W3C home > Mailing lists > Public > www-svg@w3.org > January 2010

Re: Fwd: Question on gradient userSpaceOnUse

From: Dirk Schulze <vbs85@gmx.de>
Date: Tue, 05 Jan 2010 17:31:06 +0100
To: Jeff Schiller <codedread@gmail.com>
Cc: Erik Dahlstrom <ed@opera.com>, www-svg <www-svg@w3.org>, Alex Danilo <alex@abbra.com>
Message-ID: <1262709066.25524.54.camel@dirk-laptop>
Why should there be a difference, if you move the scal from the rect to
the g?

If you do that, the following will happen:

The rect has no transformation, so the rect user space is not
transformed. Now you apply the gradient. It is drawn in the rect user
space (no transformation).
Now you scale the whole rect with the gradient (because of the scale in
g). The result is the same.

-Dirk

Am Dienstag, den 05.01.2010, 07:45 -0600 schrieb Jeff Schiller:
> I should have prepended my statement with "if you think the transform
> attributes have an effect on user space then..."
> 
> It also seems odd to me that a transform attribute with a scale should
> have the ability to "reach" in and change the gradient vector's
> endpoint in that way 
> 
> On Jan 5, 2010 7:32 AM, "Jeff Schiller" <codedread@gmail.com> wrote:
> 
> 
> 
> The fact that you moved the scale to a parent g with no differences is
> weird to me.  Note that there is already a translate of 100,100 on a
> parent element that has NO apparent effect on the rects user space
> 
> > On Jan 5, 2010 4:08 AM, "Erik Dahlstrom" <ed@opera.com> wrote:
> > 
> > 
> > 
> > On Tue, 05 Jan 2010 09:18:17 +0100, Alex Danilo <alex@abbra.com>
> > wrote: > --Original Message--: >> ...
> > 
> > ...
> > 
> > >> >> The spec says userSpaceOnUse "shall represent values in the
> > coordinate >> system that results...
> > 
> > > > Just to test for consistency I also verified that moving the
> > transform attribute from the rect ...
> > 
> > <?xml version="1.0"?> <svg xmlns="http://www.w3.org/2000/svg"
> > xmlns:xlink="http://www.w3.org/1...
> > 
> >    <g transform="scale(0.25,1)">      <rect width="400" height="50"
> > fill="url(#g1)" />      <!-- sh...
> > 
> > > >      <line x1="0" y1="0" x2="400" y2="50" stroke="black"/> >
> >  </g> > >    <rect y="100" width...
> > 
> > > There is clarification for the weird perpendicular behaviour of >
> > the gradients for 'objectBound...
> > 
> > > > I agree that it makes sense to make the gradient perpendicular
> > to the gradient vector for both ...
> > 
> > /Erik -- Erik Dahlstrom, Core Technology Developer, Opera Software
> > Co-Chair, W3C SVG Working Grou...
> > 
> 
Received on Tuesday, 5 January 2010 16:31:42 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:44 GMT