Accessible Platform Architectures Working Group Teleconference

28 Feb 2018


janina, Joanmarie_Diggs, JF, pal, IanPouncey, Nigel, Becka11y, MichaelC, marisa_demeglio
Leonie, Gottfried, david_singer



<janina> scribe: jasonjgw

Agenda Overview; Introductions go around

Janina: the main agendum is to discuss issues presented to TTML and WebVTT, and attempt to move closer to resolution than has been achieved via e-mail.
... beyond that conversation, the usual agenda apply.

Nigel: notes that he chairs the TTML work and that WebVTT is separate. Nigel notes issues with timing requirements for captions/subtitles.

<janina> We can add Styling below the Captioning issue as a topic

Styling issues with captions and CSS.

JF: there are cofnlicting user needs and conflicting understandings of what we need and the optimal user experience.
... he is concerned that there be nothing in the caption-related specs that prevents users from resizing/presenting the caption differently according to their needs.

Janina: would like to continue collecting topics before entering into the details. Styling issues have been identified.

Participants introduce themselves.

APA Captioning Comments -- Discussion with TTML and WebVTT

<JF> examploe here: http://john.foliot.ca/demos/captions.png

Janina proposes that APA concerns could be easily addressed. Normative changes don't appear to be necessary and an explanatory note (clarifying that the author's supplied presentation of captions isn't set in stone) may be sufficient.

<JF> in practice here: https://www.3playmedia.com/solutions/features/integrations/brightcove-integration/

Janina: anyone who needs significant magnification of captions is likely to ahve a strategy for dealing with the consequences of doing so.
... there are important, but marginal, use cases: user needs supersede default authoring decisions - the author proposes, the user disposes. Most consumers will receive what the author has created, but there are edge cases. JF has offered an example of a tool demonstrating such cases.

Nigel: TTML allows the author to define presentaton of captions, but it is open to implementations to present captions otherwise than as the author has requested. He notes that users can enlarge the entire video or present it on a larger display.

<JF> https://www.w3.org/WAI/PF/media-a11y-reqs/

JF: He notes the concern that captions could completely obscure the content in the viewport. Media Accessibility User Requirements discussions envisage possible reliance on a second screen that presents the video in parallel with the primary display, but with the captions enlarged. He argues that preserving visibility of video content in the viewport isn't a critical requirement.

<JF> http://john.foliot.ca/demos/captions.png

<JF> https://www.3playmedia.com/solutions/features/integrations/brightcove-integration/

JF: refers to a caption increased to 400%, achievable (with other modifications) by adjusting options in the video player.

WebVTT and TTML should not specify anything that frustrates these user adjustments.

Janina proposes an accessibility impact statement that could note the issues.

Nigel notes that the user needs to be able to obtain equivalents of both the audio track and the video track in order to understand the program content.

For example, the user needs to be able to see that there's something important in the video while reading captions.

<JF> clarification around the 2/3 viewport: I have been told this anecdotally

pal: there's nothing in the TTML spec that prevents the viewing system from transforming the video itself or the caption text.

Presentation systems can display teh timed text on a separate device, increase the size, etc. It isn't clear that any changes are needed.

Andreas: it's a question of responsibility - of where the customization task should be handled. There is no layer to negotiate subtitle customizations.

Typically the customizations are performed by the application.

JF: part of the cdore concern is wehre the customization happens and who/what is responsible. While agreeing with Nigel that the video and audio complement each other to comprise the entire presentation, there are users who need to access it in ways that don't respect this totality (e.g., viewing the video more than once).

He reiterates the question of who's responsible for customizations and how they will occur.

Janina: doesn't notice any substantive disagreement - timed text specialist appear to acknowledge the principle that users may need to customize the timed text to ensure it's usable by them.

Janina would expect users requiring significant enlargement to have strategies for working with it (additional monitors, large displays, etc.).

Janina: perhaps we need to focus on the particular language in the TTML spec which may appear to be inconsistent with this principle (of users being able to adjust the presentation as needed for accessibility purposes).

Janina suggests focusing on this specific language and that an accessibility impact statement may be appropriate.

