<mischat> zakim mute me
<scribe> Scribe: Carine
<scribe> scribenick: caribou
<hhalpin> PROPOSED: to approve SWXG WG Weekly -- 24 June 2009 as a true record
<hhalpin> RESOLVED: approved SWXG WG Weekly -- 24 June 2009 as a true record
<hhalpin> PROPOSED: to meet again Wed, 8th July
<tinkster> Minutes are here BTW - http://www.w3.org/2009/06/24-swxg-minutes.html
<hhalpin> RESOLVED: to meet again Wed, 8th of July
<scribe> ACTION: bblfish and hhalpin to write up kaliya's talk for future report
<trackbot> Created ACTION-47 - And hhalpin to write up kaliya's talk for future report [on Henry Story - due 2009-07-08].
<scribe> [PENDING]
DKA: Do we endorse the Identity Commons Purpose and Principles?
<hhalpin> http://wiki.idcommons.net/Purpose_And_Principles
<jsalvachua> +1 sounds good (as principles)
<AndreaP> +1 from me too
<hhalpin> The only issue is that we're not a long-standing group.
Self-organization?
<tinkster> +1
hhalpin: I don't know what that means
... I mostly agree with the principles
<DKA> PROPOSED RESOLUTION: The SWXG supports and endorses the Identity Commons purposes and principles; we share these principles as part of the core values of our effort.
hhalpin: but we should be clear that we do have a charter
<mischat> perhaps we could feed back to the ID commons
<DKA> +1
<hhalpin> +1
<AndreaP> +1
<jsalvachua> +1
<MacTed> +1
<mattroweshow> +1
<tinkster> +1
<FabGandon> +1
<DKA> RESOLUTION: The SWXG supports and endorses the Identity Commons purposes and principles; we share these principles as part of the core values of our effort.
<mischat> +1
RESOLUTION: The SWXG supports and endorses the Identity Commons purposes and principles; we share these principles as part of the core values of our effort
<hhalpin> ACTION: hhalpin to send e-mail to id commons saying we endorse their principles, have a link to us
<trackbot> Created ACTION-48 - Send e-mail to id commons saying we endorse their principles, have a link to us [on Harry Halpin - due 2009-07-08].
<hhalpin> http://www.w3.org/2005/Incubator/socialweb/wiki/RequirementsAndUseCases
<scribe> ACTION: [DONE] hhalpin to create draft use-case document in W3C cvs
<scribe> ACTION: [CONTINUES] danbri follow up re whether to use survey tool at w3c or elsewhere
<mischat> +1
<cperey> +1
<rreck> +1
<hhalpin> It's just a form without content, but it does look pretty and shows you can edit a full W3C spec using a wiki.
<DKA> +1
<AndreaP> +1
<cperey> yes, Harry, i think that was the spirit/intent
<hhalpin> The audience of the survey is non-W3C member.
<cperey> or surveying businesses which are NOT currently part of W3C
DKA: if we want to do survey, it makes sense to use the W3C tool
<hhalpin> It's a survey of *non*-members
DKA: but beyond the WG, another tool would be easier
<cperey> yes, makes sense
<tinkster> yes, makes sense to me.
<cperey> I use surey gizmo
WBS has capability for public questionnaires, AFAIK
<cperey> surveygizmo.com
<mischat> are we intending to send these surveys to end users or to service providers ?
Christine: I use surveygizmo, it's free, lots of reporting capabilities
<rreck> sounds like an endorsement to me
Christine: very flexible
<DKA> Another one Techcrunch seems to be using: Polldaddy.com
<hhalpin> Since Christine is probably going to taking the lead on the survey, then I'd say survey gizmo is right.
<cperey> ah hah!
<mischat> :)
<DKA> PROPOSED RESOLUTION: We'll use internal WBS tool for internal W3C polls, we'll use some tbd external system (e.g. surveymonkey, surveygizmo) to do external surveys.
<hhalpin> Sounds good to me.
<hhalpin> +1
<AndreaP> +1
<DKA> +1
<mischat> +1
<jsalvachua> +1
<cperey> +1
<tinkster> +1
RESOLUTION: We'll use internal WBS tool for internal W3C polls, we'll use some tbd external system (e.g. surveymonkey, surveygizmo) to do external surveys.
<rreck> i wish user stories were called use cases
<hhalpin> tpa?
<scribe> ACTION: [DONE] tpa to W3C-style "edit" user stories
<cperey> can someone past the URI to the wiki page with user stories?
<cperey> paste!
Tim: not much to say, first round of editing done
<mischat> http://www.w3.org/2005/Incubator/socialweb/wiki/UserStories
Tim: transformed in templates, people should fill in blanks
... we may allow more time for adding more stories
<rreck> i have a couple i could add
<cperey> I think it needs to be expanded
DKA: it seems that we need more material
<cperey> what about having use cases be an item in the survey?
