Date: Sat, 25 Mar 2000 03:02:50 GMT
Reply-To: Theo DP <theodp@AOL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Theo DP <theodp@AOL.COM>
Organization: AOL http://www.aol.com
Subject: Re: SAS estimating how long it takes to write SAS programs
>Subject: SAS estimating how long it takes to write SAS programs
>From: SRhoades@US.IMSHEALTH.COM (Rhoades, Stephen)
>Date: 3/24/00 7:14 AM Central Standard Time
>Our department has an effort underway to improve quality in the software
>development area. We are revisiting our documentation on unit testing,
>system testing and QA (quality assurance). At a committee meeting to
>discuss this we were told that we have a mandate from management to insure
>all of the testing is done prior to project completion and that we could go
>over our immediate manager's head if we were told to skip some portion of
>the testing. There was not a lot of enthusiasm by most members of the
>committee, my personal opinion was challenging a manager like this is a
>lose-lose situation. The second problem is there are no guidelines on how
>long it should take to develop a program and typically completion dates are
>established without knowing how many programs need to be written or how long
>it will take just to write them.
>I am interested in how other shops estimate how long it should take to write
>a SAS program. In other languages the approach is to come up with a time
>estimate for an easy program, a medium one and a hard program along with a
>factor for skill level. SAS in particular is a difficult, as you can have a
>whole mix of Data and Procs steps within a single program. There are also a
>number of other factors that go into a project whether it be single report
>or a whole system. Do any of you have any guidelines that you would be
>willing to share?
The old 'easy', 'medium', and 'hard' program routine...
Makes me rember how long it's been since I've had a PHB...
BTW, the 'Dogbert Management Handbook' is not actually intended to be used to
run your organization...