<AWK> +AWK

<Srini> tpresent+ Srini

<KimD> +KimD

<jeanne> pressent+ jeanne

<AWK> Kathy will scribe on Jan 24

<scribe> scribe: alastairc

<AWK> Laura will scribe next week

Survey: https://www.w3.org/2002/09/wbs/35422/WCAG21_SCs_1/results

<adam_solomon> try this: https://www.w3.org/2002/09/wbs/35422/WCAG21_SCs_1/results

<David-MacDonald> https://github.com/w3c/wcag21/issues/57

David-MacDonald: I've defined a second version of this

<Joshue108> Heres the one to merge https://github.com/w3c/wcag21/issues/58

David-MacDonald: the second to last comment https://github.com/w3c/wcag21/issues/57#issuecomment-271204200
... it incorporates terminology used elsewhere in WCAG, so it is a little simpler.

JF: Where did the 25 characters come from?

Wayne: We wanted a line length that wouldn't have excessive hyphenation required or excessive line breaking. 25 chars was relatively practical for western languages and how word wrapping works. It is slightly shorter than a standard newspaper column.
... you can get efficient word wrapping down to 15 characters, but we didn't push it that far.

David: I made it 'blocks of text' rather than 'all text'.

Wayne: We'd removed blocks of text because there are lots of uses of text not in blocks.

Josh: Agree with Mark's comment about "mechanism", appears to require developers to create something to satisfy this.

David: I've done a little testing re-sizing columns, there are some examples (e.g. CNN) and their columns are around 60 characters long, but I defer to the group on that.
... Mechanism is a stock phrase in WCAG 2.0, it is an update on "until user agents", it could be a content or a UA thing. Some things we thought UAs would do, so the mechanism could be the UA.

Mark: Are we at the point removing that language?

<Zakim> AWK, you wanted to say that there is a meaningful difference between "blocks of text" and "all text"

David: That's a WCAG 2.0 concept, and gives us the ability to pass things that are already dealt with in user-agents.

AWK: There isn't a meaningful difference between blocks and multiple-sentences. Text isn't always available in blocks, and should still apply (from the LVTF). We need to decide whether it needs to be blocks or not.

<David-MacDonald> Updated the proposal to "All Text" refresh...

Adam: Regarding screen resolution & magnification, does this relate to a particular size?
... also if you start with very small text and it doesn't get to the required size, is that an issue?

JF: I have a concern about articulating a number of characters, e.g. for German or Welsh, or RLT/ top-to-bottom languages?
... don't see a relationship here to font size. If I have a small font size how does that affect it?

<Joshue108> +1 to that also

<kirkwood> +1 to JF

<gowerm> https://github.com/w3c/wcag21/issues/78

<Joshue108> maybe it doesn't have to be so proscriptive

Mike: Regarding "Mechanism", it implies a burden on the author which is confusing to address.

<Joshue108> +1 to Mike on the mechanism thing - we obviously need to revisit this.

<KimD> +1 to wording (Mike's comment)

Mike: for line length, I'd like clarity on how resize, reflow and line length interacts.

<David-MacDonald> There appear to be 9 SCs that use a Mechanism is Available

<Lisa_Seeman> 1.4.8 Visual Presentation: For the visual presentation of blocks of text and objects, a mechanism is available to achieve the following (Level AA):

<Lisa_Seeman> Foreground and background colors can be selected by the user.

<Lisa_Seeman> Width is no more than 80 characters or glyphs for Latin and Semitic based languages; or 40 for Chinese, Japanese, and Korean; or can be selected by the user.

<Lisa_Seeman> Text is not fully justified (aligned to both the left and the right margins), or justification can be set by the user.

<Lisa_Seeman> Line spacing (leading) is at least space-and-a-half within paragraphs; and paragraph spacing is at least 1.5 times larger than the line spacing.

<Lisa_Seeman> Text can be resized, without assistive technology, up to 200 percent in a way that does not require the user to scroll horizontally to read a line of text on a full-screen window.

<Lisa_Seeman> Increased line and border spacing can be added around blocks of text and objects, such that they can be increased up to 200% without loss of content or functionality.

