RE: DOW & DOU

Tim

At 03:04 PM 5/8/1998 -0400, you wrote:
>On Friday, May 08, 1998 1:52 PM, boothm@ibm.net [SMTP:boothm@ibm.net] 
>wrote:
>> I am confused.
>>
>> If I was going to read the file from beginning to end why wouldn't I just
>> use the rpg cycle? If I am just processing a section of the file, and 
>want
>> to get out of a loop in the odd event that I am at a group that happens 
>to
>> contain the last record in a file, then this structure serves no useful
>> purpose.
>
>Booth,
>As you know I am relatively new to the AS400.  But I can tell you the first 
>time I used an IP (Input Primary) at my current site, you would have 
>thought I was the anti-Christ!  I was condemned as an heretic, and told 
>never to use the RPG cycle again.  I would also like to be led by some of 
>the bright people on this list.  I think you should use WHATEVER works the 
>best... namely... RPGII had neat level break opportunities, as well as the 
>Primary file process technique, wheras the Combined Full method of program 
>screen handling of RPGIII and later is a perfect manner for screen 
>handling.  As you are, Booth, count me slightly confused on this issue. 
> Someone give us the dope!

I'll be the dope!

Your imagery is appropriate. This could well be one of the all-time
religious battles in the works.

Personally I think there's no faster way to write or run simple reports,
than to use control level breaks (whatever they're called—you know what I
mean). There've been good examples in this former thread of ways to avoid
the 'ugly' LH-indicator on every C-spec style. Likewise, matching records
processing can make certain things very easy to write, I hear. In fact, one
of our programmer's been asked to use them just now. To be honest, I've not
been around long enough to use M1-M9,MR, but I think I like it for some
stuff. It 'automagically' gets you unmatched records as well as matching,
right? And you don't have the overhead of either OPNQRYF or the various
uses of STRQMQRY (albeit good in their places). I admit that it can suffer
from wide separation of indicators, which can be a problem—document,
document, document.

I agree with you—I like to use the best tool for the job at hand. My very
recent experience with a little DOW/DOU benchmarking suggests that what we
think 'looks' good in theory may, in fact, be implemented in a less
efficient manner by the compiler/interpreter/whatever.

Alright, send in the flames!  :-)

Not Dopey anymore, just Sleepy

Vernon Hamberg
Systems Software Programmer
Old Republic National Title Insurance Company
400 Second Avenue South
Minneapolis, MN  55401-2499
(612) 371-1111 x480


+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---