Wrong optimizations for 64 bit CPUs

Question

Wrong optimizations for 64 bit CPUs

Closed this issue 10 years ago · 9 comments

On 64 bit CPUs, some functions return invalid values due to the optimization that seems to be specifically made for 32 bit. CGVectorMultiplyByScalar for example.

samgreen commented 10 years ago

Fixed!

Answer 1 · 2015-01-27T22:12:52.000Z

Can you elaborate on this for me? Would love to fix it, also I will happily accept a pull request.

Answer 2 · 2015-01-27T22:30:05.000Z

@angelog Found the problem. It's due to CGFloat changing types of 64bit architecture. Resolving the issue now, will release ASAP.

Answer 3 · 2015-01-28T00:19:11.000Z

Awesome! 👍

Answer 4 · 2015-01-28T08:24:35.000Z

This shows an empty change, are you sure you pushed it correctly?

Answer 5 · 2015-01-28T16:45:13.000Z

Doh! Thanks noticing. Will repush in an hour or so.

Answer 6 · 2015-01-28T19:12:56.000Z

@angelog Thanks for your patience on this. The header file and Readme are now updated.

I would love any additional feedback you have on this library, it's worked well for me in a few games, but I'm always looking to improve it.

Answer 7 · 2015-01-28T19:50:55.000Z

Awesome, thanks! We're using it for a game we'll be releasing on the app store soonish.

Personally I've added a function CGVectorFromPoint (which converts a CGPoint to a CGVector), but perhaps this is already implemented by Apple somewhere that I've missed.

It's been very helpful so far!

Answer 8 · 2015-01-28T21:39:44.000Z

@angelog Please submit a pull request with your additional features, I'd love to merge them in!