Issue 9 Possible Fix (Box-to-Self implementation)

[olkivan]

Hi, Steve, could you please take a look at my fix for the Box-to-itself case? Here is an algorithm and a sample component which utilizes the algo. Could you please provide your opinion, might it solve the problem? I haven't provided the test cases yet, until you confirm it's the right way to go. I'd appreciate any feedback. Thanks!

[steveruizok]

Hey, I'll take a look at this over the weekend!

[olkivan]

Great! Looking forward to your feedback!

[steveruizok]

Hey Ivan. Thanks for the PR, I've taken a look at it now. I'd like to bring its API into line with the other functions in the library (which are unusual, I know).

I'm also considering applying some of what I learned while working on this project, both to this problem at to some of the other arrows, too. That would be a bigger change though—moving from quadratic curves to ellipse segments, as you're doing here—and might be better for a 1.0.0. release.

So I need to make some decisions before I can get it ready to merge. Until then, your implementation is pretty well isolated so you should be able to use it on the side if it's an urgent need for a project. Thanks again :)

[olkivan]

Steve, maybe it's a good idea to have both options - to use quadratic curves or to use ellipse segments (if it even possible). Anyway thanks for the feedback and I'll be waiting for the 1.0.0 release :)