Re: cvs-versions do not match + bug -- LINUX5250




On Tue, 3 Jul 2001, Patrick Bielen wrote:

> Hello guys,
> 
> Seems that the cvs-version of cleveland and sourceforge do not match.
> After i compiled from the sourceforge-version my end-key does not
> work anymore, at the cleveland-version it just jumps to the end of the
> previous typed text. As i said before, it is very confusing to manage 2
> different cvs-versions at two different locations... expecially when they
> do not match each other.

We are NOT trying to maintain two different CVS repositories.   
CVS.CLEVELAND.LUG.NET is no longer being used by this project.  Jay Felice
has already stated that he's going to shut it down as soon as he can get
all of the projects off of it.

DONT USE IT.


> About the bug... i reported it a month ago already, and i do not know
> if it is fixed in the cleveland cvs version, but it is not fixed in the
> sourceforge-version, that's for sure.
> 
> The bug is, when you type for example 100 and the "-" -key it normally
> converts to 100} but it rips one zero to much, so it becomes 10} instead
> of 100} , it's not a large bug, but very anoying :-)

This isn't a bug.  This is the correct behavior, I've tested it
thoroughly.

The only thing that I can think of is that maybe you're using a different
type of field for input that I am?   Can you send me DDS source of a
screen where the number is given a value of -10, instead of
-100? 


Here's the logic behind it:
----------------------------
10} is -100.   That's how zoned decimal works.  The number is changed from
positive to negative by changing the zone of the last number from an 'F'
to a 'D'.

in EBCDIC, 100 is expressed as x'F1F0F0'  (if you have an EBCDIC table
handy, you'll see that x'F1' is the number 1, x'F0' is the number 0.
Therefore, x'F1F0F0' is 100.

To make it negative, you change the final x'F0' to be a x'D0'  This gives
you x'F1F0D0'.   If you look at an EBCDIC chart, x'D0' is the '}'
character.  Therefore, x'F1F0D0' looks like 10}.

If you added an extra zero, and made it '100}' it would be an EBCDIC value
of x'F1F0F0D0'.   You can make that a positive number by changing the
x'D0' to a x'F0'.   Therefore, '100}' when converted to a positive number
is x'F1F0F0F0' or 1000.   This means that '100}' is -1000.


Here's how I've checked it, double checked it, etc:
---------------------------------------------------
I've tested this by writing RPG programs that test the value of a field on
the screen where '10}' appears, and in fact it does register as -100.
I've tested this same behavior on Rumba, Client Access, MochaSoft and an
actual IBM 5250 terminal, and all behave this same way.



+---
| This is the LINUX5250 Mailing List!
| To submit a new message, send your mail to LINUX5250@midrange.com.
| To subscribe to this list send email to LINUX5250-SUB@midrange.com.
| To unsubscribe from this list send email to LINUX5250-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---