|
|
The only thing I can see is cosmetic, when you're using them.
I'm wondering if one way compiles faster than the other (I assume method 1, because there's less stuff to parse), and if they produce the same code in the exe (I assume they do)?
// Method 1
template < class T >
class Utils
{
public:
static void doSomething(T* object);
static void doSomethingElse(T* object);
};
// Method 2
class Utils
{
public:
template < class T >
static void doSomething(T* object);
template < class T >
static void doSomethingElse(T* object);
};
// A class
class ClassA
{
public:
};
ClassA a;
Utils<ClassA>::doSomething(&a); // Method 1
Utils::doSomething<ClassA>(&a); // Method 2
Just curious...
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
|
Why not?
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
There was mistake from my side. Yes there is only cosmetic change as u have mentioned only static function in both classes.And who will compile faster,both will take same time . Difference in compilation speed will be there if u mention a non templated member function in both the class
-- modified at 3:37 Friday 4th August, 2006
-- modified at 3:37 Friday 4th August, 2006
-- modified at 3:38 Friday 4th August, 2006
-- modified at 3:47 Friday 4th August, 2006
never say die
|
|
|
|
|
Thanks.
I was just thinking method 1 would compile faster, because it wouldn't have to parse all the extra "template < class T >" statements.
I was trying to put all the implementation in a .cpp, but after reading this http://www.codeproject.com/cpp/templatesourceorg.asp[^] and the linker problems (which I've already run into), I've decided to put the implementations of each function at the bottom of the .h, so compile time will take a hit anyway. Oh, well...
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
Nother thing I want to add is that in class 2 u can it is not neccesary to put same parameterised variable
class Utils
{
public:
template < class T >
static void doSomething(T* object);
template < class E >
static void doSomethingElse(E* object);
};
but in first u can do that
never say die
|
|
|
|
|
You can't do that?
Basically I'm writing templated functions, so I thought I could re-use < class T >, since it only applies to each function.
Of, course I'm getting an INTERNAL COMPILER ERROR from MSVC6 right now, so who knows what's going on. 
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
It compile well even if i use E in place of it in VS2005
never say die
|
|
|
|
|
I angered the MSVC6 compiler with Method 2 (INTERNAL COMPILER ERROR), so I guess method 1 will have to do.
I guess it doesn't like template functions within a class. Has anyone have heard of this? Is my syntax wrong?
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
I build it on VS 2005 it is compiling without any error
-- modified at 5:02 Friday 4th August, 2006
never say die
|
|
|
|
|
|
try it on gcc compiler . :-OI prefer open source product
-- modified at 5:21 Friday 4th August, 2006
-- modified at 5:23 Friday 4th August, 2006
never say die
|
|
|
|
|
|
Thats why use open source product and u can make them work as u want 
never say die
|
|
|
|
|
That's true. If only there were 42 hours in the day, then I would have the time to make it work the way I want. Unfortunately, I don't have the time to make everything work the way I want, so I have to use the programmer's best friend - "the workaround".
Thanks for the help anyway!
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
Steve Echols wrote: If only there were 42 hours in the day ... so I have to use the programmer's best friend - "the workaround".
I do hope you realize the irony of this statement. Many times, it takes one far more time to "workaround" an issue than it does to do things the correct way.
GCC is far easier to use from the command line than Microsoft's compiler.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Zac Howland wrote: I do hope you realize the irony of this statement.
Yes. I'm a bit of a smart ass sometimes.
Zac Howland wrote: Many times, it takes one far more time to "workaround" an issue than it does to do things the correct way.
Unless the compiler is flawed, and won't let you do it the correct way, then you have to work around it, or take a different approach.
Zac Howland wrote: GCC is far easier to use from the command line than Microsoft's compiler.
No doubt, but I like hitting F7.
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
Steve Echols wrote: No doubt, but I like hitting F7.
As long as you realize that you are spoiled, I guess there is no problem with it 
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
sunit5 wrote: try it on gcc compiler . I prefer open source product
gcc is pretty much OK, but the debugger (gdb) sucks.
|
|
|
|
|
Steve Echols wrote: I've always heard that VC6 is a crappy compiler and isn't up to standards
At the time of its release, it was almost 95% compliant with the standards. At the time, that was the highest of any compiler out there. There is a reason that VC6 is still widely used for Windows development.
That said, when dealing with templates, you have to be very careful with any compiler. they are all rather picky about them.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Zac Howland wrote: At the time of its release, it was almost 95% compliant with the standards. At the time, that was the highest of any compiler out there. There is a reason that VC6 is still widely used for Windows development.
I've been using MSVC, since they first came out with it, and have always loved it. I've just heard other people saying how it isn't compliant, etc, etc, but I've never had a problem with it until this. Guess I'm really pushing the edge of the envelope now! 
- S
50 cups of coffee and you know it's on!
|
|
|
|
|
Not that I like to defend Microsoft for anything ... but in their defense, almost every compiler has problems with templates. That is an area of the standard that wreaks havok on every single compiler I've used (MSVC, Intel, GCC, Borland).
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Hi all,
I have to compare two double values, say x and y. I compared using this simple method x < y, but the result seems to be varying in different cases. Sometimes eventhough x and y are equal, the output may be x is less than y and viceversa,
So Please help me to have a good method for comparison of double values.
Thanks in Advance
Velayudhan
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
I've always heard that VC6 is a crappy compiler and isn't up to standards, now I'm beginning to agree...