Re: Subsettability of the SVG document

Hi Sairus,

I am not aware of solutions that subset SVG in a way similar to what you are describing for font subseting. Subsetting glyphs that are not in the subset list should be a matter of taking out all the non-matching ids. To remove animation elements, since I believe we are talking about declarative animation, I think we could just remove the all animation elements.

Kind regards,
Vincent.

From: Sairus Patel <sppatel@adobe.com<mailto:sppatel@adobe.com>>
Date: Tue, 3 Jan 2012 14:15:10 -0800
To: "public-svgopentype@w3.org<mailto:public-svgopentype@w3.org>" <public-svgopentype@w3.org<mailto:public-svgopentype@w3.org>>
Subject: RE: Subsettability of the SVG document

Hello all,

I haven't heard from anyone, so let me highlight a particular question from my message below as a starting point for discussion:

Is this sort of subsetting of an SVG document something that's done today for other reasons? Or would be this a new thing?

I'm wondering about the state of tooling in this area. Ease/practicality of subsetting is an important consideration in the new format we're designing.

If you think I should ask this on www-svg@w3.org<mailto:www-svg@w3.org> I'd be happy to -- let me know.

Sairus

-----Original Message-----
From: Sairus Patel [mailto:sppatel@adobe.com]
Sent: Thursday, December 01, 2011 10:54 AM
To: public-svgopentype@w3.org<mailto:public-svgopentype@w3.org>
Subject: Subsettability of the SVG document

For SVG OT, we've been talking about a define-it-as-you-will SVG document that simply has to have elements that can be identified in a standard way by the font engine as glyphs (e.g. id="g<glyphID>").

Now, OT font engines often need to subset a font before embedding it or streaming it for other purposes.

So the task of subsetting an SVG OT font will involve extracting only the required element id's from the SVG document, and possibly only the animated or static representations of those id's (e.g. if I'm embedding a font into a document format that doesn't support animation, I'd want to jettison any animation-related matter from the font). The rest of the subsetting task (e.g. subsetting the GSUB and other tables) will be the same as for TT or CFF OT fonts.

Is this sort of subsetting of an SVG document something that's done today for other reasons? Or would be this a new thing?

So far, we haven't talked about imposing any additional structure within the SVG document. This provides maximum flexibility for the vendor to organize their glyph descriptions, but it does increase the complexity of subsetting.

One option would be to simplify things and require a simple list of <g> elements, one after the other. But this already starts going down the path of defining some new kind of SVG <font>-like format, and I don't know if it would be worth it.

Thoughts?

Sairus