Re: [svgwg] Rect decomposition is invalid (#753)

I would personally never want to use (and even less implement) a marker algorithm that automatically removes duplicate markers caused by zero-length segments. Such algorithms are subject to implementation discrepencies due to floating point rounding errors, and also subject to animation flickers (a zero-length segment becoming non-zero length). Also, again for animation, they make implementations harder to optimize due to the possibility of non-constant number of markers during animation. Again, the number of markers is a topological quantity: it shouldn't depend on geometrical attributes such as vertex positions.

If there is a need for such "auto-removal of markers in case of zero-length", it's only because path data is imperfect in the first place, and forces you to use zero-length segments for non-linear closed paths. The solution is to improve the path data syntax to allow defining such paths without zero-length segments, not to complicate markers even more. Segment-completing Z are trivial to implement, easy to understand for author, give more compact file size, and make more sense theoretically. There's nothing not to love about them.

-- 
GitHub Notification of comment by dalboris
Please view or discuss this issue at https://github.com/w3c/svgwg/issues/753#issuecomment-567502263 using your GitHub account

Received on Thursday, 19 December 2019 14:03:49 UTC