On Sep 28, 2012, at 10:38 PM, Rik Cabanier wrote: > That's an oversight. Neither should be allowed to be negative. > > Rik > > On Fri, Sep 28, 2012 at 5:55 PM, Dirk Schulze wrote: > Hi, > > Reading the current draft, there seems to be an inconsistency. While 'r' is not allowed to be negative, there seems to be no restriction for 'fr' [1]. Is that the intention? As a note, Canvas gradient does not allow to be any radius negative[2]. > > Greetings, > Dirk > > [1] https://svgwg.org/svg2-draft/pservers.html#RadialGradientElementFRAttribute > [2] http://dev.w3.org/html5/2dcontext/#dom-context-2d-createradialgradient > > On Sep 4, 2012, at 11:11 AM, Jasper van de Gronde wrote: > > > On 2012-09-04 18:06, Rik Cabanier wrote: > >> ... We also need to call out what happens with alpha once the circles > >> start overlapping. > >> I think use expectation is that the gradient does not interact with > >> itself. > > Good point. This also has bearing on negative radii. If a future spec > > would allow negative radii it probably should not let "overlapped" > > circles shine through (if it would, then it would clash with the current > > interpretation, as the "bottom" half of the double cone would need to be > > visible). > > > >