Silver Community Group Teleconference

31 Jul 2018


AngelaAccessForAll, Jemma, Charles, KimD, jeanne, Lauriat, kirkwood, mikeCrabb, LuisG, Jan
Shawn, Jeanne


<imelda> +

<scribe> scribe: jeanne

<imelda> +1

TPAC Reminder!


Today is the last day for TPAC registration for early bird rates.

<Charles> I registered for Mon-Wed

I registered Mon-Wed

Shawn: I registered.

Questions for EO on prototype

<Lauriat> Jeanne: We looked at the form for Silver and plain language. Wanted to talk about the basic questions that we want to ask about the flavor and conformance prototypes.

<mikeCrabb> https://w3c.github.io/silver/prototypes/FlavorPrototype/site/home.html

<Lauriat> Jeanne: I'd ask "How do people like flattening it?"

Jeanne: having the core information at the same level as the general information

<Charles> I would ask people about what roles are useful to parse by

<Lauriat> MikeCrabb: Another thing is "What's missing?" If you'd add something else onto this, what would it be?

<Lauriat> …do we want to have it as a static web and dynamic web at the same level as others, or as some kind of hierarchy, with things under web?

Jeanne: How do you like the labels?

<Lauriat> …Another thing could be about how people are seeing these different things, even though some of them aren't available? Should we hide those unavailable, or display them and have them disabled?

<Lauriat> Jeanne: That's a good one. Charles also added a question in IRC, I like how Charles put it.

<Charles> another question would be “how does one locate all of the success criteria that relate to {static web} (development type)?”

<Lauriat> MikeCrabb: With the way we have it implemented, this is only with one individual SC and techniques. Would people like this in the guidelines, or just how people would conform to the guidelines?

<Lauriat> …Other issues we'll have going forward: what happens when more than one person submits, how would people like to see different methods displayed to them?

<Charles> another question: “should examples be enumerated?”

<Lauriat> Jeanne: Skipping back to a previous question, I don't think we need to assume that we'll have guidelines and principles, except maybe as introductory material.

<Lauriat> Jeanne: Somewhere in here we have to fit the testing as well.

<Lauriat> MikeCrabb: Is this going to the Working Group?

<Lauriat> Jeanne: We told them it exists and they'll look at it, but we'll specifically show it to the EO WG this week for feedback.

<Lauriat> MikeCrabb: Another question, "What do we do if someone proposes a bad technique?"

<Lauriat> Jeanne: We write a nice note back and close it.

<Lauriat> MikeCrabb: What happens if someone suggests a new development tag?

<Lauriat> Jeanne: Kind of a separate discussion, yeah.

<Lauriat> …More in maintenance, so let's not try to fit it in here for now.

<Lauriat> +1 to that.

<Lauriat> Cybele: If two different groups need the same information in different ways, does that get duplicated in different languages/detail? Where's the separation of the duplication?

<Lauriat> Jeanne: The general information tab, we can have a number of different ways of displaying that information. We could link in some of the videos demonstrating how different people with disabilities use the web.

<Lauriat> …I could see most of the plain language implementation in that tab, with more technical information, techniques, and tests in the more specific technical tabs.

<Lauriat> Cybele: It matches what I was thinking, but in practice I wonder if that gets complicated around where that division lies when things (like a tree diagram) start to separate.

<Lauriat> MikeCrabb: The way I've looked at this, the general information is "Here's what you should do" and the other tabs are more "Here's how you do it."

<Lauriat> Jeanne: We have to assume that people know the terms of their field, so I don't think we need to worry about plain language in the specific tags, though we should of course make it as clear as possible.

<Lauriat> …If you're a coder, you know how to read code.

<Charles> all document format prototypes should include that “what” and some common “how”. some document format prototypes will extend to “why” and “for whom” and “what problem is being solved”

<Lauriat> Lauriat: The language around the code should still be plain language.

<Lauriat> Charles: Plain language + their jargon.

<Lauriat> MikeCrabb: Make the code as easy to read for people, and then add language to support what's going on.

<Lauriat> +1 to both of those.

<Charles> part of the plain language style guide would be to determine when jargon needs to be parenthetically or in-context defined

<imelda> *1

<imelda> +1

<Lauriat> [general happiness for Charles' suggestion for adding it to the style guide]

<Charles> and there was much rejoicing

<imelda> +1

<KimD> +1

<kirkwood> +1

<Lauriat> MikeCrabb: We'll put the questions in the form, linking to the applicable prototype as necessary. We can also put a link to the form from each prototype index page for people to offer feedback.

+1 for improved accessibility of Forms rather than Github issues

<LuisG> +1

<Lauriat> +1

<imelda> +1

<Lauriat> Jeanne: I started a Google Doc in the prototypes > plain language folder, for kicking off the style guide.


Conformance Prototype questions

<Lauriat> Jeanne: Let's switch to the conformance prototype and come up with some questions for that, too.

<mikeCrabb> https://w3c.github.io/silver/prototypes/ConformancePrototype/index.html

<Lauriat> Jeanne: This one doesn't have an index page, maybe we should move this to another page and add an index page with a description.

<Lauriat> …I think we want to ask people if they like the way we've adapted the Leeds model to Silver.

<Lauriat> Lauriat: I'd rather keep the questions open ended, rather than yes/no.

<Lauriat> Jeanne: We should ask about keeping WCAG AA as Bronze level.

<Lauriat> …We have more to work out around the points system, though.

