On Jul 9, 2005, at 4:03 PM, Bill Klein wrote:
<much snippage>
Concerning the COBOL compiler message:
"IGYPS0157-E A shift-out was found in column 50 without a matching
shift-in in a nonnumeric or national literal. The literal was
processed as
written."
The original problem has been "diagnosed" and hopefully solved.
However,
there has been a (semi)reasonable discussion on whether any/all COBOL
programmers would know what to do with this message. Tom Ross of
IBM has
already indicated that - unlike most compiler messages - this one
MIGHT need
additional information.
It seems to me that the "problem" (with the message) is that two very
different types of programs/programmers MIGHT receive this message.
A) Programmers creating Unicode or other multi-lingual applications
*AND*
who use DBCS characters within alphanumeric literamals.
B) Programmers who have either themselves or via a preprocessor/
translator
had data inserted into their source code.
I would segment group A into two groups -- those that are actually
using DBCS and those that are only wanting to leverage the UCS system.
Just doing the LRM example of converting EBCDIC to ASCII requires one
to specify the DBCS option to get the Display-Of and National-Of
functions to work. That is a compiler option that makes me want to
pull my hair out.
Why isn't there a UCS|NOUCS compiler option? It just seems strange
to have to turn on DBCS to get access to UCS...
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
