[2733] in Release_7.7_team
An example of foreshortened scope of AUI
daemon@ATHENA.MIT.EDU (Bill Cattey)
Mon May 7 18:45:08 2001
Message-ID: <8uxmJl5z0001A85cAv@mit.edu>
Date: Mon, 7 May 2001 22:45:05 +0000 ()
From: Bill Cattey <wdc@MIT.EDU>
To: rferrara@MIT.EDU, vkumar@MIT.EDU, longpd@MIT.EDU
CC: release-team@MIT.EDU, ganderso@MIT.EDU, rar@MIT.EDU
Bob, Vijay, Phil:
Last tuesday, when members of my team and jweiss's team discussed issues
of resourcing with Vijay, one of the examples given, by Greg Hudson, I
believe, was that the scope of AUI had to be cut back. A comment to
some code being reviewed reminded me of an important cost we need to be
aware of:
Originally the AUI project discussed a rewrite of the xlogin program
(which is the thing that people speak to when they "Press any key to
start..."). We didn't have time or personnel to do so.
When Greg Hudson added code to xlogin, the side comment he made to
source-reviewers deserves wider circulation, I think, in order to help
understanding:
Excerpts from mail: 5-May-101 xlogin option for dash-base.. Greg
Hudson@MIT.EDU (16798)
> ...This patch adds a new xlogin option to select the old dash-based interface.
> (I had no idea xlogin sucked so much. Well, maybe a glimmering. I hope
> we don't have to add any more login options before it's rewritten or
> otherwise disposed of.)
Yes, we rewrite code for the sheer joy of coding, but sometimes we do it
because we anticipate it will cut effort to make changes further down
the line.
Vijay is right to ask which is more important, projects to clean up the
old or to projects to deploy the new. I am advocating the position that
we need to strike a balance, and that the balance right now is tilted a
little too far away from taking care of the old. We want to be able to
take pride in our work -- both the old and the new.
-wdc
