[boost] Re: Misc library documentation issues

From: Reid Sweatman (reids_at_[hidden])
Date: 1999-07-13 12:38:16

> I am against the use of macros in general, and their 
> abuse in particular. Macros are invasive, and very 
> often lead to obcsure code.

I couldn't agree more.  In fact, I spent an hour last night chasing down a
bug that turned out to be an obscure interaction between a macro defined in
Windefs.h, then redefined (but #ifdef'ed out so it wouldn't collide with
MS's macro), then overridden in a bizarre manner in a module in the legacy
code I hadn't even been aware was being included in my code.
 
> In some cases, macros are fine.

Yup.  Although I find that when I'm most tempted to use them is when
deadlines are fast approaching and the damn thing's just got to work, come
hell or high water.  But then, I'd call that an expedient kludge (probably
the cause of 95% of legacy code problems, IMHO).
 
> In this case, the problem isn't macros but exception 
> specifications, which are considered as a Bad Thing by 
> some people (Dave and probably others).

I'm a fence sitter; I want to believe, but I want proof first.  What would
St. Paul say? <BG>.  I think B. Gates is more of the St. Augustine camp when
it comes to implementing the spec:  "Please, Lord, give me faith!  But not
yet, not yet!" <EBG>

 

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/boost
http://www.egroups.com - Simplifying group communications