@MIT.EDU
Cc: ganderso@MIT.EDU
Subject: Notes from our 10/1/99 meeting
Date: Fri, 01 Oct 1999 16:45:01 -0400
From: "Naomi B. Schmidt" <nschmidt@MIT.EDU>


Attending were: Abby Fox, Bill  Cattey, Alexp Prengel, Bob Basch,
and Naomi Schmidt

1. IRIX Releases
   -------------

a.  IRIX 6.5 will have a long lifetime
b.  We're now at 6.5.3 and expect 6.5.7m to come out in February.
	m at the end means maintenance with bug fixes
	only.  f at the end means with new features.
c.  Why upgrade?  Bug fixes, mainly.  Before decided whether or
	not to upgrade, we need to weigh importance of bug
	fixes vs difficulty of doing the upgrade.
d.  The likely scenario is that we'll upgrade to 6.5.7m or 6.5.8m
	for Athena 8.4 and possibly freeze there.
e.  Another possibility is to use the 6.5.3m source code that we
	just received to fix 6.5.3 bugs rather than going to 
	6.5.7 if that's too burdensome

2.  FY2001 renewal time frame
    --------------------------
Mar, April - discussions w/in owls, acmg, itlt, with vendors
May - make a decision and place orders

This year factors in the discussion will be Linux Athena and
Pismere.  We could possibly stagger cluster replacements, doing
most over summer and some in IAP 2001.

3.  3partysw
    --------
a.  We should collect logs for about 1 month re use of BioSym and
MSI from license server logs.

b.  Vendors that Alex will write to are:

	- Renderman
	- Utah Raster Toolkit
	- Studio
	- Gauss
	- Macromodel
	
[Alex amends: I was intending Studio, Biosym/Molecular
Simulations, Gauss and Macromodel (Renderman only if no Sun version
exists, Utah Raster Toolkit only if we don't have access to source
code which I think we have).]

We can send email to AO's in Chemistry, Biology, ChemEng, and
Material Science, to say:

"Here are the faculty and courses that we know about that use
SGI-only 3partysw.  Who else in your department should be included
for interviews?"  Ask for name, course, # students

4.  Bill's questions to SGI - In addition to the questions he sent
us, he should also ask about their follow-on plans to the 300 MHz
O2 over the next 12 months.

5.  We went over Alex's list of questions for vendors and Alex's
interview w Oliver re impressions and impact on Athena OLC.

6.  Bob reported on his interview with Jonathon re Athena Server
Ops.  They wouldn't cry if SGI went away, but it's not a large
impact on their work.  Some issues are:

- Size of the OS and the amount of AFS space that it takes up

- The need to run an SGI license server for just one pkg

- Cost of supporting one more platform: times to propogate
releases, another platform to test on, mkserv issues

- SGI specific bugs

- Extra work in de-synching the release because of the amount of
time the SGI propogation takes

7.  Things still to do:

- Naomi interview  Brian
- Alex interview Dorrie
- Bill interview Heather Anne

We'll gather vendor data before speaking with faculty

Abby will grep through FL stuff to find other possible faculty to
interview

Naomi will craft a list of questions for faculty

Naomi will create a time line over the next 4 years
of the various possibilities of replacing or phasing out current
SGI's.
