Robert Vazan <[EMAIL PROTECTED]> writes:

> On Mon, 13 Oct 2003 14:51:51 -0400 David Abrahams
> <[EMAIL PROTECTED]> wrote:
>
>> Robert Vazan <[EMAIL PROTECTED]> writes:
>> 
>> >>  Looking at the test matrix at boost.org, VC6 does support these classes...
>> >
>> > I am talking about optimizations. Templates don't have to be instantiated
>> > for every type, but compilers won't realize it.
>> 
>> I don't know what you mean.  Nothing about the boost shared_ptr
>> design causes aggressive template instantiation.
>
> I know it, but did you test it with Gcc and VC6? 

Yeah, and about 15 other compilers.

> The official benchmark, I linked before, shows that Gcc slows down
> even if you just wrap raw pointer in template. 

What do you mean by "slows down"?  Do you mean that the compilation
times get longer?

> I don't believe that templates work. 

I guess for some definition of "work", that's true.  They won't polish
my kitchen floors, but they do what they were designed to, and in fact
a whole lot more.

> I am using them like macros and I shift as much code as possible to
> non-template functions.

With all due respect, it sounds like your view of templates is
FUD-laden.  That sort of relationship with any language feature
usually means you never get to the point of being able to use the
feature powerfully, and all your experiences tend to reinforce the
FUD-dy viewpoint.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
SourceForge.net hosts over 70,000 Open Source Projects.
See the people who have HELPED US provide better services:
Click here: http://sourceforge.net/supporters.php
_______________________________________________
Mahogany-Developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mahogany-developers

Reply via email to