• "Uwe Salomon" <post uwesalomon.de> Apr 06 2005
• Stewart Gordon <smjg_1998 yahoo.com> Apr 06 2005
• "Matthew" <admin.hat stlsoft.dot.org> Apr 06 2005

"Uwe Salomon" <post uwesalomon.de> writes:

Hello,

what exactly is the reason for the existence of a default opCmp() for  
classes? Every class that an opCmp() really makes sense for has to  
implement it anyways, and the others don't want to be compared i guess. If  
there weren't a standard implementation (or the standard would only throw  
an exception) this would help detecting programming errors. Or is the  
operator necessary for the AA-implementation?

Ciao
uwe

Stewart Gordon <smjg_1998 yahoo.com> writes:

Uwe Salomon wrote:
 Hello,
 
 what exactly is the reason for the existence of a default opCmp() for  
 classes? Every class that an opCmp() really makes sense for has to 
 implement it anyways, and the others don't want to be compared i guess. 
 If there weren't a standard implementation (or the standard would only 
 throw an exception) this would help detecting programming errors. Or is 
 the operator necessary for the AA-implementation?


I'm not sure if anyone knows the original reason for putting it in.  And 
several of us have been wanting to get rid of it for ages.

digitalmars.D/10558

Stewart.

-- 
My e-mail is valid but not my primary mailbox.  Please keep replies on 
the 'group where everyone may benefit.

"Matthew" <admin.hat stlsoft.dot.org> writes:

"Uwe Salomon" <post uwesalomon.de> wrote in message
news:opsotlm9jm6yjbe6 sandmann.maerchenwald.net...
 Hello,

 what exactly is the reason for the existence of a default opCmp() for 
classes? Every class that an opCmp() really 
 makes sense for has to  implement it anyways, and the others don't want to be
compared i guess. If  there weren't a 
 standard implementation (or the standard would only throw  an exception) this
would help detecting programming errors. 
 Or is the  operator necessary for the AA-implementation?


I think you have it right. FWIW, it's a very unpopular feature.