<rreck> Oshani and I came up with some we think are interesting
<mischat> i am happy to go through / edit the Privacy and Context subsection, we will see if have any use cases in Garlik with we use to protect our customers privacy
<cperey> question on survey: rank these use cases? are there any use cases missing?
<cperey> what about stories of the past?
DKA: there are stories about the present and about the future
<Tim> I like the idea of prioritization - this could be an internal survey at first.
<mischat> there are good stories around Jennifer Rigley and context
<mischat> in the past
<hhalpin> +1 on noticing "future" vs "past" issue
<cperey> yes, prioritization
<rreck> can we just add metadata describing the type of use case?
<cperey> +1
DKA: maybe we want to let more user stories be added before prioritization
<tinkster> I still have one user story I need to add.
<tinkster> close ACTION-19
<trackbot> ACTION-19 Document developer stories on wiki. closed
<hhalpin> maybe generate more cases first before prioritization.
<cperey> I like adding some metadata to the general template (suggestion by ron)
DKA: anybody who wants to add a story can take an action
<mischat> we could add metadata to the stories
<tinkster> ACTION-42?
<trackbot> ACTION-42 -- Toby Inkster to document danbri's microblogging provenance question as a user story. -- due 2009-06-24 -- OPEN
<trackbot> http://www.w3.org/2005/Incubator/socialweb/track/actions/42
<rreck> Ill take an action to add a usecase
<rreck> unless Oshani wants it
Tim: talk in a week or two when we have more stories
<cperey> can we add to the template an element about the user's device/access network? or do we assume that all are PC + mobile access?
<tpa> cperey, good point
<tpa> cperey, basically we don't have any stories that hinge upon differences in access
<cperey> but would it not be valuable for there to be such differences?
<rreck> i dont know how to action on myself
<tpa> ACTION: tpa to look for similar use cases list to expand ours
<trackbot> Sorry, couldn't find user - tpa
Christine: In the template, can we note if it is assumed that the service is accessed by PC or mobile
DKA: "is it a use case primarily in mobile?"
<rreck> should we define what is meant by mobile vs PC?, like the criteria cperey just mentioned
DKA: I'd not categorize as PC vs. Mobile
<rreck> ok, then let's just add metadata. yeah pre-conditions
DKA: mobile environnement vs. "desk"
<jsalvachua> perhaps could be scenario with a moblie part on it
Christine: "this user is mobile"
<rreck> +1 cperey
Christine: before we edit stories, add it to the template
Soren: it's more about which data is available
... e.g. location
<cperey> but would bandwidth be the "quality of the data"?
<mischat> who was speaking ?
<tinkster> Maybe that should be the beginning of a user story - i.e. why don't we have the same information available whether mobile or otherwise?
<cperey> quality of data seems ambiguous to me?
Soren: conditions in which the network is accessed
<rreck> i think the possible preconditions might be enumerated
<mischat> ff 3.5 gives geolocation
<mischat> hehe
<cperey> context could be "I'm in call"
<cperey> that's probably something that phone (device) knows but a PC does not
DKA: it's becoming an increasingly fuzzy distinction
<mischat> you can be on a call on skype your PC could know that
<cperey> but isn't it our responsiblity to clarify its defnition?
DKA: we should have something in the template
<cperey> haha!
<cperey> who is speaking?
<mischat> :)
<mischat> ?
Soren: I agree, browsers everywhere, not only on PC
<cperey> but would bandwidth be the "quality of the data"?
<cperey> yes, I agree
<cperey> can we please get the name of speaker?
Soren: differences in data transmission (bandwidth, security...)
<cperey> I would like to work/collaborate on that
<cperey> with someone else
<rreck> bandwidth is variable
<cperey> I would like to collaborate on additional definition on template for use cases
<hhalpin> ?
<rreck> my phone switches between networks everywhere i go thereby giving me different bandwidth all the time
<cperey> to clarify use case mobile/fixed
<hhalpin> Notice that we're already 30 minutes in....
<rreck> Ronald Reck
<rreck> im muted
<cperey> I will be happy to work with Ron on this. Misha spoke?
<mischat> me ?
<mischat> no i wasn't speaking
<mischat> the person sounded germna
<mischat> german to me
<cperey> ok, thanks for clarifcation
<cperey> ah!! thank you!
<cperey> I am happy to assist/collaborate but no actions for me
<cperey> agreed
<cperey> :-)
<cperey> I'm learning
<scribe> ACTION: Sören to propose addition to the template for user stories for conditions
<trackbot> Sorry, couldn't find user - Sören
<tinkster> happy to leave them
<tinkster> danbri not on call, but on IRC.
<hhalpin> I can suggest some another speaker
<DKA> Danbri do you have someone from Opera to talk about widgets?
