See also: IRC log
<trackbot> Date: 23 August 2012
<scribe> ScribeNick: JeniT
<scribe> Scribe: Jeni Tennison
noah: approval of minutes of 12th July?
<noah> http://www.w3.org/2001/tag/2012/07/12-minutes
minutes approved
noah: minutes from 2nd August aren't available yet
... we will have a call next week (30th) but not the following week (6th Sept)
<Ashok> Regrets for 8/30 call
<noah> The TAG will meet at BCS, The Chartered Institute for IT (I.e. the British Computer Society), in London, UK, 7-9 October
noah: who will be there?
<timbl_> yes
yes
<noah> Yes
<jar> Likely regrets for F2F
<ht> yes
ashok: will be there
<jar> possible afternoon attendance by phone
<noah> ACTION-720?
<trackbot> ACTION-720 -- Noah Mendelsohn to make sure we have somewhere to meet in London 7-9th October -- due 2012-06-20 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/720
<noah> close ACTION-720
<trackbot> ACTION-720 Make sure we have somewhere to meet in London 7-9th October closed
noah: please look at your actions to see what can be brought to the F2F
<masinter> we had a hiccup
ashok: I published on 15th: http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2012-08-15.html
... Larry made changes, published on 17th: http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb.html
<noah> Changes are minor
ashok: I don't think there's any significant differences
<noah> To do: snap a new dated version
<masinter> i fixed some markup problems due to respec pipeline issues
ashok: we two telcons, Larry, JAR & I
... discussed what we should be talking about
... I've edited this, and I'm comfortable with it
... let me talk about the edits
... changed the abstract, according to JAR's preferences
... edited the intro
... added intra-document pointers
... particularly to point to the technical mechanisms
... the other big thing was to take out the best practices
... JAR wasn't comfortable with them
... the text goes into a lot of detail
... we can put them back in if people disagree
... we'd like to publish this version as FPWD
jar: what are we trying to accomplish in this session?
noah: I assume it's to decide whether to publish
jar: I should have submitted comments earlier
noah: what are your comments? we can decide what to do
jar: I think it's getting there, much improved
... there was something missing from the introduction
... it talks about the complexity, that we have transforming, caching, archiving and so on
... a naive reaction is to get rid of complexity and make it more like print publishing
... so there needs to be justification for all the things that are going on
... the propaganda would be to say that it's about innovation
... it's good that people are doing these things
... just a sentence to defend against that attack
<masinter> I don't think we need to defend against attacks we haven't received
<masinter> most people accept that the web is complicated for good reason
ashok: there's a section on caching, would you like a couple of sentences talking about why caching is part of the architecture?
jar: no, just one general sentence, I don't know where, that says that the reason we have this complexity is that people are innovating
... and that's a good thing
... we need to defend the complexity, and why it's different from print publishing
<masinter> the "Tussle" paper gives that framework, perhaps just expanding that reference
ashok: we could talk about that before the Tussle paper
jar: it's very similar to the Tussle point
... that's a small thing, just a plug
... also, where it talks about T&Cs
... about where browsing indicates acceptance of the terms
... there's some precedent that suggests that's understood
<masinter> i think rewriting the citation, "Clark et.al. [TUSSLE] characterize the Internet as a system where different kinds of entities interact, often with conflicting goals. Managing these conflicts, they call them "tussles", they argue is crucial to the harmonious development of the internet."
noah: Well, the fact that you don't see copyright notices until you've gotten at least one page is a technical characteristic of the system. We could talk about that.
jar: I think there are some legal precedents about that being understood
ashok: jar, you and I may have some offline conversation about this
jar: I would be direct and say that the person who puts the T&Cs there has something in mind
... also, set of bullets under scope of the document
... the fourth bullet puts us out on a legal limb
... we can't say how to meet any legal restrictions
<noah> "describe the technical measures that websites can take to reinforce any restrictions that they place on the use of content they make available on the web"
<noah> How about s/ensure/attempt to/
<masinter> suggestion for intro last paragraph: The Web is a complex system in which different kinds of entities interact; often they have conflicting goals. These conflicts, called "tussles" in [[TUSSLE]], lead to conflict. This document explores some of the conflicts in the area of publishing and linking."
