We have run into similar problems.
There are several issues.
We do not have time to check every audit trail.
It should not be neccessary for our people to key stuff in, check the 
screen they keyed right, check the audit trail they did it right, have 
someone else check the audit trail.
Instead, we should have people on the payroll who have an extremely high 
rate of accuracy in what they do, and we should have equipment and software 
for them to use which makes it very easy to catch any mistakes that may 
occasionally occur.
We have exception reports looking for certain kinds of mispostings that 
need correction.
We have too many exception reports.
We do not have time to check them all.
We have made our own custom audit trails that are easier to use than those 
that came with BPCS.
Typically this is a Query/400 of someone's input, showing only the key 
fields whose accuracy is most critical.
Most of our users do not check any of the audit trails, BPCS, ours, or our 
exception reports.
The user in question knew that a mistake had occurred and volunteered the 
fact, asking for a correction.  Many human errors are not so easily 
discovered.
Some users do not even know they made a keying error.
Some people who know they made a mistake are afraid to report it because of 
what they fear about corporate culture.
Some people who know they made a mistake, they try to fix it without 
telling anyone, and sometimes they make it worse.
We have found transactions posted with dates outside of our fiscal 
calendar.  We run month end, or finish a physical inventory, then someone 
posts transactions that change the month end totals, by back dating them.
In the example, by the time this came to your attention, the Inventory 
transaction might have been communicated to the General Ledger.  Depending 
on the transaction with wrong date, the data can be replicated to other 
accounting records.
Thus if you use DBU or DFU or SQL or anything else to "fix" the original 
transaction, that has not fixed the files it got perpetuated to before this 
came to your attention.
Thus doing a negative transaction on the wrong date, then a positive 
transaction on the right date, gives you the assurance that every place the 
data is perpetuated to is also fixed.
We now have a corporate rule ... thou shalt not post a transaction using a 
date other than today date, the date you keying in the transaction.  If the 
work was actually done yesterday or last Friday, we treat it as if it is 
the date you keyed it in.
Some files contain date of actual data entry and date the user claimed for 
the transaction.
We can run query to list all postings to those files in violation of the 
company rule.
We rarely have time to check this.
Some custom programming we did was to look at fields on some screens, that 
are sometimes cleared from transaction to transaction, and to not blank 
fields that are usually the same from transaction to transaction, such as 
today's date.
Where your shop order reporting is apparently via INV500 M transaction, our 
shop order reporting is via JIT600 which might not be available with BPCS 
4.0.03.  Basically JIT600 supports keying in both labor and inventory with 
a single transaction.  Some BPCS programs do not support negative 
transactions, so you have to know the relevant work around.  Some BPCS 
programs do not support entering the same transaction again corrected, so 
you have to know the relevant work around.
I think that if HOW TO FIX this or that was generally made available to 
people in their documentation of how to do things, people would be more 
likely to fix things correctly and promptly.
Al Macintyre
BPCS/400 Computer Janitor at 
http://www.globalwiretechnologies.com/
See Al at 
http://www.ryze.com/go/Al9Mac
Find BPCS Documentation Suppliers 
http://radio.weblogs.com/0107846/stories/2002/11/08/bpcsDocSources.html
Folks --
We are BPCS 4.0.03 with customer Y2K mods.  The email below was sent to
me asking for help.
==User error email ==============
As you know, Lupe is currently training me in the process of
"reporting" shop orders.  This transaction is under "INV", "1" Inventory
transactions, "M" Multiple Issue with Receipt.
When I reported this SO #19377 (JC Pasta Ole), I entered the incorrect
date of "10/29/03" when it actually should be "01/29/03". Please make
the necessary transactions to reflect the correct date.
==END user email ==============
The previous analyst would manually (with DBU) correct such errors.  I
wish to avoid such silliness; I am a huge fan of audit trails.  My
initial response was to advise the user to reverse the erroneous
transaction by entering it the same but with negative quantity.
a> Is this a good response?
b> Is there any way to limit the dates that can be entered into
inventory transactions short of custom programming?
Thanks for any advise you may have.
_______________________________________________
This is the SSA's BPCS ERP System (BPCS-L) mailing list
To post a message email: BPCS-L@midrange.com
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/bpcs-l
or email: BPCS-L-request@midrange.com
Before posting, please take a moment to review the archives
at http://archive.midrange.com/bpcs-l.
As an Amazon Associate we earn from qualifying purchases.