<hhalpin> Eran Hammer-Lahav
<hhalpin> OASIS XRDS-Simple
<tinkster> Now called "XRD" IIRC.
<cperey> harry, can you please adjust your microphone/mouth piece?
<cperey> better
<mischat> XRIs ?
<hhalpin> LRDD Update (Resource Descriptor Discovery)
<mischat> they are they used in OAuth endpoints
<cperey> <yawn>..how relevant to SWXG?
<hhalpin> HTTP and HTML.
<hhalpin> Link Headers and Link elements in HTML.
hhalpin: we need a standardised way to discover info about someone on the web
<mischat> it is used in Oauth and in openid, as far as i am aware
<hhalpin> Mischa - what's your opinion of XRD-Simple implementations?
<mischat> I can report on this
<tinkster> XRDS is in OpenID 2, but not 1.x. XRD/XRDS-Simple is not.
Kaliya?: XRDS is part of OpenID spec
<cperey> agree 100%
<mischat> excellent
<Adam> some quick overview: http://www.hueniverse.com/hueniverse/2009/03/the-discovery-protocol-stack.html and the internet draft: http://tools.ietf.org/html/draft-hammer-discovery-03
<mischat> privacy jungle +1
<danbri> DKA, re opera people ... yes once we send a formal invite... should i just invite an agenda topic and invite them?
<danbri> ie. it's agreed in principle and definite interest
<cperey> move ahead go Soren/Joseph
<danbri> but they want to see exact topics before deciding who to send
<danbri> and when
<hhalpin> ACTION: hhalpin to contact eran to talk, also about LRDD.
<trackbot> Created ACTION-50 - Contact eran to talk, also about LRDD. [on Harry Halpin - due 2009-07-08].
<hhalpin> danbri - I just thought Eran could explain hook up between LRDD and XRD/XRD-Simple
<Adam> +1
DKA: we need to articulate why openness is good for social network business
<Adam> good for businesses +1 :)
<danbri> sure. am talking re opera action only
<hhalpin> Since he is already actively seeking comments on it.
DKA: I'd like to take the resolution that we will work on that
<hhalpin> Oh, I thought opera action was based on w3c widgets/open social compatibility
<cperey> yes
<cperey> I've posted to the mailing list
DKA: if we want to convince people of the relevance of the work and also bring industry people to the group
<tinkster> Michael Hausenblas (DERI) is also pretty interested/informed on LRDD/XRD/etc.
<cperey> it would be nice if there was work item which is focusing on why good for business
http://preibusch.de/publications/social_networks/privacy_jungle_dataset.htm
<cperey> agree...
<hhalpin> soren: if we look at alexa, it's unclear if it's social networking sites
<cperey> blogging service is that a socialnetwork?
<hhalpin> soren: stripped it down to 45 sites, got rid of site not in English as we can't discover then.
<mischat> i dont see why not, you have trackbacks
<cperey> I agree with this premise 100%
Soren: only sites where you can make friends on the site
<Adam> http://jcmc.indiana.edu/vol13/issue1/boyd.ellison.html
<rreck> IMHO, the paper is very impressive
<mischat> +1 rreck
<cperey> social networking platforms= white label social networking servers
<hhalpin> quick question - why exclude content-sharing sites like Flickr?
<jsalvachua> +1 rreck too
<hhalpin> user counts unreliable
Joseph: some sites that claim to be "leading" sites are not
<mischat> good question hhalpin
<cperey> I wonder if Joe and Soren are on the IRC
<hhalpin> maybe not.
<mischat> it doesnt seem like they are
<rreck> i hadnt even thought of that
<tinkster> Joseph: statistical data about social networks given to advertisers seems more reliable than data given in press releases and to general public.
Joseph: categories for the sites: languages, metadata, data collected, privacy controls offered to users
... key aspects of privacy policy of the sites
<hhalpin> "advertise on us" links or sites put through "advertising networks"
ZZZ: how did you access this info? (number of users...)
Joseph: fact sheets or surveys
... we could have included flickr
... their main use cases is vacation pictures to be shared with friends
... you can consider a site to be a SN depending on how people use it
... it's hard to draw a line
<cperey> and there could be more user populations on any service
Joseph: the research in privacy is not new, and the lock-in effect has existed before
... e.g. Amazon
... in SN it's worse because there are all your friends and they would not follow you if you change
<tinkster> Hopefully you'd just lose your links to your friends - you wouldn't actually lose all your friends.
<cperey> agree with this definition
<cperey> of lock-in effects of services
Joseph: Users can't take data to another network
... sites do not promote privacy as an argument on the front page
<mischat> +1 on that this is a problem
<rreck> +1 for sure
<AndreaP> +1
<cperey> problem but the user education is not there
<rreck> i think they are pandering to the lowest common denominator
<oshani> I think "privacy as an argument on the front page" will chase some users away
Joseph: it's not that it is not mentioned at all. You can still dig up a privacy policy
... they claim how good their practices are
<mischat> i think Soren is speaking now
Joseph: sites communicate on privacy depending on the targeted audience
<rreck> i think this means ->privacy doesnt sell
<mischat> :)
<rreck> this was very useful stuff
<hhalpin> I think we should keep going
<hhalpin> but people can leave...
<rreck> great question
Mischa: did you study what happens when you delete info from the sites?
... e.g. if data is really removed?
Joseph: most sites don't claim they will remove the data
... some sites do it partially, but most sites don't make any claim
<cperey> how does Joe conduct this research?
Mischa: EU regulations? govt regulations?
Joseph: experiment on photos, they are still there
<DKA> Chair: HHalpin
Joseph: some sites outsource the photos hosting
... to delivery networks
... facebook uses akamai
<hhalpin> No problem DKA
Joseph: so if you delete the photo on Facebook, the photo might be available for 30 days after
<rreck> i concur 30 days is not reasonable
Joseph: privacy policy says "it will be deleted after a reasonable period of time"
... the UK data protection act and EU data protection directive lay out things
... it is questionnable if 30 days would be ok
Mischa: Do they tell if data will be sent to other countries for storage?
Joseph: No. The "content delivery network" is under their control, they have the master copy, but the rest is unknown
... once a photo has been seen in the UK from the akamai, there's local caching
... it's a huge grey area legally
... the larger sites are the tip of the iceberg in privacy practices
... the smaller sites often started with bad privacy
<cperey> been interesting, very good work, we should build on this.
<mischat> dito in orkut is massive in iran too
<cperey> ++ to what Joe is saying about a broad view of what social networking should be defined. Great deal of diversity
<cperey> our responsibility is to have broad view
Mischa: recent story about last.fm
... are there different levels of privacy?
Soren: problem with affiliate companies
... law is less strict when it comes to transferring info to affiliated companies
... it's difficult for the user to follow where the info is going
<oshani> another similar story: http://tech.slashdot.org/story/09/06/26/1331218/Of-Catty-Rants-and-Copyrights?from=rss
Soren: it's very hard to make a guess from their terms and conditions
... never seen 'affiliates' defined anywhere
hhalpin: geographical diversity pb
... from the W3C perspective, interest in i18n
... in some of your metrics, balance in non-english sites?
Soren: assessed the sites on accessibility, mobile OK, ...
... low scores, disappointing news
... we included non-english sites but difficult to assess privacy policies in foreign languages, and also vs. local culture
hhalpin: unclarity of privacy policy phrasing
<hhalpin> harry: can we cluster privacy practices?
Soren: it was hard to distinguish clusters
<hhalpin> harry: due to privacy policy?
Soren: it's hard to see prevalent practices
Joseph: we might try to diff the privacy texts, to see the %age of wording reuse
<hhalpin> And POWDER?
<hhalpin> P3P hasn't taken off.
<FabGandon> beleives Lorry Cranor did a lot on that subject indeed when working on P3P
Soren: P3P was designed to communicate privacy policies, but less than 10% of the sites implement P3P
... and some are doing it wrong
harry: is it doable to present a human-readable privacy policy?
Soren: several projects going this way
<AndreaP> POWDER might be good if you would like to state which are the privacy policies adopted by a given set of resources.
Soren: it's worth investigating
... also somewhere where W3C could step in
Harry: I'm interested in links to these projects?
Soren: I will try to put something together<tinkster> Perhaps a cut-down dialect of P3P in RDFa might increase update? I think danbri was looking at something like that.
harry: Can we use some of your results in the XG?
Soren: yes, publicly available
<hhalpin> that was a pretty hacky list :)
Mischa: .... list of SN
<hhalpin> should blogging be under social networking?
Joseph: we did look at some of the ones you have on your list
<tinkster> danbri, while you were gone I said: Perhaps a cut-down dialect of P3P in RDFa might increase update? I think danbri was looking at something like that.
<hhalpin> soren: youtube, blogger, take off
Joseph: youtube not really a SN
... friends' communication is limited
harry: whatever has a friends list could be a SNS
<hhalpin> soren: Gmail is pretty huge social network
<hhalpin> mischa: gmail using facebook connect
Mischa: there's a connect between facebook and gmail
Joseph: 2 ideas in textual analysis of policies
... identify reuse of boilerplate text
... difficulty to read
... it might also be interesting to identify relevant dimensions of a privacy policy
... and see how they differ
<melvster> in the uk i believe they have to give you all the data they have on you, but you need to pay a small fee
Joseph: for some observations you need a real account
harry: the study is very educational
... interested in discussion on the mailing list
<tinkster> melvster, organisations have the right to charge a £10 fee for gathering, printing and sending your data, but many will waive that.
<hhalpin> Call adjourned