[
Lists Home |
Date Index |
Thread Index
]
Well, what I'd hope (probably fruitlessly) to hear was
1.that the denormalization had been done for performance, and could
likely go away after some benchmarking, and
2. That the names had made sense on an even earlier version of the
product and had been retained to comfort (now long gone) old timers.
Thus the names too could be fixed.
I'd settle for hearing that "it was very late that night." or "We
thought it made sense but found out too late that we were wrong." At
least then I could do what I wanted with a safe feeling.
If I don't find anything, well, the chance that cleanup will become a
can of worms is a budget and schedule risk factor. And of course even
with a solid answer budget and schedule.
Frank
Bullard, Claude L (Len) wrote:
>That is certainly a good first step. Given the system,
>that information may or may not exist, may be contradictory,
>and has to be factored into other intelligence (eg, conversion
>costs and opportunities).
>
>What would you be looking for?
>
>
|
