Steven Watanabe wrote:

...

This is not entirely optimal though as it requires 3 MOV instructions to create the scs void* copy when SYNCH is false (and I'd rather not lean on compiler optimizations even if they will help in this case).

Why not? Have you actually looked at what compiler optimizations do? The optimization required is really not very sophisticated.

Mmm. I know that compilers do much more sophisticated optimization than this but I just figured that there was a more elegant way to express my actual intent here. Using a 'void*' when it isn't ever actually used/read seems like a 'hack' [and I don't like writing superfluous code]. However, it seems that there's no better approach in this case, so it will be fine. Thanks, Jonathan

Just a question on your discussion. Why not to let mpl::if choose an appropriate overload, like: //... somewher in c private: inline void functionality_impl()const/or non const { //make impl here } inline void f() { f::type>(); } template<class T> void f(); void f<ScopedCriticalSection>() { scs(cs); functionality_impl(); } void f() { functionlity_impl(); } Greetings, Ovanes

Nice idea! I'll do this instead. Thanks, Jonathan From: boost-users-bounces@lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Ovanes Markarian Sent: Tuesday, April 07, 2009 7:18 AM To: boost-users@lists.boost.org Subject: Re: [Boost-users] [mpl] newbie question about if_c et al actually, specialization and not overload should be correct. But idea remains the same. On Tue, Apr 7, 2009 at 4:13 PM, Ovanes Markarian mailto:om_boost@keywallet.com> wrote: Just a question on your discussion. Why not to let mpl::if choose an appropriate overload, like: //... somewher in c private: inline void functionality_impl()const/or non const { //make impl here } inline void f() { f::type>(); } template<class T> void f(); void f<ScopedCriticalSection>() { scs(cs); functionality_impl(); } void f() { functionlity_impl(); } Greetings, Ovanes