– DRAFT –
Silver Conformance Options Subgroup

sgenda?

Agenda Review & Administrative Items

JS: pretty standard - remind that we need to create an April report

propose we will deliver that on the 30th, but also note that the 29th is a full-day workshop

Jeanne: p[erhaps have the report ready for the 29th then?

JS: OK, we can add that to the list

there is a list with items that we should look at, and there are also some github comments

Assigned github issues https://github.com/w3c/silver/labels/Subgroup%3A%20Conformance%20Options

<sajkaj> https://github.com/w3c/silver/labels/Subgroup%3A%20Conformance%20Options

JS: there are two issues we need to review

Kim: looks like comment from an email

[reads comment aloud]

JS: Kim read issue 450

believe this mirrors one of our existing points in our google doc - principle 5 (thrid usecase)

Blog posts, tec.

JS: summarizes that if we are too strict it will be ignored or will drive people away

Jeanne: notice that it splits 3rd party content into 2 categories: contractual or copyright implications, and then 3rd party 'user-generated' content

were forcing that kind of content into accessiblity may be difficult

so found the split useful

to think about'

JS: agrees - two very different scenarios, and thinks we've already started down that path

example - travel site

so that things that are contracted have a higher bar. But then there are items like payment processing where you can demand minimums in the contract

other services like google, facebook, AWS, etc. - all have sign-in services (federated authentication)

those could likely have higher bars

but being overly strict may introduce problems

Jeanne: thinks there are 3 categories: personal users - low burden, contracted material - higher bar, and then copyrighted material (eg Library of Congress)

JS: interesting use-case, but believes there are exemptions in place to address the needs of a11y

BB: +1 - have seen that before and it is quite specific

JS: relays an experience of encountering copyright barriers

JF: asks about how to integrate copyrighted content into a conformance model

<Zakim> JF, you wanted to ask where this will be leading us to?

JS: think we can't just throw our hands up

BT: usecase: If I go to a restaurant and posts a photo of a menu item and post it on social media, is it the same as, say, the restaurant owner?

Jeanne: maybe they need different rules, but thinks the owner is responsible for text alternatives for their menu. But if it's a patron, then the responsability falls to the poster

so we may need seperate rules to cover scenarios like that

not sure where the line is however

BT: and that's what I am trying to understand - is it based on role, or on something else?

where does the decision take place

JF: how do you annotate that?

JS: concerned that we are getting outside of what W3C can normatively say

believes this is Best Practice guidance, but worried that we are crossing a line

this discussion is already fairly US-centric, need to think about i18n

Jeanne: at this point, just trying to think about use-cases.

Kim: wanted to ask if this is a use-case. Have struggled with this and am unsure

we have a product that courts use, where people can upload evidence for 'sharing'

here there are very complex 'images' or (content) - and how do we provide text alternatives there?

JS: This is a very interesting question

JS: describes experience of being on a jury and dealing with visual evidence

Iim: what I've recommended that the system must *allow* the addition of alt text, but not 'required' (enforced)

<JF> s/Tim/Kim

JF: Plus 1 to Kim

JS: we have a second issue

Next issue is #362. Like that #450 has surfaced 3 different scenarios

[Kim reads out comment]

Jeanne: I think this item is mis-filed - not relevant to this group

Jeanne will re-assign this to somebody else

Action: Jeanne to re-asing Issue #450 to another group

<trackbot> Sorry, but no Tracker is associated with this channel.

Use Cases Discussion (Continued)

JS: there are others we've not touched on yet

<sajkaj> https://docs.google.com/document/d/1GyUYTnZp0HIMdsKqCiISCSCvL0su692dnW34P81kbbw/

[looking at other issues on the google doc]

Looking at Principle 5 Use cases (B & C)

JS: notes similarities to Issue #450 - but bullet list is more expansive

<Zakim> JF, you wanted to comment on "highly technical"

defining "highly technical" will be problematic

Jeanne: last bullet sounded like if yo don't make your software, you're not responsible

JS: Agree, can't use that as an excuse

BT: what were the 3 initial buckets?

Jeanne: personally generated 3rd party, copyrighted material, contracted third-party (login services, news feeds, etc.)

JS: attempting "highly technical" - consider EDU content (advanced math) - but we already are starting to see how to address this (ref: content Usable)

question around submitting work by student

BT: likes seeing the split between University and student - whatever technology supplied by EDU owns the responsibility

JS: takes us back to Kim's comment

KD: this may not be the right spot - what about patent applications? Does that fit into the same or similar use-case

+1 Kim

Kim: does this mean we might need an excemption category?

JS: we could likely draw a line of what can and cannot be achieved - where we know how to implement and can advise that

Kim: continues to trip up over plain language

Jeanne: what plain language talks about is giving orgs the ability to write plain language summaries

JS: Kim is suggesting that they can't - the judge can come back and say you've mis-represented what i meant

it's a legal thing - but can be applied to medical or other fields as well

Jeanne: would be curious to hear from Bruce about how the US Fed deals with this

they have statutes, etc, already

Jeanne: this is something that the plain language folks need to work out

JS: thin this may still fall to us - this is going to require a negociation

Jeanne: think we need to figure out what/where this will apply

JS: We have the responsibility of defining what 'success' looks like, not who is responsible for implementing

<KimD> JS: where summary provides a danger or legal interpretation, perhaps an exemption

<KimD> Kim: perhaps different levels of plain language summaries? One that doesn't involve interpretation as much

<KimD> i.e., more summarizing content than interpretation