1 - Yes you could probably replace the data with the file upload. iSeries
developers don't often recommend that for a number of reasons
a - The data on the i could have had records added, deleted or modified
since you downloaded your spreadsheet.
b - Someone could have modified the number of columns in the spreadsheet
and unaligned the field to field relationship
c - Someone could misunderstood and deleted some rows in the spreadsheet
thinking it would just not update those, but would now leave those out of
the file with option *replace.
d - Us developers are just so much cleverer and slow down and think
(and, we need the job security)
If you decide to pursue this option I might suggest copying the old file
before the upload. Then compare number of rows, etc.
Most developers would recommend option 2. Again, appropriate backups of
the data. Pay close attention and decide what you want your final product
to be, and make sure your SQL can handle it, if you have
TABLE A
REF DESCR
1 BLACK
2 WHITE
3 RED
TABLE B
REF DESCR
1 PURPLE
3 GREEN
5 MAROON
3 - Been years since I touched Access. And I'm rather out of touch with
the tool of the day popular with PC development but a fair developer
should have some idea of how to code ODBC. If not, this is as good as any
first assignment to get their feet wet. First, I'd copy the table on the
i to ensure they don't muck it up during development and testing.
For some general statements with no detail you can try
http://publib.boulder.ibm.com/infocenter/iseries/v5r4/topic/sqlp/rbafydynamicsqlclient.htm
Don't know if this would be any better
http://publib.boulder.ibm.com/infocenter/iseries/v5r4/topic/rzaii/rzaiiodbcadm.htm
Rob Berendt
midrange.com