<Lauriat> …We have overall conformance, individual testing, and then "accessibility supported" including the browser, assistive tech, authoring tools, etc.

<Lauriat> MikeCrabb: If we're looking at a points-type system, do we have an upper limit to the points you can get? Or the more you add the more you get, with no real upper limit to it?

<Lauriat> Jeanne: The way we were looking at it, the automated testing, the clear measurable things would sit at the lower level. We'd also have points in categories, so you couldn't heavily load up on the simple things from automated tests and ignore the more-difficult-to-measure.

<Lauriat> …So you need to do your alt text, your keyboard access, etc.

<Lauriat> …At an upper level, we'd have more on usability testing and that sort of thing that takes more effort.

<Lauriat> …We can have different point scoring systems for different things. Mom & Pop vs. an e-commerce site.

<Lauriat> …For an e-commerce, you *must* have an accessible shopping cart and such. For massive social media sites, we may not score against the individual images, but more around the capability around making the images accessible.

<Lauriat> MikeCrabb: The conformance model from table 5 in the design sprint, their prototype was more around the activity for testing.

<Lauriat> Jeanne: Yes. Section three is around the conformance and section four around how to measure the conformance.

<Charles> based on this line of conversation of “what if”, it is important to note that a prototype should answer those questions. THEN, the survey questions should be around whether or not those answers are valid and useful and achievable.

<Lauriat> Luis: I have concerns about the points system and whether it gets too complicated.

<Lauriat> Lauriat: The more I look at this, the more I think we need to provide results of conformance for different sites to show how it works and where different sites fall and why.

<Charles> to Shawn’s point, an actual score and how it was calculated would need to be included to understand the model

<Lauriat> Right, exactly.

<Lauriat> Jeanne: I've asked the task forces for examples of SC that didn't get into 2.1 because of the requirement for strict testability, along with how to test them.

<Charles> and if the example site only had 11 applicable SC and passed them all, it would presumably score higher than a site that had 53 applicable SC and passed 51 of them.

<Lauriat> …I set up an example of how you can use a rubric for testing one from the low-vision task force.

<imelda> +1 to both Shawn and Charles' feedback

<Lauriat> Jeanne: I'll write up some things and then Shawn and others can correct it, so that we can have that.

<Lauriat> …We also have an example of alternative test, to do a heuristics evaluation test rather than a flat "does it exist or not" which doesn't address the helpfulness of it.

<Lauriat> …We could add a test showing whether the alternative test was useful.

<Charles> heuristic evaluation can be biased by the subjective conclusions of the evaluator

<Charles> it is useful, but ultimately not a reliable measure

Shawn: It is one part of usability testing.
... it mostly illustrates that usability testing can be done by devs, it's not all UX-lab testing
... it's part of the spectrum of measurement.

<Lauriat> Jeanne: Let's move on, but thank you everyone for the discussion and great questions!

Information Architecture

<Lauriat> Jeanne: I did get an email from Jennison on the IA participation email he sent out - I'll follow up pointing them to the document MikeCrabb set up and another document related to it that we want them to look at.

<Lauriat> MikeCrabb: Yeah, we just need to get people to comment on the document and then we should be set for creating something fairly quickly.

<Lauriat> …With the way that it's working, it should be fairly easy to create a fairly basic version of Silver without the tagging, and then add in the tagging system on top of it.

<mikeCrabb> https://drive.google.com/open?id=14sfnvy-px5p_xNcIM0lyKse9jKCdLCV5

<Lauriat> Charles: I had a comment on the plain language survey. The second question says "Which audiences should Silver target when writing in plain language?" and I'm opposed to that.

<Lauriat> …We want to not have exclusionary language. We want to include more audiences, not exclude any particular audience.

<Lauriat> Jeanne: I'm good with that, I'm not attached to the wording.

<Lauriat> Cybele: Not sure if it's the right wording of the question, but I know we've had a back-and-forth on the audience issue. I think sometimes it's good to break down the diversity of audiences.

<Lauriat> …Not just in terms of the language, but in terms of the information each audience needs to get.

<Lauriat> …I think it's important to think about what audiences there are.

<Lauriat> Charles: I think it's important, but I think that's why the research shows that we need to use plain language. I more meant that the use of the word "target" implies that we want to specifically provide information for one audience.

<Lauriat> …"What are some of the potential benefits of using plain language?" and someone might offer up "So the general public can read it."

<scribe> New document to comment on the Plain Language form <- https://docs.google.com/document/d/1I0VbLM2i3xE9q6dUpOhLPBFBba4VzQ-TzWhVdmN-SpM/edit?usp=sharing

<Lauriat> Angela: I wonder if there's an opportunity to ask a question like that, but address the specific need of a particular audience.

<Lauriat> Jeanne: What you don't see, is we have a lot of additional questions making the form fairly long, so I think we can just drop this question. If it doesn't offer a lot of value, let's just drop it.

<Charles> please send notice of updated survey if done before Friday

<Lauriat> trackbot, end meeting

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2018/07/31 14:32:19 $

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)

Default Present: AngelaAccessForAll, Jemma, Charles, KimD, jeanne, Lauriat, kirkwood, mikeCrabb, LuisG, Jan
Present: AngelaAccessForAll Jemma Charles KimD jeanne Lauriat kirkwood mikeCrabb LuisG Jan
Found Scribe: jeanne
Inferring ScribeNick: jeanne
Found Date: 31 Jul 2018
People with action items: 

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]