<Joshue108> Actually looking at the single column issue and the request to merge them - that may be an easier way to frame the use case https://github.com/w3c/wcag21/issues/58

Lisa: We have an SC (51?), we wanted to change visual presentation and upgrade it (faded out). we should merge it into that bullet point, rather than have two proposals about width of text.
... Nothing against the SC at all, but makes sense to merge them.

<marcjohlic> @josh - yep I agree. Might be easier to start w/ talk of merging

Lisa: also, we are making (in ARIA WG + COGA), personalisation settings, and saying that it should work with the plugin would be much easier.

<Lisa_Seeman> https://github.com/w3c/wcag21/issues/51

Lisa: We've made a proposal (51) we have changed "VIsual Presentation" to be AA, and I think we should merge it.

<Zakim> MichaelC, you wanted to comment on viewport / exception 1 and to disambiguate line length restriction from zooming and to +1 I18N concerns and to -1 over-combining with

<Joshue108> +1 to Lisa and merging where possible

MichaelC: Underscore internationalisation aspect, when numerating characters. Also, on zooming & line length we need to be very clear. Line length restriction is related to people's ability to track and stay on a line. But they inter-relate, and meeting one might hamper the other.
... Personalisation is a technique, it sounds good, but I don't think we should combine those as we'd end up with one SC of "make your site personalisable".

<MichaelC> https://www.w3.org/TR/CSS21/visuren.html#viewport

<David-MacDonald> Alastair, can you jump in on ViewPort

MichaelC: not sure if the exception can be reconciled with the definition of the viewport.

Wayne: Font size & line length are related, but there are people with low-vision have limited or no peripheral vision. It is very much the case where you would want short length, not a fact of large print.
... there was also no intention that authors would create the mechanisms. It just means that a mechanism exists, like programatic determinism. UAs have user-style sheets, and what I really want is something more like: If no user-agent for a given technology has a mechanism, then the author is not responsible to create it. However, that isn't possible under WCAG 2.0 framework, where some UAs can't do it.

<Lisa_Seeman_> got it working now

Wayne: also, I don't think the line length should be combined with the reflow criteria. The ability to control line length within table cells isn't possible unless it is separate.

<Lisa_Seeman_> https://github.com/w3c/wcag21/issues/51

<Lisa_Seeman_> but can we merge them so it does address them?

<AWK> Alastair: two points

<AWK> ... marc asking about the relationship between reflow and resizing

<AWK> ... currently resizing doesn't talk about horizontal scrolling so every site passes

<AWK> ... the assumption on the resize content is that it will work with branding etc in place

<AWK> ... first request for resizing was 1000% which seemed to be a problem for many sites to manage

<AWK> ... 400% seems more feasible

<AWK> ... while maintaining the order and presentation

<Zakim> AWK, you wanted to speak to "mechanism"

<David-MacDonald> +1

AWK: Mechanism - is from WCAG2.0, was used and may yet have good utility, and it is key to technology independance of WCAG. There are things you can do as an author in one technology, but taken care of by technology on another platform.

<Zakim> MichaelC, you wanted to say we don´t want SC that only apply if UAs do it and to say ¨mechanism¨ is a bugaboo from WCAG 2.0 that we might want to clear up in Silver and to say

AWK: Mechanism - is from WCAG2.0, was used and may yet have good utility, and it is key to technology independence of WCAG. There are things you can do as an author in one technology, but taken care of by technology on another platform.

MichaelC: When people need things outside of those limits, what are the reasonable limits for the guidelines?

<Zakim> Joshue, you wanted to ask about the scope for merging techniques

Josh: Personalisation, and what we can/can't merge. The initial issue (57) could probably dove-tail with 58 quite well? we should look at merging them, not to create a whitewash of generalisation, but when they have a genuine nexus of things.

<David-MacDonald> I have to drop off in 2 minutes to give a lunch webinar

<Joshue108> +1 to Mike

Mike: the table cells aside, I think they are interdependent: with reflow, with the exception of things like tables everything can reflow into a single column. Once there, I don't need to control the line length anymore, because I can resize my browser width. Once it's in one column I'm not sure what relevance this SC has.

<AWK> @gower it would be good to see the suggestion in text to help think about - a comment for #57/#58?

Mike: Oh, the table cells exception! I think we need a specific technique or SC for table widths.

<JF> +1 I have to agree with Mike's logic here

<gowerm> Thanks Makoto!

Makoto: As JF pointed out, we need to think about internationalisation. The Japanese (and other) characters have 1.4.8 which defines 80/40 chars. If the number will be 25, it might be 12/13 for CJK chars. But I need to check with the LV community in Japan.

<AWK> @makoto - checking with the LV community in Japan would be helpful

<laura> Thank you, Makoto.

Wayne: 1st, the line length question - even though you can reflow, there are UAs that don't allow you to shrink the window. Reflow and word-wrap are different things. Reflow is a linearisation of elements, important because [sorry, missed a bit], it is safer to have a narrow column to search down to find things. Word-wrapping is what we mean in that sense. For the practicality, a 13" laptop starting with 12pt font will easily fit 25 chars on a screen.
... 400% is a practical size, so that's why we had different ones.

<Zakim> alastairc, you wanted to talk about things that would be prevented by no allowing mechanism terms.

<AWK> Alastair: talking about mechanism

<AWK> ... on desktops/laptops you can zoom and the browser will reflow with media queries / responsive design

<AWK> ... on mobile (iOS/Android) this isn't the case

<AWK> ... when you use pinch zoom it expands beyond the layout

<AWK> ... so there is a divide between what "mobile" OS and desktop OS systems currently support

<AWK> ... how to handle? an exception for mobile? (this is what that language was for)

Jon: I'd agree with Alastair & Wayne, you can't just get a bigger monitor. Some people with LV can't use widescreen monitors. Other situations you might be on a mobile device and have a fixed size screen. Magnification is not optimal for a lot of users, you can't see all the information. We need *some* requirement for reflowing content.

JF: I have a few concerns, the other being actual font-size, and starting points.
... I think we need to declare a font size, like the colour contrast SC does.
... thinking of a webpage with H2/3 is set at 22pt, then I resize it I get horizontal scrolling quickly.

AWK: David will have to work through these comments.

Zakim take up item 2

SC Manager questions

AWK: I asked if anyone had any to survey, haven't had a flood of those. Do people need to know more? Is it a time issue?
... Telling me would be a good start. In the process on the SC manager page, it goes into a pull request.
... We haven't resolved the editing current SCs aspect, which might hold some people back.

Rachael: How do we handle overlapping with other SCs which don't have managers?

AWK: I'll check with Josh and MichaelC about that and share with the list.
... any other thoughts/questions please send them to Josh or me.

Voice summary demos and cognitive roadmap (Willie Scott from IBM)

Will Scott at IBM, I'm the software architect for this technology. working on a tech for people with cognitive issues.

Will: Please ask questions as we go through.
... Called "Content Clarifier", which is a natural language generation tool.
... take content and manipulate it, to help people with intellectual disabilities, and extend to other audiences.
... first thing is simplification, to improve comprehension.
... second is content enhancement, which is easier to understand from the demo. We take semantic text and add it to the source.
... we can pull in symbols and more.
... Also doing summarisation, pulling out the most important information from content.
... finally, doing a second pass on summarised content to simplify it, like in step 1.
... demographics that might benefit are people with Intellectual disabilities, older people, and people with second language issues.
... Overlap with WCAG additions, such as replacing complex words with short, simple words. Replace non-literal phrases, idioms etc with literal wording.
... can add iconography, convey meaning visually, and combines with simplified text.
... uses an open library of icons, but that's configurable.
... Some of the key components are IBM Watson, a scalable API, semantic web and "IBM Bluemix".
... the technology is delivered as an API, so a programmer can pull things into their own app/site. The Bluemix is the cloud platform, which has benefit of scalability.
... Watson, IBM's super-computer focused on cognitive computing. Exposing that to developers via APIs, so combining use of Watson, and doing creative things with manipulating content. A lot of the manipulations of the content are based on semantic web. Linked data sources on the web, which are queried to retrieve information.
... Any questions before the demo?
... Demo, showing a content clarifier, (interface looks a bit like a google-translate, but with options of conversational / news/ technical / corp content.
... this is just a basic interface, the API is how we'd expect people to use it.

JF: Is anything being used that relies on Flesch–Kincaid / readability scores?

Will: sort of, sometimes though it will make the content longer. measures for simplicity can complain in that aspect.
... for example, if we insert a simple english definition, it will make it longer.
... replaces things based on many factors. Also, can go out and get definitions for things like "mobile phone", turning the page into a self-glossary.
... Using a News content source, with definitions selected. Inserts explanations inline, after the particular words it selects.
... The content is longer, but it is explaining complex terms in detail.
... it pays attention to the context when adding to or enhancing the content.
... example shown which shows AAC symbols, inline with text. Helpful for the developer because it comes back with the attribution information for the icons. You can change many things, like size, alignment, what the developer wants to do.
... if a human wants to do this, you have to Id the symbols, review them, select the symbol for the wording, and manually insert it. This is automatic. Same goes for the other enhancements shown.
... example of natural language generation, an abstractive approach. Not just extracting sections.

<Joshue108> Very interesting Will -appreciated.

Will: Ultra mode - take the summarised content, do a second pass with simplification.

q

AWK: How does a dev make use of this?

Will: this is deployed via bluemix, IBM's cloud offering, so it works as a cloud API. Once you have an account, you can call it.

<laura> http://ageandability.com/2016/09/21/simplifying-content-for-people-with-cognitive-disabilities/

<Mike_Elledge> How does it deal with pictorial languages

Will: Can follow up with me on email.

Mike_Elledge: Language support?

Will: Use cases for english at the moment, thus far that is the focus. Multi language support is on the roadmap.

<Mike_Elledge> Thanks Will!

<AWK> +Wayne

<laura> Thanks. Bye.

<Jim_S> Thanks!

trackbot, end meeting

This is scribe.perl Revision: 1.148  of Date: 2016/10/11 12:55:14  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/flash kincade/Flesch–Kincaid/
Found Scribe: alastairc
Inferring ScribeNick: alastairc
Default Present: AWK, JF, steverep, Greg_Lowney, Lauriat, KimD, Laura, Kathy, MichaelC, Jim, kirkwood, Lisa, Srini, alastairc, marcjohlic, David-MacDonald, adam_solomon, MikeGower, Makoto, Rachael, Joshue108, Pietro, JaEunJemmaKu, jon_avila, Elledge, Wayne, Jim_S

WARNING: Replacing previous Present list. (Old list: (no, one), JF, steverep, Greg_Lowney, Lauriat)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, JF, steverep, Greg_Lowney, Lauriat, KimD


WARNING: Replacing previous Present list. (Old list: AWK, JF, steverep, Greg_Lowney, Lauriat, KimD, Laura, Kathy, MichaelC, Jim, kirkwood, Lisa, Srini, alastairc, marcjohlic, David-MacDonald, adam_solomon, MikeGower, Makoto, Rachael, Joshue108)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, JF, steverep, Greg_Lowney, Lauriat, KimD, Laura, Kathy, MichaelC, Jim, kirkwood, Lisa, Srini, alastairc, marcjohlic, David-MacDonald, adam_solomon, MikeGower, Makoto


WARNING: Replacing previous Present list. (Old list: AWK, JF, steverep, Greg_Lowney, Lauriat, KimD, Laura, Kathy, MichaelC, Jim, kirkwood, Lisa, Srini, alastairc, marcjohlic, David-MacDonald, adam_solomon, MikeGower, Makoto, Pietro, JaEunJemmaKu, jon_avila, Mike, Elledge)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, JF, steverep, Greg_Lowney, Lauriat, KimD, Laura, Kathy, MichaelC, Jim, kirkwood, Lisa, Srini, alastairc, marcjohlic, David-MacDonald, adam_solomon, MikeGower, Makoto

Present: AWK JF steverep Greg_Lowney Lauriat KimD Laura Kathy MichaelC Jim kirkwood Lisa Srini alastairc marcjohlic David-MacDonald adam_solomon MikeGower Makoto Jim_S
Regrets: Detlev Katie Wilco

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 10 Jan 2017
Guessing minutes URL: http://www.w3.org/2017/01/10-wai-wcag-minutes.html
People with action items: