On Mon, 22 Nov 2010 04:38:37 -0600, Chris Mason wrote: > >Is there any impact whatsoever in having or in not having the macro in the >source code on the reliability of the generated logic?: > >Of course not - but it's here that the timorous might allow themselves to be >influenced by the FUD factor.
Yes, there is, and it is not FUD. Timorous? You are repeating yourself. It is not based upon fear. I will agree that it is true that one can, with sufficient effort, generate the parameters correctly without using the macros. The difficulty comes later when someone else needs to modify the code. I submit that, having taken the trouble to discern what the macros do, it is no more difficult to use the macros to generate the code. More importantly though, when the WTO code is generated using the macros it is much easier for someone else to modify it later. In fact, I have had several situations in the past few years when I have had to modify someone's WTO code that was created without the use of the WTO macro. Perhaps they found it too difficult to get the results that they desired using the macro. I have no way of knowing that. When I figured out what they were trying to do, I had no difficulty at all coding a macro that gave the same results as they had, but it took considerable time to decode the code that I had to modify. Having coded the macros (L and E) to replace the DIY code, it was a simple matter to modify it to suit the new requirements. Yes, I read where you wrote that when doing this kind of thing it is important that the code be adequately documented. Unfortunately, when reading 30 year old code that has been modified several times, there is no way to know whether the comments are correct. Indeed, even if the author tried hard to get the comments right and no one has modified the code or the comments, there is no guarantee that the documentation is correct. For me, the only good documentation as to what the code was trying to accomplish is the macros that were used to generate the code. They are more or less clearly defined and they are guaranteed to match the generated code. Please do not misunderstand. Comments are important and I rely on other programmer's comments every day. And I try to provide useful comments for those who will follow me. Still, it is important to verify comments because sometimes they do not match the code. When I find code such as what I have described, I can't help but to conclude that the author was enamored of his own brilliance and uninterested in any difficulties that his coding style caused us mere mortals. Frankly, I am not impressed. -- Tom Marchant ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
