Speaking unofficially as vice-chair of X12C, which is responsible for
the 997, I will confirm the comments that Jonathan Allen (chair of
X12C/TG4) made concerning the 997. Due to strong opposition from a
small number of voting members, we failed to achieve the consensus
required to modify the 997 to report against implementation guides. The
997 is strictly defined to report against the standard. If I remember
correctly (and I'm sure Jonathan will correct me), the 824 is in the
process of being modified to report against ICs. I will also confirm
that there is no case where two 997s are valid. If you need an official
statement on either point, you may file a request for interpretation
with DISA, X12's secretariat, and we would be happy to prepare you one.
Taking my standards hat off and putting my consultant hat on, I will say
that despite what the standard says, all of the translators I've worked
with except one actually validate against several aspects of an IC and
not strictly against the standard. For example, in GENTRAN Director and
NT, you need to explicitly enable segments that you are expecting to
receive. If you receive them and they aren't enabled in the map, a 997
reject will be generated. Other translators also give you the option of
restricting code sets used for validation.
Cheers,
--
Michael C. Rawlins, Rawlins EC Consulting
www.rawlinsecconsulting.com
=======================================================================
To contact the list owner: mailto:[EMAIL PROTECTED]
Archives at http://www.mail-archive.com/edi-l%40listserv.ucop.edu/
