- From: Boris Dalstein via GitHub <sysbot+gh@w3.org>
- Date: Thu, 19 Dec 2019 00:33:27 +0000
- To: public-svg-issues@w3.org
Well, the "real problem" is not just the markers. There are are two somewhat compounding problems: 1. The markers 2. The fact that, as per SVG 1.1, segment-completing closepath operations are only defined for straight lines. I agree that it would be inconsistent to use segment-completing Z for sharp rects and not the other shapes. However, the approach in the current draft is far superior and future-proof in my opinion: using segment-completing Z for all shapes, including rounded rects and circles/ellipses. We just need to fix the minor mistake that sharp rects are not currently closed. In the future, if we want to properly fix the markers, we'll need segment-completing Z for all path segment types anyway, so it's good if equivalent paths already use them (even if the path data syntax don't support them yet). This is because the "number of markers" is a topological quantity, and any definition of a topological quantity requires a clean definition of topology, which requires segment-completing Z for all path segments. In particular, a fundamental truth is that open paths have N+1 vertices and N edges, while closed path have N vertices and N edges. You need segment-completing Z to capture this (skip the "N+1"th vertex for closed paths) and make an explicit, author-controllable difference between "topologically closed" and "geometrically closed", difference which matters a lot in the definition of markers. That's the two cents of an algebraic topologist anyway. -- GitHub Notification of comment by dalboris Please view or discuss this issue at https://github.com/w3c/svgwg/issues/753#issuecomment-567275102 using your GitHub account
Received on Thursday, 19 December 2019 00:33:29 UTC