Re: read/chain/write/update with *input/*output/*all DS

Since we can't predict future, wouldn't this means that we should avoid
using global result-field DS? I think local result-field DS should be
fine since the result is only intended for local procedure consumption.


"Jon Paris" <jon.paris@xxxxxxxxxxxxxx> wrote in message
news:<mailman.3224.1282085837.2619.rpg400-l@xxxxxxxxxxxx>...

On Aug 17, 2010, at 6:01 PM, rpg400-l-request@xxxxxxxxxxxx wrote:

How/what would be the best way to handle above if I prefer to always

use
result-field DS when performing file IO opcode?

I agree with Scott that to rely on the pointer approach is a risk not

worth taking.

There are often very valid cases where use of result field I/O makes
perfect sense and offers significant benefits. It seems to me though
that in this case is not one of them. You are jumping through through

a lot of hoops and making the code harder to maintain and more fragile

simply to be able to use result I/O. To me that makes no sense.


Jon Paris

www.Partner400.com
www.SystemiDeveloper.com