OT: Coding revolution, was: Seeking: IBM Coding Forms

As you all know, I'm heavily biased towards the hardware side, but much of it is the same there... Agreed... 'Banning' certain things becuase they make code (or hardware) difficult to undersnd or maintian is silly IMHO. I've sene pefecrly understnadable progeams with GOTOs (or their equivalent). I've seen right messes trying to avoind them (condtionals around just about everything). I've sene right messes of hardware built with monostables all triggering each other so if one drifts a bit the whole lot falls over. And I've seen them used properly. I've seen a right mess which claimed to be a state machine too... Again I think it's a matter of (good) judgement. You can add small extra feautres ot hardware or software without problems. But as you said, if you want a major change, it's time tore-think the deisng from the start. I would add soemthing else that managers don't like, and that is that using 'standard solutions' is not always best. Of course subroutine libraries are a Good Thing. But over-use of them, when you mess about wit hthe problem so that you can use one library routine after another, this routine for this case, and that routine fo thtat case, often makes a less reliable solution than writing the special routine to solve that problem. Similarly in hardware, I don't like starting from a stnadard module (microcontorler PCB, etc) and somehow making it fit the problem. I find it better to actually think of the solution. -tony