Re: [Paths] Catmull-Rom and other new path commands
Dirk Schulze <[email protected]>
| Newsgroups | gmane.comp.web.svg |
|---|---|
| Message-ID | <[email protected]> |
> On Jul 31, 2015, at 2:28 AM, Dean Jackson <[email protected]> wrote: > > I'm slightly concerned about adding new path commands, such as Catmull-Rom. I have nothing against the command itself, but I don't like that there is no nice fallback mechanism. > > How am I supposed to author content that works in an engine that supports this feature, but also works in an engine that doesn't? > > Maybe all new commands should be in a d2 attribute? Yuck. If you use the style approach, at least you could have multiple "d" properties. > > Of course, as soon as you're providing fallback, it makes the argument for providing the new feature less compelling. An alternate proposal is to support a d CSS property and have different path syntaxes separated by functions. Exactly like the clip-path property does today with basic shapes: d: “oldPathSyntaxString" d: path(“oldPathSyntaxString") d: catmull-rom(catmull-romSyntax) d: polygon(setOfCoordinates) d: …. I agree with your general concern and brought up the same issues before. For this and other reasons we develop the new path syntax in a new module to not block SVG2. > > Dean > > PS. And if we're going to break the syntax, why didn't we add an arc command that mortals can understand? :) We discussed to align with Canvas Path API and add new segments but we simply run out of meaningful letters ;) Greetings, Dirk