On Mon, 31 Mar 2003, Peter Sergeant wrote: > > If your code issues > > warnings, it's wrong. > > It is? Warnings are just that: warnings. Information to make you sit up > and say 'is that what I meant to do'? If your code doesn't compile, it's > probably wrong, whether or not it issues warnings is irrelevant.
This is horribly remeniscent of the argument, "It compiled, so it must run". Warnings are things that tell you when you did something you shouldn't. If you're aware that you're deliberately doing such a thing, then you should disable the particular warning for a small portion of the code where you do that thing. Odd, this is the same "correctness" argument from firewalling applied to correctness of code: Deny everything which could be bad for you, then allow what you explicitly want. But nobody would argue the point in the case of firewalls. Why do so just because it's correctness of code that's at stake? Code is seen as "less critical" somehow? Less subject to the same stringent measures of correctness? Even when 99% of the time, a code error will have a far greater effect than a hole in a firewall. Anyway... S. -- Shevek http://www.anarres.org/ I am the Borg. http://design.anarres.org/