‎2009 Jun 12 4:08 PM
Hi,
I'm just interested to know what people make of this article:
Malcolm.
‎2009 Jun 12 4:19 PM
Nothing new in it. See and many other posts in this forum.
Rob
‎2009 Jun 12 4:36 PM
Rob, correct me if I'm wrong, but what is usually said is that MOVE-CORRESPONDING has a negligible impact on performance, and that its advantage would outweigh this impact.
What the graphs in the article show is that MOVE-CORRESPONDING actually has better performance.
‎2009 Jun 12 4:45 PM
I believe he was saying that the impact of using it on was small. Therefore the disadvantage of using it does not outweigh any small advantage.
Rob
‎2009 Jun 12 4:46 PM
Funny noticing, though, what Matt wrote at the end of [this thread|; (MOVE vs =), on Apr 23.
>
> Not really - it's posted about once a month. Next month, you can watch the one about MOVE-CORRESPONDING.
This was more than a month and a half ago, so I guess these threads are appearing less often than they used to..
‎2009 Jun 12 4:50 PM
Maybe I should attach this article to the thread I mentioned above.
Rob
‎2009 Jun 12 5:33 PM
Imagine all the nice logic that would not be possible without CORRESPONDING, e.g. just add a field to your DDIC-ALV-structure and whoops it is being read from the database and shows up in your list (provided your program is coded accordingly). Or append a field to BSIS with same name as in BKPF (e.g. CPUDT), and it will be filled from now on by the SAP standard.
There will be many more examples.
My point is: forget about the minimal performance aspect, use it for purely for program logic where applicable.
Thomas
P.S. just repeating what I said in one of the threads linked in the sticky note