W3C home > Mailing lists > Public > public-fx@w3.org > April to June 2014

[geometry] DOMMatrix with DOMString argument

From: Dirk Schulze <dschulze@adobe.com>
Date: Fri, 23 May 2014 21:27:03 +0000
To: FX <public-fx@w3.org>, www-style <www-style@w3.org>
Message-ID: <15153935-DF83-4815-9510-5762D4895C1C@adobe.com>
Hi,

Geometry Interfaces defines a constructor for DOMMatrix that takes a DOMString as argument[1]. The DOMString may be a <transform-list> as defined by CSS Transforms[2].

DOMMatrix does not have any context. It is independent of a Document or Element. However, DOMString can have transform function values with relative or percentage units.

	“translate(2em, 3rem) translateX(50%) translateY(10ex)”

These values can not be resolved without the context of an Element.

A possibility could be to pass a second optional Element argument. DOMMatrix would use the context of a passed Element to resolve these values.

	Constructor(DOMString, Element?)

However, SVG transform attributes traditionally don’t have any value units at all and are context unbound. That is why I would like to have the Element argument optional.

So we would still need a fallback behavior if no Element was passed. I am also not sure what happens it Element is not attached to a document. How can the viewport or relative values be resolved then?

To summarize:
1) Shall we support Element as an optional argument to provide a context for the parser? What about unattached Elements?
2) If no Element was passed, what is the fallback for relative of percentage values?

Greetings,
Dirk

[1] http://dev.w3.org/fxtf/geometry/#dom-dommatrix-dommatrix
[2] http://dev.w3.org/csswg/css-transforms/#typedef-transform-list
Received on Friday, 23 May 2014 21:27:40 UTC

This archive was generated by hypermail 2.3.1 : Monday, 22 June 2015 03:33:52 UTC