Long variable prob with break processing still

[John Esak]

	> > again. Declared variables do NOT carry the 10 copies of
> themselves that 2
> > character variables do. This is what is causing your problem. I
> do not need
> > to try the "sample" code. It is obvious from looking at it.


>
> Thanks for your explanation John.  I guess I missed that thread
> entirely as I always try to keep
> messages from the list which are pertinent and could not find
> anything in my archived messages.  I
> found the messages to which you refer in the list archives from
> Oct last year and yes it was all
> explained very clearly.
>
> The only thing missing is for fP Tech to explain and document it
> very clearly that filePro is
> "designed" in this way so more time is not wasted by developers

I could not agree more. As I told you in my first message, this bit me
pretty good too... twice as a matter of fact. Once about 7 or 8 years when I
first discoverd it... and once again about last October (thanks for dating
it... I couldn't find it) when I forgot about it and tried it again.  It
should be very clearly spelled out in the docs, and it isn't. :-(

> on issues such as these.  Or fix it
> of course!

There in lies the rub.  There are very good reasons for the declared
variables NOT to behave as the 2 character variables do with regard to
subtotalling. (Just as there would be good benefit to having them carry the
subtotals...)  When Ken suggested I put the question out for debate, the
response was pretty thing... except for maybe something from Nancy and maybe
a couple others.  The pros and cons were not very well laid out, and there
are some really far reaching ramifications to changing how the declared
variables work now... most of them bad.

John