[Issue 3847] To avoid a C code bug

[d-bugmail at puremagic.com]

http://d.puremagic.com/issues/show_bug.cgi?id=3847



--- Comment #12 from Stewart Gordon <smjg at iname.com> 2010-09-20 18:04:33 PDT ---
(In reply to comment #11)
> Some C++ programmers don't even know about 'and' and 'or' because 
> the C++ compilers do nothing to suggest, encourage or force the 
> usage of those more readable and less bug-prone operators.  The D 
> compiler can avoid that trouble if somehow && || become a legacy or 
> deprecated (but probably supported still) feature.

But because && and || are by far the most commonly used forms, and 
probably most modern languages with C-derived syntax are designed on 
this very basis, deprecating them would probably confuse the 
programmer by making D the odd one out.

Moreover, a trend that has distinguished C and its derivatives from 
Pascal, Fortran, SQL et al is the tendency to make use of symbols, 
thereby keeping down the number of keywords in the language.  D has 
continued this trend by doing away with the little-used and and or.  
OK, so D has is and in, but these aren't things for which other 
C-like languages have symbols (at least, I'm not sure if JS/PHP === 
is similar enough to count).

> If D style guides, standard library, newsgroups, and books use 
> 'and' and 'or' operators, new D programmers will use them.

And maybe get confused when they find that they don't behave in the 
same way as in Perl and PHP.  Probably the obscurity of C's and/or 
has meant that the designers of Perl and PHP saw nothing wrong with 
changing the precedence.  (Though why PHP's designer saw fit to 
change the associativity of ?: is a mystery I haven't solved.)

> But fixing bug 4077 doesn't help bugs like:
> 
> void main() {
>     bool a, b;
>     if (a & b) {}
> }

How's that a bug?  When a and b are bools, they behave the same!

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------