W3C home > Mailing lists > Public > www-svg@w3.org > July 2011

Re: stroke=gradient for lines or horizontal paths

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 22 Jul 2011 18:00:04 -0700
Message-ID: <CAAWBYDBkgJ-iSfYdToC-5Vt1rK4hVsyZQm-6cDMG8rg+LANWxA@mail.gmail.com>
To: www-svg <www-svg@w3.org>
(Apologies for not chiming in earlier here - for some reason I never
got subscribed to www-svg when I joined the WG.  This has been

I strongly agree that gradients should paint in 0width or 0height
bounding boxes.  The gradient is still conceptually well-defined in
these cases.

For linear gradients with spreadmethod=pad, if the gradient is aligned
exactly opposite the bbox, it's just 0-length, and thus defines a
two-color image (the color of the first <stop> and the color of the
last <stop>).  Otherwise, it's identical to if the gradient was
aligned with the bbox and was the length of the bbox.

For radial gradients with spreadmethod=pad, it always defines a
solid-color image with the color of the last <stop>.

spreadmethod=reflect or repeat are more difficult, though, because
there is no first or last <stop>, conceptually.  Right now, CSS Images
punts on the issue and requires repeating gradients to clamp their
width to a minimum of 1px.  I'm open to better solutions, though
(weighted average?), and would like to ensure that CSS gradients align
with SVG gradients here.

Received on Saturday, 23 July 2011 01:00:50 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:29:48 UTC