<masinter> "might attempt to"
jar: we can't tell people what they should do to meet legal restrictions
... it's what they can do if they want to be cooperative
... what you do if you're trying to cooperate
... the bullet before that also raises a red flag
... it's not to reinforce restrictions, it's to implement the restrictions that they want to place on the content
... basically you're talking about DRM here
... you're basically saying how you can do DRM
<ht> "which may enable them to meet"
<noah> How about: "Describe some of the technical mechanisms that can be used for controlling access to restricted material"
<ht> +1 to "implement"
<jar> wha sites can do to *implement* the restrictions that they want to impose on visitors
<masinter> can we try to get FPWD out soon after these edits? Can we agree by email, or do we have to wait until next wewek?
noah: there's passwords and logins and so on
jar: that's DRM in my opinion
<Ashok> jar: "Technical mechanisms to implement restrictions on visitors"
ht: it also covers things like not providing links to things you don't want linked to to things
ashok: the emphasis is on technical means
jar: I don't think we should take a position on how to do it
<noah> Hmm. From a terminology point of view, I tend to separate login restrictions from DRM. Login restrictions keep me from getting something; DRM tends to limit my ability to send to others or freely use content I have obtained"
<Zakim> noah, you wanted to ask about swapping 1.1 and 1.2 and to suggest clarification of note on Silk/Mini and to ask about swapping 1.1 and 1.2
noah: looking at section 1, background and scope
... there's a huge amount of background, and I wanted to know about what the document was about
... maybe swap 1.1 and 1.2 and maybe add a little introductory sentence to the Scope of the Document section
ashok: I'll look at that
<Ashok> noah: Swap 1.1 and 1.2 and add intro para
<Zakim> ht, you wanted to raise concerns about terminology
<Zakim> ht2, you wanted to prefer to wait on FPWD
<jar> Please not to take a position on whether technical barriers are preferable to legal ones or vice versa. That's out of scope. But we can talk about tradeoffs, or make people aware of options that they might not have been aware of.
<Zakim> ht0, you wanted to ask about target audience, and thus level
ht: I read this through to the beginning of section 3 very carefully
... I have copious notes
... I need clarification about what the target audience is
... a lot of my comments depend on what level of sophistication we're assuming
... Larry, Ashok, is this meant for my mother in law? my colleagues who know nothing about the web but are computer scientists?
ashok: yes, the latter
... it's meant for people who are going to write legislation and to work on legal things
masinter: I don't want to educate US representatives
<masinter> maybe their informed staffers
ht: representatives don't write legislation, their staff does
... and they're not computer scientists
<masinter> my view of goals is: minimum requirement is to make something interesting to W3C
ht: that's what I thought, in which case it's hard because we know this vocabulary backwards, and it's hard to write for an audience we don't represent
... it's not a problem for going to FPWD, we can refine as we go along
... I just think we need to be on the same page about who that audience is
... my most important point is about terminology
... because the terminology doesn't connect up well to itself
... it's the one thing I think we have to fix before we publish
... although I think we talked about whether to stick with this terminology, at the F2F
... I have strong reservations of terminology as it stands, because it gets in the way of what we're trying to achieve
... I want to ask: do people agree that we need to get a terminology that we're happy with before we go to FPWD?
... I think changing it after FPWD would be confusing
ashok: an example?
ht: the use of the term 'file'
... I have a lengthy list
<jar> wow... "file" is central to the presentation.
ashok: would it be clarification?
ht: there are some things I'd like to discuss with the group
... like why you have 'resource' and 'web document' and 'file'
... what the value of that is
... I'd like to just use 'web document' unless there's work being done by the distinction
<masinter> i'd rather note HT's issues in a NOTE and still FPWD, rather than getting everything right
noah: is this stuff easy to fix?
ht: I think it's easy to fix
ashok: if you've got stuff written up, could you send it out?
ht: I was hoping to get further before I sent it
... and you could talk me out of this
... I was wondering why the distinction between resources and web documents
<masinter> note: note http://tools.ietf.org/html/draft-barnes-blocking-considerations also
ashok: I inherited it
JeniT: A Web document was intended to be specifically an HTML page, which your browser will typically open in a certain way; resource also encompasses things like spreadsheets.
ht: it seems to be drawing a distinction in an odd place
... I can hand you a URL for an image and your web browser will display it
noah: ...and for spreadsheets, for that matter, depending on user agent configuration
timbl: that's not the point, we're talking about how things work
... it's not a property of the thing, it's how it's used in the architecture
ht: web documents are used within other web documents, using iframes for example
... the distinction between include, embed and link is really important
... but that's where the distinction belongs, not at the level of things that servers serve
... I think it's clearer to focus on how things are used
... it's harder to give a definition of the different kinds of things that has clarity to it
... arguing that an image is different from an HTML page or a PDF, which of those are web documents isn't intrinsically clear
... inclusion, embedding and linking is very clear
... anyway, that's the direction I'd like to go with this
... should I write a clean set of terms?