Responding to a question, John states that he can't find any specific language in the specification, and that the concern arises from comments received in response to the horizontal review. He wishes it to be clarified that the use cases under consideration are matters that we want to support.

JF: notes that technologies often arise for review by APA without sufficient documentation of their intent.
... suggests we need clarification that nothing prevents, for example, increasing caption size to 400%.

<JF> @Pierre: https://www.w3.org/WAI/PF/media-a11y-reqs/

: Establishing authoring guidelines for various territories and\/or languages is outside the core of the format spec, while being valuable to application implementors.

Nigel: notes difficulties that arise with changing captions over time if not all of the text is simultaneously visible due to extreme enlargement.

He notes that timed text is different from an HTML page that can be scrolled.

Unbounded zooming results in the user's missing out on content.

Michael: there's a challenge in that although the spec may not prevent zooming, the UA developers might not think to support it. We need user agent implementation guidance.
... agrees with Nigel taht there's a level of magnification that becomes infeasible for the user, and hence we tend to require it to be supported up to a specified level (400%, for example).

JF: no WCAG requirement to support zooming of captions. He agrees that authoring guidance is important.
... MAUR provides some guidance that can be drawn upon.

Michael: WCAG didn't require support for caption enlargement as this wasn't supported by user agents when the WCAG 2.0 Recommendation was published. In the present context we anticipate future requirements.

Responding to a question about whether MAUR has been found consistent with regulatory requirements (e.g., related to captions), JF notes the expertise of caption and media-related developers who contributed to the process. It hasn't been checked for consistency with regulatory requirements; it reflects user requirements as identified by the W3C.

<pal> https://www.w3.org/TR/media-accessibility-reqs/

JF agrees with the proposition that an informative note citing this MAUR document would improve the caption-related specifications.

<pal> https://github.com/w3c/imsc/issues/316

There's discussion of identifying the GitHub issue to which proposing a link to the MAUR document would resolve.

JF suggests it should be #316.

Summarizing, Janina suggests that a response proposing a non-normative note citing the MAUR would address the issues raised.

Styling Concerns

<nigel> https://www.w3.org/wiki/TimedText/CSSRequirements

Nigel: work has been done to identify which styling features important to captions are supported in CSS.
... there are some desired features that are difficult to map to CSS.

Nigel suggests taht more work needs to be done to develop specific proposals to address these issues.

<pal> Below are some of the issues filed against CSS:

<pal> https://github.com/w3c/csswg-drafts/issues/1975

<pal> https://github.com/w3c/csswg-drafts/issues/1974

<pal> https://github.com/w3c/csswg-drafts/issues/1973

Ian: from the CSS accessibility Task Force point of view, we should identify wehther the issues pertain to accessibility.

<nigel> https://w3c.github.io/ttml2/index.html#style-attribute-derivation

He favours technology-agnostic requriements with technology-specific techniques, and there may be use cases on the accessibility side for some of the desired styling features for captions (with CSS as the technology for which techniques are needed).

Nigel notes the work documented on the wiki.

Nigel notes that examples were presented to the CSS WG.

Janina: we've made good progress. There appears to be a way forward arising from the meeting. Minutes will be posted to both working groups if this is acceptable.

Janina notes likely future topics to be jointly taken up.

There's an invitation for TTML WG participants who may be interested to join the CSS accessibility task force.

Janina will establish whether the conclusions reached at the meeting sufficiently address the open GitHub issues.

<pal> See slides 8-11 at https://drive.google.com/open?id=0B_OQLEGiaO-RamRTam5Edkc1cXc

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2018/02/28 18:08:14 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/Piere/Pierre/
Succeeded: s/authoring requriements/authoring guidelines for various territories and\/or languages/
Default Present: janina, Joanmarie_Diggs, JF, pal, IanPouncey, Nigel, Becka11y, MichaelC, marisa_demeglio
Present: janina Joanmarie_Diggs JF pal IanPouncey Nigel Becka11y MichaelC marisa_demeglio
Regrets: Leonie Gottfried david_singer
Found Scribe: jasonjgw
Inferring ScribeNick: jasonjgw
Found Date: 28 Feb 2018
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)

[End of scribe.perl diagnostic output]