Re: Read all about it - 8 IBM RFE's entered. Vote now!

Yep this would be the kind of thing I mean. There are also direct historical kinds of things - we have a process for downloading invoices to field specialist, and it has an ADDDAT and ADDTIM pair of fields.

These are kept around "forever", basically - there is a kind of purge.

These are used for an "as of" process, to know what was downloaded when the invoice is uploaded back. And what text values were in force when the download occurred.

Temporal support as in place now (system time) could likely be an excellent replacement (not sure in the case mentioned, there are considerations around the kind of changes were downloaded) for files like this.

Cheers
Vern

On 9/5/2017 6:45 AM, Rob Berendt wrote:

<snip>
Now for those of us who have already implemented some kind of history,
this could be about the same usage, just that the system automatically
manages more of it.
</snip>

You mean history like this? Like our biggest library, by far, on our ERP
machine is our journal receiver library?

% of Size in
Library Owner Disk 1000 bytes
#MXJRN MIMIXOWN 39.85 2391890501.6
QGPL QSYS 11.27 676212596.7
ERPLXF SSA 9.86 591586611.2
ERPLXSAVF SSA 2.95 177205379.1
QPFRDATA QSYS 1.56 93850730.5
ERPLXUSRF SSA 1.39 83154870.3
...

See RTVDSKINF and PRTDSKINF *LIB

We do some crazy journal retention and analysis. I think we are the
target market for temporal tables. They even have slides in their
presentations mentioning journal scraping.

Rob Berendt