Andrzej Krzemienski wrote
2014-05-26 11:08 GMT+02:00 Vladimir Batov <
vb.mail.247@
>:
... I do not see Andrzej confirming anything. He indeed asked a question... which a far from a confirmation of any sort. The way I see it
user-level optional <T> res = convert <T> (from, cnv);
the standard optimization/elision mechanism is to pass "res" to convert(). So, it becomes
convert(optional <T> & res, from, cnv);
I claim (without evidence in form of benchmarks) that returning optional <T> by value should not be slower than passing it as an input argument. Most compilers elide the moves/copies today, don't they? I would recommend making measurements before applying this optimization.
Andrzej, you are correct... and I am not suggesting to change the signature. The change from user-friendly optional<T> res = convert<T>(from, cnv); to convert(optional<T>& res, from, cnv); is done by the compiler as part of optimization/elision. That technique's been around for many years. I won't be surprised if optimization is far trickier than that these days. -- View this message in context: http://boost.2283326.n4.nabble.com/review-Convert-library-tp4662821p4662891.... Sent from the Boost - Dev mailing list archive at Nabble.com.