- From: Robin Berjon <robin.berjon@expway.fr>
- Date: Thu, 10 Nov 2005 17:59:08 +0100
- To: Al Gilman <Alfred.S.Gilman@IEEE.org>, wai-liaison@w3.org
- Cc: www-svg@w3.org
Dear PFWG & WAI, thank you very much for your comments. Here are the SVG WG's responses. > This statement seems to indicate that UAAG guidelines are > still under development. It is not clear what "Once the > guidelines are completed, a future version of this > specification" is refering to UAAG or SVG. The statement has been corrected and our conformance section now requires conformance to UAAG priority one guidelines. > While the blanket references in Appendix D are necessary > for completeness, the readers of this document will understand > what you mean in various inline remarks about what processors > should do if the pertinent sections of the UAAG10 are > directly referenced at key points in the text. We have added multiple references to UAAG throughout the text, wherever we could find a section where it would be applicable. > We are not satisfied that this document yet meets > the requirement to demonstrate accessible usage > in its examples. Since examples are not normative, the SVG WG believes that they can be added during the CR phase, as editorial improvements. We would very much like to work with you to define a set of examples demonstrating accessible usage better than is currently done. We have taken some of your example suggestions into account but would like to start work with you to push this further. > Could not find where in the document it discusses user disabling play > of audios and videos. No matter, this provision where mentioned > should apply to the inhibition of animation as well and cite UAAG10 > Checkpoint 3.2 at: > http://www.w3.org/TR/UAAG10/guidelines.html#tech-configure-multimedia The accessibility chapter specifies that the UA should provide means of disabling animations, audio, and video through multiple modalities. We have modified our draft to make this point clearer. > Examples in general use mouse-specific events. This should be > upgraded in line with the intent-based event philosophy developed > by the HTML WG. Again, we would like to identify during CR which examples would constitute good targets for improvements and work with you to make them less pointer-specific, while retaining the same functionality. Please note that where such events are discussed, they are pointer events and not mouse events. > Examples use an artifical arbitrary namespace. > While the point is there to be made that SVG > does not limit the markup used here, the > preponderance of example should use realistic > namespaces such as the mobile profile of XHTML. Usage of arbitrary namespaces in SVG content is not only realistic but common. In fact, should we use an example based on XHTML we believe it would give the impression that the markup is only open to W3C specifications, while in fact we intend to get the notion across that authors not only can, but in fact should, use their own vocabularies when they need to. > <draft class="changeTo" > > 1.6 shape. A graphics element that comprises a defined combination of > straight lines and curves. Specifically any instance of the element > types: 'path', 'rect', 'circle', 'ellipse', 'line', 'polyline', 'polygon'. > </draft> We have applied your suggested change. > 1.6 User agents include graphical desktop browsers, multimedia > players, text browsers, voice browsers; used alone or > in conjunction with assistive technologies > such as screen readers, screen magnifiers, speech synthesizers, > onscreen keyboards, and voice input software [UAAG10]. We have applied this as well. > 8.3 path data > For semantic entities such as flowlines in flowcharts, how should > user information such as 'title' and 'desc' be associated with > complex paths? > How would you suggest that we get authors to observe this practice? It is not possible in SVG Tiny 1.2 to attach metadata to subcomponents of a path, though this may become possible in future versions of SVG. In the meantime we recommend that the path be separated into smaller path elements, each with their own metadata, knowing that they can easily have the same rendering. We intend to add such considerations to our authoring guidelines. > It is not clear how a User Agent or Assistive Technology is to > associate text that it finds with the structure of the scene > in the graphic depiction. > > Diagrams often embed labels, but how is this semantic > relationship (of the label text to the conceptual object > labeled) to be recognized from the SVG representation? The way of indicating that sort of semantic relationship is by putting the related items in the same group element (<g>), which can then contain text and metadata. Again, we intend to place a discussion of this in our authoring guidelines. > How would authors use this notation to identify alternatives > for audio and video objects (as you see this notation being used)? The switch element and its usage is described in greater detail in the chapter on structure. It can be used to identify a sequence of alternative SVG fragments (which may in fact be completely different in content) based on features that indicate what is used in their content (e.g. video, audio, etc.) > The examples read as if all delivery contexts support pointer events. > The provision of alternatives to pointer events should be illustrated > in the examples. We have expanded our text indicating that pointers may not always be available and that authors should use device independent events. However since this is a section on how to control pointer events we have maintained the example. > To what extent does this specification support a two-step identification of > user event handling with an intent-based event defined first and then > bound in an over-rideable way to a device-specific UI event? SVG does not define its event handling to be different from the other interaction vocabularies, and only has access to the facilities defined in DOM 3 Events and XML Events. > If an element is animated, is there a way to provide > a static alternative other than the inital state of the > animation? We have added the snapshotTime attribute which can indicate the time in an animation that can be used to create a snapshot. Also, the switch element may be used to provide alternatives to animated content. > How can an AT perceive that a given animated > whatever is in the state of ongoing animation? Assuming the given assistive technology has access to the MicroDOM, then it can know that something is being animated by catching SMIL timing events (begin, end, repeat animation), and can access the animated value through the trait system. Thank you deeply for your excellent comments, please let us know within two weeks if they do not satisfy you, -- Robin Berjon Senior Research Scientist Expway, http://expway.com/
Received on Thursday, 10 November 2005 16:59:16 UTC