Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Cursor Question

Re: Cursor Question

From: Kenneth C Stahl <kstahl_at_lucent.com>
Date: Tue, 22 Jun 1999 09:51:26 -0400
Message-ID: <376F94DE.6F530A05@lucent.com>


If you really want something that boggles one's mind with all sorts of implications, consider the setjmp()/longjmp() of C/C++. Read the man pages for those two functions and your mind will begin to race. In effect it is a GOTO, but it is only a backwards goto. It has its uses however when writing signal handlers.

The thing is, if a program is absolutely peppered with GOTOs it probably indicates a problem with the basic design of the program. But, when it is a practical way to handle a specific problem then it is an acceptable solution.

I have a story I tell on myself that cured me of weenieism. I once wrote a C program that contained conditional operators nested 5 deep (you know, the (condition) ? if_true : if false) in two different branches. It was ingenious and I thought it was a beautiful solution - until I had to modify it a few months later and found I couldn't even read my own code. Ever since then I have believed the KISS maxim.

Ken

Marc Mazerolle wrote:

> I use to think this way (back in my C/C++ days) but have you ever tried to
> follow code that uses Oracle "exception" to it's limit ? That is spaghetti.
>
> One well placed "GOTO", i think, is acceptable.
>
> Regards,
>
> Marc Mazerolle
>
> DBAOracle wrote:
>
> > Am I the only person that cringes seeing the goto in this code? As an old C
> > programer, goto's make me sad...
> >
> > Robert
Received on Tue Jun 22 1999 - 08:51:26 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US