Comments on: Memoization http://efesx.com/2010/01/07/memoization/ Technical & low level Mon, 30 Apr 2012 17:39:06 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: rmn http://efesx.com/2010/01/07/memoization/#comment-163 rmn Mon, 08 Feb 2010 18:14:22 +0000 http://cplusplus.co.il/?p=752#comment-163 It is true that Boost already contains utilities like the ones you have mentioned, thanks for bringing them up. The relevant boost::remove_reference utility can be found <a href="http://www.boost.org/doc/libs/1_41_0/libs/type_traits/doc/html/boost_typetraits/reference/remove_reference.html" rel="nofollow">here</a>. A point worth noting though, is that there is much to be learned from small snippets like the RemoveRef class mentioned here. So it only benefits more readers that way. Thanks again for the kind words. It is true that Boost already contains utilities like the ones you have mentioned, thanks for bringing them up. The relevant boost::remove_reference utility can be found here.
A point worth noting though, is that there is much to be learned from small snippets like the RemoveRef class mentioned here. So it only benefits more readers that way.

Thanks again for the kind words.

]]> By: Manuel http://efesx.com/2010/01/07/memoization/#comment-162 Manuel Mon, 08 Feb 2010 08:29:00 +0000 http://cplusplus.co.il/?p=752#comment-162 Great post rmn. Maybe you should mention that Boost provides a "remove_reference" class together with other related goodies like "remove_const" or "remove_pointer" Great post rmn. Maybe you should mention that Boost provides a “remove_reference” class together with other related goodies like “remove_const” or “remove_pointer”

]]> By: rmn http://efesx.com/2010/01/07/memoization/#comment-161 rmn Fri, 22 Jan 2010 18:16:10 +0000 http://cplusplus.co.il/?p=752#comment-161 That's exactly what I was going to say. Chips? :) That’s exactly what I was going to say. Chips? :)

]]> By: ManicQin http://efesx.com/2010/01/07/memoization/#comment-160 ManicQin Fri, 22 Jan 2010 18:14:34 +0000 http://cplusplus.co.il/?p=752#comment-160 Lately I'm asking and replying too much ;) It'll be more STL compatible and you could use bind1st and etc... Lately I’m asking and replying too much ;)
It’ll be more STL compatible and you could use bind1st and etc…

]]> By: ManicQin http://efesx.com/2010/01/07/memoization/#comment-159 ManicQin Fri, 22 Jan 2010 17:55:07 +0000 http://cplusplus.co.il/?p=752#comment-159 I think that the minute they named the lambda and made it recursive they lost the point behind lambda. :) IMO lambda should stay small and simple (It becomes unreadable). Besides, Their way is too intrusive. In your way you can just apply the functor to any code, and you can apply other functors on top of it. I see no actual reason to inherit from std::unary_function, maybe only for readability (showing other developers that this is a functor) Do you see any benefits? I think that the minute they named the lambda and made it recursive they lost the point behind lambda. :)
IMO lambda should stay small and simple (It becomes unreadable).
Besides, Their way is too intrusive.

In your way you can just apply the functor to any code, and you can apply other functors on top of it.

I see no actual reason to inherit from std::unary_function, maybe only for readability (showing other developers that this is a functor)
Do you see any benefits?

]]> By: rmn http://efesx.com/2010/01/07/memoization/#comment-158 rmn Fri, 22 Jan 2010 16:25:37 +0000 http://cplusplus.co.il/?p=752#comment-158 And that is done to allow less-restrictive usage. I like the link to the Visual C++ Team blog, even more so since it deals with Lambda functions (and few other C++0x features). However, I think the implementation here is far more elegant, and can be adapted to work in their scenario as well. We could even make Memoizer inherit from std::unary_function. What do you think? Thanks for dropping a line. And that is done to allow less-restrictive usage.

I like the link to the Visual C++ Team blog, even more so since it deals with Lambda functions (and few other C++0x features). However, I think the implementation here is far more elegant, and can be adapted to work in their scenario as well. We could even make Memoizer inherit from std::unary_function. What do you think?

Thanks for dropping a line.

]]> By: ManicQin http://efesx.com/2010/01/07/memoization/#comment-157 ManicQin Thu, 21 Jan 2010 14:20:06 +0000 http://cplusplus.co.il/?p=752#comment-157 Ignore that I didn't see that operator() is const Ignore that I didn’t see that operator() is const

]]> By: ManicQin http://efesx.com/2010/01/07/memoization/#comment-156 ManicQin Thu, 21 Jan 2010 13:46:36 +0000 http://cplusplus.co.il/?p=752#comment-156 By the way, Why is cache_ mutable? By the way,
Why is cache_ mutable?

]]> By: ManicQin http://efesx.com/2010/01/07/memoization/#comment-155 ManicQin Wed, 20 Jan 2010 12:10:40 +0000 http://cplusplus.co.il/?p=752#comment-155 A different approach to Memoization and fib http://blogs.msdn.com/vcblog/archive/2008/11/18/stupid-lambda-tricks.aspx :) A different approach to Memoization and fib

http://blogs.msdn.com/vcblog/archive/2008/11/18/stupid-lambda-tricks.aspx
:)

]]> By: alpmestan http://efesx.com/2010/01/07/memoization/#comment-154 alpmestan Fri, 08 Jan 2010 13:20:14 +0000 http://cplusplus.co.il/?p=752#comment-154 Very nice ! Thanks for the post. Very nice ! Thanks for the post.

]]>