On the 16 Mar 98, Jonathan de Boyne Pollard performed the most audacious feat 
of simultaneous gibbering and moaning. It was generally directed at me
[code fragment]
 JdB> According to my reading of the ISO C++ Standard, the answer is,
 JdB> surprisingly, yes.
Same here, using the reference manual in TC++PL.
[::foo() uses ::qwe not bar::qwe]
 JdB> I deduce this from the fact that none of the name lookup rules
 JdB> relating to class scope in section 3 of the ISO C++ Standard apply to
 JdB> the names used in definition of a function that *isn't* a member
 JdB> function.
I caught onto that code fragment when I realised that there was no 
restriction on where a qualified name can be used, or what scope can legally 
be refereed to. As well as the fact that TC++PL's ref-manual implies that the 
scope of a qualified declarator becomes active.
 JdB> Certainly it is one that many implementors have overlooked.
Except me :o) I'm current involved in a compiler-like project. I achieved 
what I assume to be the correct behaviour accidentally.
 JdB> Several compilers that I've tried will fail to compile this code.
Same here.
At I'll stand a good chance of winning any "Most cryptic&obtuse C++ code" 
awards. It's not every day you write legit code that's so weird that even 
most compiler software is unable to recognise it! :o)
Looks like *another* happy ending!
 offworld@bleach.demon.co.uk   >##<
... Ah, Dinner's ready.  The smoke detector just went off!
 >> Tag-o-Miga 1.1a - Quoted 72% - 772 tags in Taglines.txt <<
--- Spot 1.3a #1508
---------------