BG> [snip cut]
 DM> Simply trying to compile this should show you that 
 DM> either it is illegal
 DM> code, or your compiler wouldn't support it anyway.
 BG> That's not exactly a rigid formal proof :o)
No, but if you were merely writing code, I'd say that it was a _practical_ 
proof.  That is, if your compiler doesn't accept it, you couldn't write it 
that way.  ;-)
 DM> At any rate, the code makes zero sense to me anyway - what is it
 DM> exactly that you are trying to do?
 BG> It's a compiler-like program I'm writing that needs to 
 BG> read/produce C++ sources. I was wondering if the odd code was correct.
I would suggest being able to handle reading it, but not writing it.  It's 
pretty ugly-lookin' code.  ;-)
 DM> Is it just try bizarre scenarios,
 BG> Yep. The reference manual in "The C++ Programming 
 BG> Language" does tend to get somewhat vague.
... if somewhat out of date.  ;-)
 DM> or is there a problem that would be solved by this construct?
 BG> Nope.
If there was a specific problem that would be perfectly modelled with this 
code style, I'd say ask your compiler vendor to fix their parser.  :-)  
However, I think it's safely low on any compiler vendor's priority list.  Can 
ANYONE come up with a problem that would be modelled by this code style?  I 
know I can't...
 BG> ... If you can't stand the heat, turn the central heating down. Duh!
:-)
---
---------------