Problem with sub-total breaks on a report

[Kenneth Brody]

Boaz Bezborodko wrote:
[...]
>>> In @wbrk2 I have
>>> ? If:
>>> Then: ch(10,,2)=cg*ca/"100" ; cl=cl+ch
>>> The very first line of code checks if it is on a new sales rep and, 
>>> if so, sets cl(12,,2,g)=""
[...]
> When going through the data before hitting any of the breaks cl="".  
> Once the first break point is hit it should equal '0' until given the 
> value of ch which in my sample data is "62.32".  But it only shows this 
> when in the @wbrk routine.  Once it starts going through the data lines 
> again cl shows up as '/d0'.  When it gets to @wbrk2 again it shows as 
> "62.32" until a new ch gets added, in this case a value of '248.73' and 
> 'cl' then correctly shows a value of '311.05'.  At that point it hits 
> @wbrk3 and the value shows '396.49' for no particular reason that I can 
> find.  There is no '396.49' anywhere in the data set I'm working with.

The value of CL is based on CH, which is based on CG and CA, which are based 
on ???...

Without actual examples of all the data involved, we can only guess.  And my 
guess is that you are not doing what you think you are doing with regard to 
dummy fields and multiple break levels.

> I did find that I could get 'cl' to hold the right data if I put it in 
> AUTO processing.  But I'm trying to find out what FP is doing otherwise.

And that further points to my comment above about dummy fields and multiple 
break levels.  You may want to check out "Output Processing: subtotals, 
dummy fields, aliases and arrays", published in SmallTalk 9.3 (summer '92).

http://www.hvcomputer.com/download/output.txt

-- 
Kenneth Brody