Re: Rename 'd' property

Hi, folks–

I'm supportive of renaming it as a property.

I'd suggest 'path-data', to give it the clarity of context.

There is no issue of backwards compatibility; the option of setting the 
path data via CSS doesn't exist in any current or previous version of 
SVG, so there's nothing to be incompatible with. There is the issue of 
name correspondence between the attribute and property, for the sake of 
the author; if aliasing is allowed in CSS today, then we could have both 
'd' and 'path-data' (or whatever).

The issue of style-vs-content is largely irrelevant, since there's no 
software that makes a distinction; the most pressing issue there is not 
the semantics, but the ease and consistency of API access to the path 
data, regardless of where it's defined, and the consistency of 
bounding-box results.

Regards–
Doug

On 2/9/16 9:32 AM, Sebastian Zartner wrote:
> On 9 February 2016 at 11:27, Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de
> <mailto:Dr.O.Hoffmann@gmx.de>> wrote:
>
>     Hello,
>
>     the core disadvantage for renaming something in SVG is, that it is
>     backwards
>     incompatible to already published viewers and existing content.
>     For SVG 2 there is a requirement to avoid any backwards
>     incompatibilities.
>
>
> That's clear. My initial request for renaming only targets the CSS
> property, which is, as far as I know, not implemented yet.
> In case of wanting both the CSS property and the attribute consistently
> named, only aliasing would be a solution.
>
>     And as we all know - there is a certain danger, if implementor get a
>     choice to
>     do something, this typically results in trouble for authors and the
>     audience
>     ;-)
>
>
> There wouldn't be a choice for the implementors. They would have to
> support both. The choice would be at the authors.
>
>     'd' is a nice name both for an attribute and a property.
>
>
> So, we obviously have different opinions on this. Again, my argument is
> that 'd' is inexpressive, especially when it is defined somewhere else
> than at the place where it is used, i.e. as CSS property in a stylesheet
> instead of a tag attribute.
>
>     However, it is questionable to define it as a property, just because
>     in almost
>     any case the information in it is content and not just decoration.
>
>
> That is a good point, though note that the reason for defining it as CSS
> property is "because it allows non-SMIL declarative animation of shape
> morphing, using CSS animation syntax"[1] and letting it work together
> with Web Animations[2].
>
> Sebastian
>
> [1] https://lists.w3.org/Archives/Public/www-svg/2016Feb/0005.html
> [2]
> https://www.w3.org/2016/02/03-svg-irc.html#T23-18-02<https://lists.w3.org/Archives/Public/www-svg/2016Feb/0005.html>

Received on Wednesday, 10 February 2016 00:31:22 UTC