Re: Error and omission in SVG spec

Hi Ben,

At 04:20 PM 7/14/99 +0100, Ben Hutchings wrote:
>I'm currently reading through and digesting the SVG spec (1999/7/6
>It does not appear to specify which corner of a viewport is the origin. 
>Is it the top left (typical for computer graphics systems) or the bottom
>left (normal for Cartesian co-ordinates)?  Or is there a reason why it
>should not be specified?

After lengthy and heated debate, the SVG WG decided that the initial origin
is at the top left, with the y-axis pointing down. Thus, our coordinate
system has the same orientation as HTML/CSS. (We will fix the spec to make
this clearer.)

Because of the y-axis pointing down, with the initial coordinate system,
angles go the opposite direction from what we were taught in geometry
class. Thus, with the initial coordinate system, positive angles go
clockwise. (The spec will include pictures showing this.)

Text appears right-side up in the initial coordinate system. (Non-graphics
people will say, "but of course!". However, graphics folks familiar with
rendering models such as PostScript recognize that there needs to be some
minor auto-magical behavior in SVG viewers to make this happen, which has
the side effect that if you reflect the coordinate system to cause the
y-axis to point up, your text will therefore appear upside down.)

>I also noticed a minor error in section 8.5, which says:
>"If <sy> is not provided, it is assumed to be equal to <sy>."
>The second instance of "<sy>" should presumably say "<sx>".

Correct. The next spec will fix this.


Jon Ferraiolo
SVG spec editor
Adobe Systems Incorporated

Received on Wednesday, 14 July 1999 12:19:43 UTC