Re: GOTO

Charles,

In a message dated 97-07-14 16:40:21 EDT, you write:

> People!
>  
>  If even part of the time spent fighting about GOTO or not to GOTO had 
>  been spent on dealing with actual concerns (like, the end of the 
>  century!)  We'd all have less junk to deal with, and our users would 
>  have better code!

Actually, we've probably spent a good deal MORE time talking about Y2K here
than about GOTO.  I seem to recall ad nauseum discussions on tools,
techniques, and compliant vendor releases.

>  Structure, like any other part of the process is to provide a 
>  function.  That function is to make the code more understandable.  I 
>  don't care HOW you code the stuff, just so I can figure out WHY you 
>  did WHAT you did when I have to go in and fix it!

Very true, and EXACTLY why I think that the GOTO discussion is appropo.
 Syntax is as important in your "quest for the understandable program" as is
documentation.  Given the code that most of us have run across at one time or
another that was written by someone without a mentor, I think that we can all
agree that people new to the AS/400 (many of which come HERE) will benefit
from syntax discussions.  The do tend to get long-winded though, and have
often been taken off-line for continued discussion.  I don't think that this
particular GOTO thread has reached that point yet, but it has in months
past...

>  We have an industry with code generators which "document" their code 
>  by inserting the same comments we were all taught not to use, (i.e. 
>  the following line adds AAA to BBB....)

Sounds like a bad code generator or an improperly set "comment generation"
parameter on it.  The AS/Set CASE tool can get ridiculous with this if you
use the "Full" comment option on generation, yet "Partial" merely inserts
your pseudo-code statements prior to the RPG that they generated.

>  We have code generators which generate the identical subroutine 
>  multiple times.

DOES sound like a bad generator, which one are you using?

>  We have date routines slavishly copied for decades which are 
>  incorrect.

Yes.

>  We have clients who refuse the correct technical solution to a 
>  problem in order to protect their turf.

What business doesn't?

>  We may have clients sueing us all for malpractice or the equivilant 
>  for the damned date fiasco.

Possibly, but most wouldn't have a leg to stand on.

>  We still have no known methodology to determine whether or not a 
>  piece of code really does what it is supposed to.

No, but we can get darned close with a good tool and a LOT of effort.

>  We still have no good method for generation of test data to test a 
>  new program.
>  
>  We still spend too much time hand-coding simple programs which ought 
>  to be trivial to generate (i.e. lookup & reports.)
>  
>  Most computers still respond to typographical errors with the 
>  equivilant of "HUH? I don't understand."
>  
>  Any one of these issues would return more value for time spent than 
>  adding to the "Why GOTO's are EVIL./Why GOTO's are needed." debate.
>  
>  <<END SOAPBOX>>

Agreed on the latter problems, but disagreed on the usefulness of their
discussion here.  Every application is different, so generation of test data
would hold different criteria for everyone on the list.  Many query tools
will handle the lookup/report situation; unfortunately, most have to mine
databases that are hard-coded at the application level.  As for "Huh, I don't
understand", I'd rather the computer tell me that than make an assumption
like "Oh, you meant to delete all the files in the production environment --
I can handle that!".

This will be my last post on the subject though, unless someone else stirs my
interest with another post :-)!

Regards,

Dean Asmussen
Enterprise Systems Consulting, Inc.
Fuquay-Varina, NC  USA
E-Mail:  DAsmussen@AOL.COM

"One of the greatest labor-saving inventions of today is tomorrow." --
Vincent T. Foss
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* 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 MAJORDOMO@midrange.com and specify            *
* 'unsubscribe MIDRANGE-L' in the body of your message.  Questions      *
* should be directed to the list owner / operator: david@midrange.com   *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *