<Lisa> clear item 1
<thaddeus> +present
<scribe> scribe: CharlesL
We will be taking off Dec 24'th and 31st
<Lisa> we are taking off the 24th and 31st dec
<Lisa> [17:05:27] ‹Lisa› agenda?
<Lisa> https://github.com/w3c/personalization-semantics/wiki/actions
Lisa: two proposals, we have a wiki page of action items.
Janina: how do we track action items for the group.
Lisa: The other way is to use
action rss bot, or we can open them as github issues would be
another way.
... , we are using issues right now for comments and would be
using it for two things which could be confusing.
<Lisa> https://github.com/w3c/personalization-semantics/issues
<Zakim> janina, you wanted to discuss tracker
Lisa: , 1. use wiki, github issues, or tracker issues bot
<janina> http://www.w3.org/2005/06/tracker/users/my
Janina: everyone can use this
link and get only your action items from.
... , my vote would be to use the trackbot
<Lisa> https://github.com/w3c/personalization-semantics/issues
JF, I believe its called trackbot
<Lisa> https://github.com/w3c/personalization-semantics/wiki/actions
Lisa: does anyone prefer to use the wiki?
<Lisa> https://github.com/w3c/personalization-semantics/wiki/actions
John, I like tracking the issues in tracker, but if we want to do extended work in the wiki, I would support that, but using Tracker as the collector would be a good choice. They could be used together. who ever is assigned an issue could create a wiki page to discuss anything.
Janina, I would worry about them getting out of sync, and Lisa's wiki is a complete list of action items not one for each isue.
Lisa: are there any objections. usingTracker.
JF: I like tracker if everyone settles on I will be ok with a wiki.
Thad: just consistency
Steve: ideally 1 place for all my issues leaning towards Tracker.
Sharon: no preference.
<Lisa> ACTION: set up trackbot
<Lisa> ACTION: lisa set up trackbot
Janina, Roy or Micheal can set that up.
Lisa: Roy or Micheal can either of you set this up?
<Lisa> https://lists.w3.org/Archives/Public/public-personalization-tf/2018Dec/0005.html
<Lisa> topic to two implentiosn support
<Roy> Yes, I could set
<Lisa> https://lists.w3.org/Archives/Public/public-personalization-tf/2018Dec/0005.html
Lisa: continue to find the best way to implement it.
<Lisa> https://lists.w3.org/Archives/Public/public-personalization-tf/2018Dec/0005.html
Lisa: , standardize a 2nd syntax
for the header (JSON, RDFA, etc.)
... , JSON Lets for the whole site can get parsed and put into
the DOM
... , at Benetech there was a programmer that took an XML
format and put that into the page. Second xml file
staandardized in a way if a page is linked to page is contact
us you can put it iin an xml/json file and it cuts the author
burden considerably.
... , someone could fund making these files for critical
services and you don't have to negotiate with website, but
having a 2nd way as well close to the implementation layer
could free us with author burden.
<Roy> trackbot, associate this channel with #coga
<trackbot> Associated this channel with #coga.
JF: I like the "global style sheet" how does a 3rd party inject their data into someone else's file.
Lisa: the main page they would have to add a line in the header.
<Roy> trackbot, associate this channel with #personalization
<trackbot> Associated this channel with #personalization.
JF: thats the critical point if I
make something wonderful, but how would I get my data to link
from someone else's website.
... I don't understand it how it is linked.
... , how some of this could work, it could also run through a
proxy server, the proxy server could add this level of support.
I want to find the easiest solution that can scale.
Lisa: I think we can defer the
mechanics on how to agree on the 2nd implementation
... , we can try to develop that after
<Zakim> janina, you wanted to discuss IP problems with injection to 3rd party sites
Thad: is that what you intended. in terms of work, I did this for 3 years, I can work pulling in 3rd party, security concerns.
Janina: Big companies like Bank of America, there is intellectual property reasons. I am curious to know how that works.
Thad: anything external to your website is considered 3rd party
<JF> Not without understanding the mechanics of this - it feels like a distraction at this time.
<janina> -1
<JF> -1
Lisa: assuming we can address the mechanics, lets say IRC who wants to explore this after the primary implementation
<Lisa> who want to explore after the peimary implantion and current obligation
<stevelee> 0
<thaddeus> +1 if standard vetting practices for external code are in place
JF: this feels a little of a distraction, lets get the primary done, I have concerns, a linking mechanism to get to a style sheet, folks at GPII are doing this.
Lisa: no disagreement, after we get stable working drafts then we can look at this.
<Lisa> JF seasw 4 or five month down the road
JF: we do have a charter and timeline, this is interesting to pursue, but I think this should be back burner this. I support the idea but don't want to spend any time on this like 6+ months.
<sharon> +1 to JF
JF: , tools like Alexa / Siri, are coming up with a standard approach to home automation which is tangentially related.
RESOLUTION: Janina is not sure about this, but look at this after our charter obligations have been met
Lisa: we are trying to 3
attributes 1 token, text and URI's will that work how will that
work compared to multiple attributes.
... , we got in a mess when. you had 2 attributes
<Lisa> https://github.com/w3c/personalization-semantics/wiki/Use-cases#aui-attributes---b
<thaddeus> am sorry I have to drop to take a call
Lisa: , 2 tokens, if we don't allow white space it does not work.
<Lisa> <button id="2" context="high undo">undo</button>
Lisa: , context is just a
placeholder now, with a space between tokens
... , works if we allow space between them. in real we did find
people got confused, where aria got all together but this is
overloaded.
<Lisa> <button id="2" context="simplifcationHighactionUundo">undo</button>
<Lisa> <button id="2" context="simplifcationHigh actionUundo">undo</button>
Lisa: when you have high and undo
it was hard to understand.
... we still have some parsing but we have long names.
JF: so kinda feels like a micro syntax and space separate or comma
Lisa: CSS uses spaces so thats why I chose that.
JF: agreed
Lisa: just experimenting if it is
possible.
... just say "high" is ambiguous.
JF: if we did a scale, simplification=1 to simplification=5 for different levels. High, medium, critical
Lisa: you need to give people more information.
<Lisa> <button id="2" context="simplifcationHigh actionUndo">undo</button>
<JF> @Charles actually more like context="simplification1" (= "high")
<janina> +1 to Becky
Becky: now we are adding more difficulty with additional detail.
JF: but content author and you say its high, you own the content so you are in control. Simplification is always been a sliding scale.
Becky: I was thinking from user persective.
Lisa: right now we are tryingg to
see if these 3 attributes work.
... , if we want 1 attribute for tokens we may need to make the
tokens longer. Undo is an action, High is about simplification
and not arrow up. we need to also supporting a list of them
(comma's spaces etc.)
... , to me there is a cost with longer vocab names
<Lisa> see https://raw.githack.com/w3c/personalization-semantics/value-list/vocabulary-list/index.html#tokens
JF: 3 types of information, action, alternative (easy lang) etc, and finally simplification. Then each of those terms could take multiple terms?
Lisa: I am just looking to see if
this is viable. came back from TPAC with the. idea tokens, URIs
and text
... , action destination may have a problem with ambiguity,
alternative types, message importance / simplification, step
indicators
... , straw poll, should we try to merge them all into one
attribute for just tokens.
<Lisa> one attribute for all the tokens?
Steve: I think that was less to
support.
... , we probably want to explore it and see if it works.
<Zakim> JF, you wanted to respond to Becky
JF: once we have our taxonomy set
up we were going to go back to webplat,
... , here is our usecases and some patterning and bring it
back to them.
... , I agree its time to go back to webplat
<JF> +1 to Steve
Steve: they will have the experience and would know how to do this a lot better than we could
<Lisa> text : https://github.com/w3c/personalization-semantics/wiki/Use-cases#numberfree-and-extrahelp
Steve: , This is only for tokens into a single attribute
Lisa: tokens are the simpler
case.
... , can we have 1 attribute for multiple text (easy Lang and
numbers free
<Lisa> https://github.com/w3c/personalization-semantics/wiki/Use-cases#numberfree-and-easylang-2
Lisa: , someone needs both of
them.
... , 2nd use case dyslexic less words but more numbers
<Lisa> https://github.com/w3c/personalization-semantics/wiki/Use-cases#three-text-fields-usecases
Lisa: , 3 text fields I couldn't see a way that this works. Feedback, EasyLang, and another
[AUDIO cut out]
Lisa: people take a look at this how can we support multiple text or URI's.
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: janina CharlesL JF Becka11y MichaelC Found Scribe: CharlesL Inferring ScribeNick: CharlesL WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth WARNING: No date found! Assuming today. (Hint: Specify the W3C IRC log URL, and the date will be determined from that.) Or specify the date like this: <dbooth> Date: 12 Sep 2002 People with action items: lisa set 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]