<LisaSeemanKest> https://github.com/w3c/personalization-semantics/wiki/Options-for-symbol

<LisaSeemanKest> https://github.com/w3c/personalization-semantics/wiki/Options-for-symbol

<CharlesL> scribe: sharon

IP potential issues with Bliss Symbols

Charles: We will skip first agenda item. OK with IP its acceptable.
... Looks like we are in good shape for agenda item 2.

JF: We want to use Bliss's definition and taxonomy. Do we need to contact Bliss and let them know?

Charles: We'd like to bring back Matt and make sure we are aligned.

Lisa: Matt is not the person. Bliss was happy with this. At the time we thought we would use URL so we need to contact them and let them know.

Janina: Lisa can you review why we picked that set?

Lisa: Wordnet does not have every part of speech and it is in English, so it is missing non-English concepts. Bliss has a process to submit new terms and they will create one.

<Zakim> JF, you wanted to note a more formal engagement with Bliss

Janina: If we have a way to say a verb vs noun, it is key to concerns discussed last week.

Lisa: Its a very lightweight process for adding symbols.

JF: Understands that we are embarking on a relationship with Bliss, that we may want to talk to W3C for an announcement.

Lisa: We can bear it in mind when we are ready.

Janina: Let's discuss announcement with APA.

Easy Reader implementation prototyping 6 month’s left of funding

Lisa: We have an offer of real live users with researchers who could measure usability. What do we want to test?

Janina: Thinks we should test it. Even if its not the end user it would help the author.

Lisa: This would be testing specifically by a group of users.

<LisaSeemanKest> https://mycult-5c18a.firebaseapp.com/

Lisa: Test page on how to make a cup of tea. Is there a different page we want to use? Think about what we want.

JF: In favor of testing. It seem like there are 3 layers here: A new attribute that will take a numeric value or common term. A numeric value is extremely agnostic that references a taxonomy from Bliss. All the author does is add data- to a numeric value. Then linking a symbols set. Do we need to reference that int he HTML documenet? Or will there already have a mapping?

<stevelee> I tried to capture this here - https://github.com/w3c/personalization-semantics/wiki/Symbols---the-big-picture

JF: How will the presentation layer be rendered in HTML? Possibly use Ruby that allows the presentation above the words.
... Need to think about how all 3 pieces will work together? Is the presentation layer our problem?

Lisa: If your linking to map a symbol number to the one a user is using it needs to happen at the user end.

JF: That assumes symbols will be fetched from the cloud. Should we assume the symbols are in the application?

Lisa: A mapping file used at the clinic end for the equivalent of symbol. At the user end they have to map their own symbol.
... The mapping file is a string, it passes the symbol number to the extension, which could be local.

<LisaSeemanKest> (img scr=:"myimg.html>

<LisaSeemanKest> myimg.html{ symol: 2000

Lisa: Enabling a standard way to map and reference.

<JF> No, the author DOES NOT specify a symbol set, the end user does

Lisa: My symbol for warning, but then they will need their own symbol so I make a file that maps to the taxonomy so if some comes that an AC their extension can override them.

JF: Does does not agree with that. The helper application is the tool that should retrieve the appropriate symbol.

Lisa: the author putting the number in HTML we agree that its a default way to do it. If someone has a large ebook they may want a way to do it that does not involve putting it in every time and just reference the mapping.
... Instead of adding the symbol every place they use it there is a mapping. Its an alternative method.

JF: Suggesting that there must be a third party tool that will do the alternative rendering. It is that piece of technology where the mapping lives.

Lisa: Agrees with what John is saying about the user's symbol. She is coming up with an author scenario who is continually using the same symbol they don't want to have the extra wording. If they put it in line every time the loading goes up.
... Has been told that adding data- is an over head.
... They can use a one line symbol mapping and the user agent can look at the file and map it to the user set.

<JF> I really hope we are not conflating emoji and symbols

<JF> +1 to Charles

Charles: In HTML code where the symbol is you don't put anything there and the user agent looks through the text and pull it out on its own. This could be problematic.

<JF> acl Li

Lisa: We may want to do this as a stage 2.

JF: The mapping should happen at the application layer. The number value hands it off to the end user. There is a helper applications that renders the symbols. It's not the authors job.

Charles: May need to give guidance on how the user agent deals with these symbols.

<JF> https://github.com/w3c/personalization-semantics/wiki/Options-for-symbol#proposed-attributes

JF: Link is to a simply example of a home icon and he found 5 different icons. I just need to insert number and your application will pick one of the images or an unknown sixth. Its the one that works best for the user.

Steve: We agree the author can mark up areas for symbols and the user can override the symbols. Lisa is also providing a way to replace anytime you see this html replace it with this symbol.

Lisa: Was using images that they use consistently it would be replaced without having to put in a reference every time.

Charles: This could be an enhancement / stage 2.

Lisa: Easy first is doing it but in a proprietary way. Here we can do it in a way that is open.

Charles: Good use case so the author does not have to add all the data-.
... Some 3rd party looks at BofA and finds the words they use to map to certain symbols.

Janina: There is a security and privacy concern.

<stevelee> ack

JF: We previously talked about a linking mechanism using schema.org.
... Would like to see sample code.

<janina> See ya!

Lisa: Get the basics working. Look at this to make it more scalable.

This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Present: CharlesL stevelee LisaSeemanKest janina Roy MichaelC
Regrets: Becky
Found Scribe: sharon
Inferring ScribeNick: sharon
Found Date: 22 Jul 2019
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.)