Re: SVG2: wrong definition for bounding box algorithm

Hello,

the section about bounding box units mentions objects without width
and height, excluded to have a meaningful interpretation of bounding
box units, this includes for example horizontal and vertical lines (including
obviously points).

On the other hand, if those objects have a painted stroke, the presentation
becomes visible.

Because the SVG 2 draft 7.11 mentions tree types of bounding boxes,
maybe this should depend on the type:

'object bounding box' obviously requires a non zero area to become visible,
'stroke bounding box' and 'decorated bounding box' seem to be more related
to the presentation, not the core content, therefore for them such objects 
without object bounding box are meaningful, if they are presented with non
zero size.

That the origin (0,0) has such a specific or emphasized meaning
for the determination of the boundingBox is quite surprising, what makes it
different from for example (121,-137)? The algorithm should be explained 
invariant from the currently chosen coordinate system.
The related text in '7.12. Object bounding box units' is better understandable
to determine the bounding box than the new text in 7.11, and not sure, why
there is a need for another text about this issue.

Will there be an attempt to make the 'stroke bounding box' and 'decorated 
bounding box' available in a similar way as the objectBoundingBox for such
features, the objectBoundingBox units can be used? This could help for
those features like gradients and pattern etc in some use cases, for 
example if the gradient or pattern is applied to the stroke, not the fill.
Else it is not obvious anyway why to define 'stroke bounding box' 
and 'decorated bounding box' and what is written in 7.12. 
(as in SVG 1.1 and SVG tiny 1.2) should be sufficient.


Olaf

Received on Saturday, 19 April 2014 16:39:49 UTC