Some comments on <image>

Hi folks,

I'm in the midst of developing some on-line instructional material for W3C, so have had occasion, after some years of blissful vacation from it, to reopen the SVG Spec document(s). What I've been looking at mainly is Scalable Vector Graphics (SVG) 1.1 (Second Edition), but not knowing of the publication timetables for various drafts, perhaps my questions would be better addressed toward 2.0 or 1.2 or some tiny version. It is not my wish to quicken anyone's heartbeat nor forestall any last calls or anything, so please move these comments into an appropriate comment-bin if need be. 

1. The spec says "Conforming SVG viewers need to support at least PNG, JPEG and SVG format files." Why not GIF? I recall a profound nervousness that spread like squid ink through the open source community [1] 10 or 12 years ago as the holders of the GIF patent threatened to go after those who used it without license. I believe, however, that the patent has since expired. [2] A search of gif in Google images shows about a billion files with close to half that number for PNG. In many cases GIF files are smaller than PNG files, I think, and lots of the older public domain imagery sites on the web used gif because, well, PNG wasn't available then. All the browsers I know of go ahead and support GIF anyhow, but it is one thing we can be certain of that no longer has patent entanglements. PNG?? Who can ever be completely sure until the 20 years pass?

2. Maybe there is something there that I am missing, but since <image> is (most?) frequently used to bring bitmaps into SVG would it make sense to have something like an alt attribute for accessibility? The SVG Tiny 1.2 accessibility  document [3] says "When a graphic does not include explanatory text content, it requires a text equivalent. If the equivalent is complex, use the 'desc' element, otherwise use the 'title' child element." while the 2000 document [4] on SVG accessibility gives examples in which graphical elements and descriptors (title and desc) are affiliated via common membership in a <g> tag.

I suppose that <image><title>words</title><desc>more words</desc></image> affiliates descriptors with images better than the examples given: <g><title>words</title><desc>more words</desc><image /></g>, and I surmise that that is the intention of the statement in 1.2 Tiny? But shouldn't this perhaps be a bit clearer in the spec itself when <image> is discussed? 

3. In some viewers <image height="200" width="300"> works fine to convey the authors intention to resize an image to fit a particular rectangle, but in most one has to say <image height="200" width="300" preserveAspectRatio="none">. A careful reading of the spec confirms that an author must override the browser's wish to override the author's intent here, but if the author wished to preserve aspect ratio by default, would she not just say <image height="200" > and let the browser's desire to preserve the aspect ratio kick in? If an author goes to the trouble of specifying both height and width, why would we assume that she must be mistaken?

I think I may have successfully subscribed to this mailing list, so you may hear from me a bit more on issues of specifications. Please forgive me if these issues have already been discussed but my ability to find discussion threads in W3C WG archives is a bit lacking. 

regards
David


[1] http://www.gnu.org/philosophy/gif.html
[2] http://drupal.org/node/8727 or http://en.wikipedia.org/wiki/Graphics_Interchange_Format 
[3] http://www.w3.org/TR/SVGTiny12/access.html 
[4] http://www.w3.org/TR/SVG-access/