Re: Alternate syntax for required features.

While some aspects of the proposal are worth considering, I have to oppose
the element based proposal as it is outlined in the example in [1] at the
current time. The main problem with the proposed formulation is that it
necessitates formally defining a global attribute (in ttf: namespace) for
every feature. This means we will have to add definitions, with prose text,
xml syntax, and schema entries, for 123 attributes plus the 3 new proposed
element types (profile, features, extensions). This will also require tests
for each of these attributes. This would require a LOT of effort on the part
of the editor to accomplish this work, as well as on the group to develop
tests and verify these normative additions to the
DFXP language.

The current approach (based on requiredFeatures and requiredExtensions
attributes) is already well founded in existing SMIL and SVG practices,
etc., and supports conditional inclusion at a finer grained, which we could
easily add in the future by permitting these attributes to be used on
arbitrary DFXP elements.

Glenn

[1] http://lists.w3.org/Archives/Public/public-tt/2009Mar/0008.html


On 4/15/09 5:08 AM, "Sean Hayes" <Sean.Hayes@microsoft.com> wrote:

> I expect the meeting to be short (unless we see the new WD today or tomorrow).
>  
> The main decision point on the agenda is the features syntax. If you don¹t
> intend to join the meeting please indicate on the list your preference with
> respect to:
> a)     The attribute value proposal
> 
> b)      My element based proposal
> 
> c)      Something else.
> 
> d)     Don¹t care.