• Daniel Horn (19/19) Apr 29 2004 I can see why one would wish to avoid virtual template class template
• Walter (6/25) Apr 29 2004 No.

Daniel Horn <hellcatv hotmail.com> writes:

I can see why one would wish to avoid virtual template class template 
member functions to keep a fixed size vtable (I miss them however!) Is 
that a feature that will be added?

How about template struct template member functions: these are just 
syntactic sugar for making a static function...

Speaking of making a static function...I made 2 vector classes, one for 
float vectors one for double, and I wish to multiply them.

I am stuck with


XVector!(double) k=XVector!(double).opCall(1,2,3);
XVector!(float) l=XVector!(float).opCall(3,4,1);//why must call opCall?

XVector!(double) m = VecOp!(typeof(k.x),typeof(l.x)). Mul(k,l);


typing VecOp!(typeof(k.x),typeof(l.x).Mul(k,l)
is lengthy, and obscures the function of the code, which is to multiply 
k with l.

in C++ I would make a macro at the very worst (an overloaded operator at 
best) that did it
#define VecMul(k,l) VecOp!(typeof(k.x),typeof(l.x).Mul(k,l)

is there any shorthand for this, short of running gcc -E before gdc?


PS: Typeof is really cool!

"Walter" <newshound digitalmars.com> writes:

"Daniel Horn" <hellcatv hotmail.com> wrote in message
news:c6rlj9$2aga$1 digitaldaemon.com...
 I can see why one would wish to avoid virtual template class template
 member functions to keep a fixed size vtable (I miss them however!) Is
 that a feature that will be added?

No.

 How about template struct template member functions: these are just
 syntactic sugar for making a static function...

Already implemented.

 Speaking of making a static function...I made 2 vector classes, one for
 float vectors one for double, and I wish to multiply them.

 I am stuck with


 XVector!(double) k=XVector!(double).opCall(1,2,3);
 XVector!(float) l=XVector!(float).opCall(3,4,1);//why must call opCall?

 XVector!(double) m = VecOp!(typeof(k.x),typeof(l.x)). Mul(k,l);


 typing VecOp!(typeof(k.x),typeof(l.x).Mul(k,l)
 is lengthy, and obscures the function of the code, which is to multiply
 k with l.

 in C++ I would make a macro at the very worst (an overloaded operator at
 best) that did it
 #define VecMul(k,l) VecOp!(typeof(k.x),typeof(l.x).Mul(k,l)

 is there any shorthand for this, short of running gcc -E before gdc?

Will an alias work?

 PS: Typeof is really cool!

Thanks!