Rick and Bonnie, Thank you for taking this on. I read the first quarter '99 report. I am very concerned as this is my power co. We do have actual temperatures (not wind chill) sometimes of minus 20 to 50 below zero! What didn't make sense to me was spending only one-half million more dollars and going from 17% remediated at the end of 1998 to 57% remediated as of May 7,1999. However, they said this was based on the number of systems. I have a feeling this is a case of doing the simplest systems first and leaving the more time consuming till last? Do you think I'm way off base here? Nancy

-- Anonymous, May 16, 1999

Answers

I have spent the past week or so modifying some programs for a change in communications protocol. (Nothing to do with Y2K.) Most of the detailed work had already been done: I was only incorporating a new subroutine into fifty or so programs.

First, I had to find which programs would be affected. Then, I studied the programs and their interactions, so I could understand what would need to be done. Then, I made a fairly quick pass through them all to determine what precisely would need to be changed in each.

I categorized each program in turn: "Easy. Easy. Not so easy. Easy. Not so easy. I'm not sure. Easy. Easy. I'm not sure. Easy. Not so easy...."

Which modifications did I do first? The "Easy" ones. Which did I do next? The "Not so easy" ones. And, you know which I left for last....

(Two "Not so easy" modifications took about as long to do as two dozen "easy" modifications. The four or five "I'm not sure" modifications took as long as all the rest combined.)

Am I the only one I know who operates that way? Hardly. Do I extrapolate from my own experience when I try to figure out how companies are doing their Y2K remediation? Not entirely. But is it a factor I consider? You bet.

If I may, see Percentages and Budgets: What Do They Really Tell Us?, especially the last section.

-- Anonymous, May 18, 1999

I think even non-programmers work this way, Lane. It's human nature to catagorize multiple tasks in such a manner, and it's also human nature to tackle the easy ones first. That's one of the reasons that percentage aggregation of project progress metrics can be somewhat misleading, and doesn't tell the whole story. A good project manager looks at more than "percent of tasks complete" as an indicator of true project progress. In a project (or sub-project) where tasks are not sequence dependent, again, it is human nature to put off the tough tasks until the bitter end. Ergo, the 80 /20 rule (the first 80% of any project takes 20% of the alloted time; the final 20% takes 80% of the time).

Also, it's interesting that you were discussing this approach with respect to communication protocol rewrites. See this earlier thread:

http://www.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=000p p5

-- Anonymous, May 18, 1999