Re: [Re: RPGILE V4.3 Gotcha] -- RPG400-L

What's to understand?  From all the discussion I have read here, and
from someone quoting the manual, if you use eval you can LOOSE all
your decimal precision by doing combined mathematical functions.

And in regards to "any programmer use any programming language without
being intimately familiar with the corresponding documentation".

1.  I do combined mathematical formulas in LOTS of different languages,
and RPG is the only one that has this quirk.  After reading up on mathematical
formulas, how numbers are store, what a mantissa is, how computers perform
math, the difference between BCD and floating point, time evaluations on
doing a clear .vs. a move et. al.

I have read this for enough languages (FORTRAN, COBOL, Pascal, Basic,
C, etc..) so that I understand it.

Now, behind me is a book case, it is 4 feet high by 3 feet wide.  Inside of this
bookcase are 42 note books, the majority containing one book, quite a few
containing 2 or more books.  Out of these I have read maybe 10 cover to cover,
such as the CL Reference, DDS reference, Operations Guide, and Command
Reference, RPG 400 guides, et al.  Others I have skimmed thought, some I
use for reference and only open when the need arises.  Some of them I will
most likely never open.

IBM has decided that the printed manual is no longer the norm, that now books
are going to be in electronic format.  And,  much of these are scattered all
over the place, Softcopy Library, Rebook Softcopy Library, Books on Line,
PDF, and who all knows where else, I sure don't.

And, though all this mass of material, I am supposed to be able to know to
read one particular book that explains in some particular paragraph that IBM
has decided that the way all other compilers do math is wrong, that IBM has
the right way, although it will, in effect, produce less accuracy over the long
run.

Now that I know about it, I can work around it.  But a couple of points here.

1.  I shouldn't have to work around it.  Nor should others.
2.  What about all the other programmers that don't know about it?
3.  People don't read the manual to do something as simple as multiplying A 
times

     B and dividing by C.
4.  This "feature" is going to produce more bugs that programmers won't
     be able to find without careful examination of how IBM has decided to do
     their math routines.  The behavior it is trying to fix (Overflows) is so
     minute in comparison to lost accuracy.
5.  And who says I don't understand the rules?  Is there any behavior I stated
     in my posts that wasn't correct?  Would it not in fact loose accuracy as I
     stated?  If I am in error, let me know where.  If I am not in error, don't
say
     I don't understand it.

Regards,

Jim Langston

boldt@ca.ibm.com wrote:

> <SNIP>

>
> Oh dear oh dear oh dear!
>
> I hope this doesn't sound rude, but how can any programmer use any
> programming language without being intimately familiar with the
> corresponding documentation?
>
> Sure, we could have come up with better rules for decimal precision.
> But please don't criticize the existing rules while admitting you
> don't even understand them.
>
> Cheers!  Hans
>
> Hans Boldt, ILE RPG Development, IBM Toronto Lab, boldt@ca.ibm.com
>
> +---
> | This is the RPG/400 Mailing List!
> | To submit a new message, send your mail to RPG400-L@midrange.com.
> | To subscribe to this list send email to RPG400-L-SUB@midrange.com.
> | To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
> | Questions should be directed to the list owner/operator: david@midrange.com
> +---

+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---