Re:IBM says Swift is ready for the enterprise
by mpweiher on Monday September 26, 2016 @09:24AM
(#12581564)
It's not really predictable enough for a systems language.
[ Parent
| on HN ]
Re:IBM says Swift is ready for the enterprise
by pjmlp on Monday September 26, 2016 @10:09AM
(#12581905)
Can you predict what your C compiler with going to do with your code, taking into account UB and compiler specific implementation behaviors not specified by ANSI C?
Re:IBM says Swift is ready for the enterprise
by mpweiher on Monday September 26, 2016 @10:19AM
(#12581979)
More than with Swift. I routinely see 100x - 1000x and more performance difference between -O0 and -O in Swift. Considering that the optimiser doesn't give warnings or errors if it can't apply optimisations, that's out of bounds for me for a systems programming language. YMMV.
The whole UB idiocy is a different matter, though related because it's perpetrated by roughly the same group of people, for similar nonsensical and non-validated reasons. See my post http://blog.metaobject.com/2014/04/cc-osmartass.html
See also: http://www.complang.tuwien.ac.at/kps2015/proceedings/KPS_201... and Proebsting's Law.
Re:IBM says Swift is ready for the enterprise
by pjmlp on Monday September 26, 2016 @10:34AM
(#12582133)
> More than with Swift. I routinely see 100x - 1000x and more performance difference between -O0 and -O in Swift. Considering that the optimiser doesn't give warnings or errors if it can't apply optimisations, that's out of bounds for me for a systems programming language. YMMV.
Improvements required for the toolchain in a young language, than Apple, IBM and others will certainly improve.