Re: New to RPG

[Booth Martin <booth@xxxxxxxxxxxx>]

The issue, to my mind, is that way too often a program written years ago 
has quirks that solve a business problem and the purpose of the quirk  
isn't readily apparent when deciding to rewrite the program in a new 
paradigm.  The result is that suddenly a business rule is broken and 
everyone is screaming and setting hair on fire.  Or worse, no one 
notices and tons of damage is done.

It is understandable that twice burned, a manager will say "If it isn't 
broken you can't fix it; you can only break it.  Leave it alone."

My own choice, given the decision to make, would be to do rewrites every 
generation or so by application, not by program.  A cycle of maybe 12 
years?  I have code out there still running from the 1980s.  I shudder 
to think of how brittle that code must be by now.


On 4/30/2017 8:10 AM, Buck Calabro wrote:
> I agree with your assessment, but I especially agree with this.
>
> Copying code without understanding it is killing us.
> If we understood it, we'd never copy it -- we'd write better.
> Not because the code is old per se, but because that pattern is no
> longer the best fit to the problems it is being twisted into solving.
>    --buck