Re: transform as a presentation attribute

> Presumably SVG 2 will adopt this from CSS 2.1, this will apply as well.

Has there been any discussion on adding an ability to alter the default DPI?

For instance, a new attribute/property "unit-dpi" which has a lacuna value
of "96 96",  Perhaps also with a special value "auto" in which the renderer
chooses a device DPI based on values obtained from the OS.

I know there have been questions on Stack Overflow from people with designs
and plans in SVG format that they wanted to print to-scale.

Paul




On 25 June 2014 19:56, Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de> wrote:

> The relation between the unit px and local units did not change in SVG.
> What was changed in CSS 2.1 is the relation between absolute units
> and px. There are different interpretations of the wording, but because
> absolute units are already defined by other organisations, 1px is fixed
> to 1/96 inch (with 1 inch = 2.54 cm as usual). But in typical viewers px
> and
> absolute units are presented rescaled somehow, therefore with CSS 2.1
> absolute units become useless in practice. Presumably SVG 2 will adopt
> this from CSS 2.1, this will apply as well. In practice viewers do it
> already
> for documents written in SVG 1.x (and fail therefore typically for
> absolutely
> scaled documents, if the screen does not have exactly a resolution of
> 96dpi -
> printers may do it better, but one has to take into account, that viewers
> can
> pass the obfuscation to the printer ;o)
>
> The other thing changed in the CSS transform drafts and in the SVG 2 draft
> is, that one can use units at all for transform, not just numbers with the
> meaning of local units.
> Therefore effectively, if one uses 100% for width and height for the root
> svg element and no units inside the document, the presentation remains
> the same.
>
> The joke to suggest to write transform="translate(200px, 200px)"
> instead of transform="translate (200 200)" has mainly the function to blow
> up the source code and ensures, that older/current viewers will fail to
> interpret this.
>
> Olaf
>
>