You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Meeting this Objective
EPUB Creators MUST provide synchronized audio playback via Media Overlays for all textual content.
This statement could be interpreted as saying to provide synchronized audio using the Media Overlays mechanism, and not to use another mechanism.
Since the main purpose of this clause is to say that the complete audio narration should also be provided along with the full text, we believe that "for all textual content" is a weak explanation and should be "for complete textual content".
Kudou-san is a member of the JDC technical committee.
Does the sentence quote by Kudou-san mandate the use of MO? I guess not but it can be interpreted so.
Here is my interpretation of "via Media Overlays all textual content"
If an element within the XHTML body element has child text content, this element MUST be referenced by some PAR element or MUST be a descendant of an element referenced by some PAR element.
Does the sentence quote by Kudou-san mandate the use of MO?
No, there is no requirement to use media overlays to meet the specification. And even if you use media overlays, you are not required to meet the objectives yet. We've kept that optional, as it says in the application section:
To maximize the effectiveness of Media Overlays for people with different reading needs, however, EPUB Creators are strongly encouraged to meet the OPTIONAL objectives defined in the next section.
Any text element within an SVG content document MUST be referenced by some PAR element or MUST be a descendant of an element referenced by some PAR element.
It becomes a big more convoluted than that, @murata2makoto. It is not a par element that refers to a text element, it is a text SMIL element within a par that does that (via a src attribute).
Maybe it is enough to replace the reference to par to a text, and not refer to par; the structure of a valid SMIL document would require that text to be part of a par anyway.
It becomes a big more convoluted than that, @murata2makoto. It is not a par element that refers to a text element, it is a text SMIL element within a par that does that (via a src attribute).
Stand corrected. My bad.
But how can we make sure that all text element within an SVG content document is covered by SMIL?
There's probably a larger question here about why we're referencing media overlays directly if we want this to be technology agnostic. If media overlays are replaced in a future epub, we have to rewrite all the objectives to account for the new technology.
The section should probably be "Synchronized Text-Audio Playback" and we should write out the specific references to media overlay documents and par/seq elements. That explanation belongs in the techniques.
Assuming we move in this direction, the one good thing is that the techniques don't go to CR so we don't have to have any new additions in place before we transition. We'd just need to clean up the specification.
Activity
murata2makoto commentedon Apr 5, 2022
Kudou-san is a member of the JDC technical committee.
Does the sentence quote by Kudou-san mandate the use of MO? I guess not but it can be interpreted so.
Here is my interpretation of "via Media Overlays all textual content"
mattgarrish commentedon Apr 5, 2022
No, there is no requirement to use media overlays to meet the specification. And even if you use media overlays, you are not required to meet the objectives yet. We've kept that optional, as it says in the application section:
mattgarrish commentedon Apr 5, 2022
It would have to account for SVG, too.
murata2makoto commentedon Apr 6, 2022
Any
textelement within an SVG content document MUST be referenced by some PAR element or MUST be a descendant of an element referenced by some PAR element.iherman commentedon Apr 6, 2022
It becomes a big more convoluted than that, @murata2makoto. It is not a
parelement that refers to a text element, it is atextSMIL element within aparthat does that (via asrcattribute).Maybe it is enough to replace the reference to
parto atext, and not refer topar; the structure of a valid SMIL document would require thattextto be part of aparanyway.murata2makoto commentedon Apr 6, 2022
Stand corrected. My bad.
But how can we make sure that all text element within an SVG content document is covered by SMIL?
mattgarrish commentedon Apr 6, 2022
There's probably a larger question here about why we're referencing media overlays directly if we want this to be technology agnostic. If media overlays are replaced in a future epub, we have to rewrite all the objectives to account for the new technology.
The section should probably be "Synchronized Text-Audio Playback" and we should write out the specific references to media overlay documents and par/seq elements. That explanation belongs in the techniques.
mattgarrish commentedon Apr 6, 2022
Assuming we move in this direction, the one good thing is that the techniques don't go to CR so we don't have to have any new additions in place before we transition. We'd just need to clean up the specification.
add change log entry to core spec for #2221
Merge pull request #2249 from w3c/editorial/add-cl-entry