| Newsgroups |
gmane.comp.audio.supercollider.devel |
| Message-ID |
<[email protected]> |
On November 5, 2017 00:52:26 julian.rohrhuber-QYZGCWsIODmAF8UT6DzBU6xOck334EZe@public.gmane.org wrote:
> thank you for doing these tests, James. Much better than fading memory.
I learned something too -- I honestly thought ? was a normal method call. I
was surprised when it was performing about the same as the inlined
short-circuiting operator. The ControlOpcode explains the result.
Part of the point is to be cautious about making style recommendations for
the class library. If it were just a "huh, didn't expect that" curiosity, I
might have passed it by. When it started being recommended, then some more
evidence would be wise.
The really precise conclusion is, if it's a ? b (no method call):
- If you expect a to be nil most of the time, a ? b is, on average, very
slightly faster.
- If you expect a to have a value most of the time, a ?? { b } is slightly
faster.
- If it's 50-50, the speed differences cancel out.
... which is a lot for an average user to remember :D
hjh
Sent with AquaMail for Android
http://www.aqua-mail.com
_______________________________________________
sc-dev mailing list
info (subscription, etc.): http://www.birmingham.ac.uk/facilities/ea-studios/research/supercollider/mailinglist.aspx
archive: http://www.listarc.bham.ac.uk/marchives/sc-dev/
search: http://www.listarc.bham.ac.uk/lists/sc-dev/search/