Open tabatkins opened 1 year ago
I have a proposal to fix this in https://github.com/w3c/fxtf-drafts/issues/504#issuecomment-1458831844
Letting shape()
cover this use-case is also OK to me, so we don't have to change anything to path()
(e.g. add at <position>
to its syntax).
Note that shape()
always has the starting point (i.e, initial position), per its current spec.
Do we want to distinguish between a path that has an explicitly given initial move-to, and let ones without such a move-to use the 'offset-position' as their initial position?
Per SVG1 (https://www.w3.org/TR/SVG11/paths.html#PathDataBNF), and SVG2 (https://www.w3.org/TR/SVG/paths.html#PathDataBNF), it seems the syntax of svg path should always have the initial move_to
command (if it's not an empty string).
So I expect path()
should always specify the initial position, by the initial move_to
command. That means it should be fine to let path()
just ignore offset-position
because the initial position should always be there, per its syntax. (This may be the same as url()
because I expect the SVG geometry elements always provide the initial position.)
Currently, 'offset-path' defines that the "initial position" of a path() function is given by the first move-to in the path().
This makes complete sense to me when there is an initial move-to in the path string. But this is pretty limiting; you can't use any of the CSS units or percentages there.
Do we want to distinguish between a path that has an explicitly given initial move-to, and let ones without such a move-to use the 'offset-position' as their initial position? Or should we just let
shape()
cover this use-case instead, since it can specify its initial move-to using CSS units?