That would work Wallace and in a new design would be easy to incorporate
but as a modification where it would require that dates be assigned
which they currently are not, this entails modifying the structure of
the shipto table and a script to retroactively fetch the dates from the
other tables to back fill from previous history oralternately the
queries could be restructured to fetch the dates from the other tables
at each instance.
The idea for a table format result requires no modification to the
tables or the queries, only the application of existing formatting code
to the result of existing queries. Not quite cut and paste but close.
Jeff Roberts
J.R. Electronics
604-241-1362
On 02/08/2012 01:39 PM, Wallace Roberts wrote:
On Wed, Feb 8, 2012 at 1:26 PM, Jeff Roberts<[email protected]> wrote:
...
From what I can see when you ask for shipto addresses it queries the shipto
table for all unique addresses that are associated with that customer from
the entire history. The trouble may be in determining what the criteria for
"unused" is. If you do it by date you could end up re-typing addresses just
because you haven't shipped to that location in a year, etc. You could then
add a date selection process somewhere to determine how far back to look for
addresses eg: 6 months, 1 year, all... but that adds extra clicks to the
process that may not be desirable for others.
Why not have the system simply select the next N most recently used
shipto addresses for that customer, where N is a configurable
parameter with a reasonable default?
_______________________________________________
SQL-Ledger mailing list
[email protected]
http://lists.ledger123.com/mailman/listinfo/sql-ledger
