Quoted text is from <3D808EC801AED111B0100008C75D5DDC0722E5AD@roc05bxgei
sge.is.ge.com>, by Bayes, Charles (GXS) <[EMAIL PROTECTED]>
>be in your data. The second "key character" is the 105th character of the
>ISA record. This is the sub-element separator. Although sub-elements are
>not used as extensively in X12 as they are in EDIFACT, there are times when
>they can be used, so this value is always defined. Again, this character
>CAN NOT be used anywhere in your data. The third "key character" is the
>106th character. This is your segment terminator, and is typically a
Jonathan Allen has pointed out in this list that it is theoretically
legal to send a negative interchange control number, such that the field
length is 10 characters. Personally I suspect that the probability of a
negative value is itself also negative, but purists might want to look
at the 100th character in ISA, and if not the same as the 4th, use the
106th and 107th characters for sub-element separator and segment
terminator.
I also suspect that a negative interchange control number would break
most of the existing EDI converters, and generate such hostility that
the sender would not repeat it in a hurry.
Regards
Chris
--
Chris Johnson +44 (0)20 8501 1490 (home)
EDIMatrix Ltd +44 (0)20 8559 2454 (work)
+44 (0)20 8559 2497 (fax)
http://www.edimatrix.co.uk
=======================================================================
To contact the list owner: mailto:[EMAIL PROTECTED]
Archives at http://www.mail-archive.com/edi-l%40listserv.ucop.edu/
