• dan <dan_member pathlink.com> Jan 10 2004
• "Matthew" <matthew.hat stlsoft.dot.org> Jan 10 2004
• "Jonathan Turkanis" <technews kangaroologic.com> Jan 11 2004

template< typename Y >
class x
{
friend class Y;
};

Is an error, but compiles in DM.  Comeau on-line compiler says:

"ComeauTest.c", line 4: error: template parameter "Y" may not be used in an
elaborated type specifier
friend class Y;
^
Cheers!

 template< typename Y >
 class x
 {
 friend class Y;
 };

 Is an error, but compiles in DM.  Comeau on-line compiler says:

 "ComeauTest.c", line 4: error: template parameter "Y" may not be used in


 elaborated type specifier
 friend class Y;
 ^
 Cheers!


This is about the least valid error in the business. Check out
http://www.cuj.com/documents/s=8942/cujweb0312wilson/, and also note that
Comeau have recently added a --friendT option to v4.3.3 to specifically
legalise it when compiling in --strict mode. This behaviour is one of the
one's most likely to
be legalised in the next version of the standard, and is profoundly useful,
so it'd be crazy for any vendor to attempt to change their support for it.

Cheers


-- 
Matthew Wilson

STLSoft moderator (http://www.stlsoft.org)
Contributing editor, C/C++ Users Journal
(www.synesis.com.au/articles.html#columns)

"If I'm curt with you it's because time is a factor. I think fast, I talk
fast, and I need you guys to act fast" -- Mr Wolf

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

"Matthew" <matthew.hat stlsoft.dot.org> wrote in message:
 template< typename Y >
 class x
 {
 friend class Y;
 };

 Is an error, but compiles in DM.  Comeau on-line compiler says:

 "ComeauTest.c", line 4: error: template parameter "Y" may not be




 an
 elaborated type specifier
 friend class Y;
 ^
 Cheers!


 This is about the least valid error in the business. Check out
 http://www.cuj.com/documents/s=8942/cujweb0312wilson/, and also note


 Comeau have recently added a --friendT option to v4.3.3 to


 legalise it when compiling in --strict mode. This behaviour is one


 one's most likely to
 be legalised in the next version of the standard, and is profoundly


 so it'd be crazy for any vendor to attempt to change their support


 Cheers


You probably noticed a recent post by Andrei Alexandrescu on
comp.lang.c++.moderated refering to your article. (See
http://www.talkaboutprogramming.com/group/comp.lang.c++.moderated/messages/153854.html.)
He asked "are friend template arguments a de facto feature ready to
make it into the standard?" I was hoping some standards committee
members would respond, but none did.

What is your evidence that this will likely be legalized? I certainly
hope you are right.

Jonathan