RE: Using SAV & RSTCHGOBJ as a method of saving records that

[Evan Harris <spanner@xxxxxxxxxx>]

Hi Dave

I seem to have missed the gist of what you are trying to do 
somewhere, so apologies if this doesn't help you with your problem.

It seems to me that saving and restoring objects just to identify 
only changed records would also require a significant amount of 
programming. I'm figuring maybe one program per file, although I 
suppose you might be able to automate the creation of an SQL 
Statement for each file which only retrieved records that were 
changed (I have no idea how to do this, but it strikes me as 
theoretically possible given that SQL is essentially text).

I think the journal suggestion would probably be easier and more 
readily extendible (unless you can make the SQL thing work) as the 
change record entries are easily identifiable using the journal code.

Again, without knowing what your actual problem is, is this something 
for which a trigger program might be a better vehicle for solving the problem ?

Regards
Evan Harris

At 08:06 a.m. 1/12/2005, you wrote:
> Michael,
>
> You wrote:
>
> >>You might be able to do a SAVCHGOBJ (there is no RSTCHGOBJ) and then
> a
> RSTOBJ, but I would write something using journalling myself. <<
>
> (BTW, is there an editor or mechanism that will allow me to
> automatically capture a person's reply on here and put into a reply to
> the forum and put in the necessary >> marks or the like?   I'm used
> something much more sophisticated than the copy and paste I'm doing but
> that was on a mainframe back in the '80s using REXX and XEDIT.)
>
> On to what you wrote...  I don't mind the idea of going after the
> journaled data, but that seems like a fair amount of programming unless
> something like it is already written and I can get a copy.   What are
> the advantages/disadvantages between the approach I offered and going
> against the journal?
>
> Thanks much,
>
> Dave