Re. Estimating customization times

[Geoff Hart <geoff-h -at- MTL -dot- FERIC -dot- CA>]

     Patty Ewy asked about how to estimate customization times for
     (presumably documentation of) software products:

     Here's a dark secret from my past: If you think editors are "anal
     retentive" (why don't we just say "constipated" and save a word?), you
     should try research scientists. Having been one, and currently being
     the other, I'll suggest a research approach: if you don't have any
     good statistics already, go out and collect them. Do a trial run for a
     conversion. Pick the hardest part of a full conversion project
     (instead of doing the whole conversion), convert that part, and then
     use this to estimate the timerequirements for the easier parts. This
     will probably overestimate the time you'll actually require, but the
     approach has two advantages:

     1. Until you've done a few conversions, the process will be slower
     than you expect (particularly since you'll have to stitch the
     component parts together, and this will take additional time not
     accounted for in the quickie estimate I suggested). What starts out as
     an overestimate may prove an uncomfortably tight deadline the first
     few times around.

     2. If it really does turn out to be an overestimate, think how good
     you'll look when you beat all your deadlines! <grin> Remember the Star
     Trek/Next Generation episode in which Scotty joins the crew for an
     episode? Picard phones down from the bridge to ask Geordi to do
     something typically impossible and Geordi responds that he'll have it
     done in a day. Scotty, amazed, reproaches the younger engineer in that
     patented Scots accent: "Och, mon, have ye learned nothing from your
     years at Starfleet? You tell them it'll take two days, then when you
     do it in one, they think you're some kind of genius."

     Hope you end up looking like a genius!

     --Geoff Hart    #8^{)}
     geoff-h -at- mtl -dot- feric -dot- ca

     Disclaimer: If I didn't commit it in print in one of
     our reports, it don't represent FERIC's opinion.