[222] in ad-lib
Re: LCN's
daemon@ATHENA.MIT.EDU (Grant Young )
Fri Mar 24 11:47:06 1995
To: Sarah Mitchell <smitchel@MIT.EDU>
Cc: ad-cat@MIT.EDU
Date: Fri, 24 Mar 95 11:43:24
From: gyoung@MIT.EDU (Grant Young )
If you were depending on the LCN's that resulted from the Circulation back
transfer, those are likely to change in the second GMA. If you're using
LCN's for locating existing MRMS records, that might work. I believe LCN's
are generated as the record is loaded. The GMA loads records in GRSN order
first and then Circ records are back transferred. If the Circ record finds
a match with the GRSN/MRMS record already loaded it gets added to the MRMS
record. If it doesn't, a new record with a new LCN is created. The first
LCN for a circ record is one higher than the last GRSN record. Since the
2nd GMA will have a bigger GRSN file to work from (regular loads and such
in the meantime since the 1st GMA), the LCN's for the Circ records will
start at a higher level which is unpredictable. The GRSN number is pretty
stable. So, too, is the LCN once it is assigned. But trying to guess that
the LCN will stay the same after the second GMA is not a sure or
guaranteed thing. That's why we loaded the GRSN number into the GMA record
so we could be sure we got a match and could compare the GLIS record.
-- Grant
