Jerry after George:

>> !DSCOPY PDB.DBASE TO ,FS2;FCODE=-400;REP
>> !DSCOPY PDB##.DBASE TO ,FS2 ;FCODE=-401;REP

> I'd be *very* careful....  This won't pick-up any POSIX files
> (b-trees and jumbo sets) and will also won't work with OMNIDEX
> index files....

We used to do a version of the above too, but as Jerry already
correctly pointed out it's easy to miss something with DSCOPY.
Besides B-Tree and JUMBO files (B-Tree FCODE = -412),
there are other traps that can sneak up on you.....  Here's one:

You have a database with <= 99 datasets.  The above DSCOPY
syntax works (as long as no B-Tree, JUMBOs, or OMNIDEX)...
BUT:  Let's say that some time after putting the above in a stream
file, you or a co-worker end up adding a couple new datasets with
a database tool, that pushes total number of datasets to >=100...
big time oops:   The last two characters of your dataset names
will now be A0, A1, etc...   and the "##" in the above DSCOPY will
*silently* ignore any dataset > 99....

My preferred method to copy databases:  Use a database tool.

Next best, if for whatever reason the above is not possible:
Jerry's suggestion:  LZW....  Although when HP makes STORE
to Disc (STD) universal like they just said they were going to do,
that should provide another option....

Ken Sletten