Re: Basement tinkerers and inventors


Pete Dashwood <dashwood@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:5kbg35F2u8h5U1@xxxxxxxxxxxxxxxxxxxxx


I am now reticent about it and gave up long ago trying to persuade anyone
to
this approach. The example you gave, while I agree it is not contrived in
a
non-OO environment, is a "contrived" example for an OO solution. The
approach would be different if OO was in use. (There wouldn't be an
edit/file update, for a start... the decomposition levels would be lower,
with different methods of the same class communicating with each other and
the outside world. I tried to imply that in my response.)

Pete.
--

I didn't state, because it wasn't needed for the purposes of the question,
that there were daily, week-to-date, weekly, month-to-date, monthly,
year-to-date and yearly requirements for reporting. Nor did I state, for
the same reason, that a complete time record involved from one to six raw
records from the clock. But now you've flummoxed me. I don't see how you
can possibly NOT have an edit-file update function, unless the approach is
to retain all the raw data and reiterate the edit/assembly/resolution
function every time a report is run.

?????

PL


.