[353] in magellan
Braindump from today's magellan meeting
daemon@ATHENA.MIT.EDU (Tim McGovern)
Thu Dec 16 16:40:06 1999
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Message-Id: <v04020a05b47f086f3ff5@[18.152.1.21]>
Date: Thu, 16 Dec 1999 16:39:47 -0500
To: Greg Anderson <ganderso@MIT.EDU>
From: Tim McGovern <tjm@MIT.EDU>
Cc: magellan@MIT.EDU
Greg --
We had a small, but spirited, group on hand today, and we tackled the three
topics more or less all at once, and it proved to be quite a good
discussion.
I humbly ask my colleagues who were present to chime in with additional
thoughts, ones I may have missed, and corrections to what I have here....
All errors in this regard remain mine.
* there was a general feeling that the Cutter Consortium categories just
didn't feel right for Discovery projects
* there was sentiment that the categories made more sense at/after the
actual deployment of some new idea/thing. Later, we discussed whether this
might make the Cutter metrics better adapted to the initiatives of the OP,
rather than to specific projects
* we tossed around some ideas about checklist / metrics for Discovery:
level of customer input, check-ins with stakeholders, rating by Delivery
teams, how close we came to estimated effort
* the question arose as to whether getting good measures always means that
the project was a success...hmmmmm.
* there was a feeling that different projects need different measures
* the suggestion was made that a few simple metrics -- on-time, on-budget,
etc. -- might be all that's needed, assuming we actually do a serious job
of estimating those in advance
* given that Discovery proejcts spend a good deal of time in the
pre-planning space, a suggestion was made to measure Discovery project
success by asking:
-- did we have the right team?
-- did we identify the right question(s)?
-- did we have a successful integration checkin?
-- did we identify the right deliverables?
-- and there would be more...
* we segued from there into talking about Discovery reports at about this
point, and there was the notion that the templates are a bit constraining,
and that they shouldn't be used as forms to be filled in. Different
projects (with different kinds of questions) often need different kinds of
reports.
* the notion of a "successful integration checkin" led us to discuss those
events in some detail. The general feeling was that they need to become
better structured, more facilitated discussions, in a more hospitable place
(Not E40-302). Another thought that came up were that there should be a
very formal role for the ITIT at the project announcement point, and that
should aim to head off projects that might get creamed when they get to the
checkin.
* finally, there was a general belief that there should be a more explicit
statement of Jeff's role as the "senior systems architect" and what that
means, so that newbie project leaders understand why it's so critical to
consider Jeff in almost all projects as a key stakeholder.