<noah> I think I agree with Henry.
masinter: it's ok to note things for ashok and myself
ht: it's difficult to tell from my notes where I think we should end up
<Zakim> masinter, you wanted to point to http://tools.ietf.org/html/draft-barnes-blocking-considerations and to speak about audience
noah: can we talk about timing? we want to get this out quickly
... we have a call next week, not the following week
... can we get this done for next week?
ashok: I'm on vacation next week
... I can update it for the 13th call
<masinter> i could work on it
noah: it really fits well with the F2F to have something out early September, so we can get feedback for the F2F
ht: I will get my comments written by midday tomorrow my time
masinter: I can do it; I'm eager to get to FPWD
... if there are issues with the terminology, we'll try to fix them
... and if we can't resolve it, we can write a note to warn people the terminology might change
... so we can get feedback from the intended audience rather than simply talking amongst ourselves
ashok: Larry will attempt an update which you can talk about next Thursday
... and I'll do the following editing
noah: the crucial thing is to pass it to Yves to publish as FPWD
... I'd like to get a decision on that next week
<Zakim> noah2, you wanted to suggest clarification of note on Silk/Mini
noah: so I'm hoping that ashok & Larry will tell us if that's appropriate
... the note about 'split browsers'
... the intention is fine, I'm nervous about the word 'server' in that note
... I think we should reserve 'server' for HTTP server
... maybe use "execute on a centralised machine" or "host", or something that clarifies that the 'server' is not the same as the servers we were talking about before
... architecturally it's a user agent
<masinter> "Split browsers (such as Opera Mini and Amazon Silk) where some software components execute on a server and some software components execute on a client device are not discussed further in this document. Since, in these cases, the client and the server are tightly coupled, they could be considered to be a single User Agent rather than a client and a proxy."
<masinter> i don't see any problem with this
ashok: can we replace 'server' with 'host'
masinter: I like the wording as it is
noah: elsewhere in the document where, until you put proxies in, the web consists of servers and clients
... the servers are providing content, and clients are running user agents
... this bit is talking about a user agent being implemented in a distributed way
... what Amazon is doing is running the back end of the user agent
... I don't know if they're using HTTP, they needn't be
... it could use a proprietary protocol
... for web architecture, the HTTP GET emanates from the Amazon site
... and everything else could move tomorrow
... if you want to use 'server' to mean anything centralised
... and distinguish the servers that respond to GETs and POSTs, you could do that
<masinter> "Server" is a generic term
<masinter> just like "client" in "client/server"
<masinter> We use more specific terms "web server" "proxy server" "origin server"
<masinter> http://en.wikipedia.org/wiki/Server_(computing)
<masinter> We use the word "serveR" in its generic sense
masinter: server is a generic term of art
... there are web servers and proxy servers and so on
... in the context of HTTP, we might say 'server' without saying 'HTTP server' but that's because of the context of the document
<masinter> the real problem is that the "note" is an excuse for not discussing something important
noah: maybe add a phrase to say "(not typically the origin servers or proxy servers described here)"
masinter: the problem is that the note was added to prevent talking about something we should be talking about
... I get the point and I'll add something qualifying term
... I had a pointer to another document
<masinter> http://tools.ietf.org/html/draft-barnes-blocking-considerations
<masinter> it's about one of the remedies used
masinter: I wanted to point out that one of the ways to stop publishing/linking you don't like is blocking
... I'm not sure we can reference this document, but we might want to review it at some point
noah: are you suggesting an action for someone to read it and prepare something for us?
<masinter> it talks about related issues, i guess i'm looking for volunteers
<noah> ok
<masinter> i don't think we need to review before FPWD, but possibly ask for IAB review once we do have FPWD
masinter: I don't think we need to review it before we get to FPWD
... once we have a FPWD, maybe we can reach out to communities working on similar issues
noah: so you'll bring this back to us
masinter: yes
noah: informally, assuming the edits go well, can you indicate whether you will vote to go to FPWD next week?
<ht> +1
<masinter> optimistic +1 for FPWD
+1
<jar_> +1
<Ashok> +1
<timbl_> +1
scribe: no dissent, that's good
... let's hope we have a quorum next week so we can do this
<noah> ACTION-727?
<trackbot> ACTION-727 -- Ashok Malhotra to with help from JAR and Larry to work on a plan for taking a slightly stronger version of the Copyright and Linking draft forward -- due 2012-07-17 -- CLOSED
<trackbot> http://www.w3.org/2001/tag/group/track/actions/727
<noah> ACTION-667?
<trackbot> ACTION-667 -- Noah Mendelsohn to check, when publishing and linking wraps, whether it's time to reinvest in http://www.w3.org/2001/tag/2011/12/evolution/Registries.html -- due 2012-07-15 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/667
<noah> ACTION-721?
<trackbot> ACTION-721 -- Noah Mendelsohn to update product index to target date on publishing & linking -- due 2012-06-20 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/721
<Ashok> Larry, so you have the pen next week. I will pick it up after Labor Day
<noah> ACTION: Larry to prepare FPWD on Publishing and Linking - Due 2012-08-21 [recorded in http://www.w3.org/2001/tag/2012/08/23-minutes#action01]
<trackbot> Created ACTION-732 - prepare FPWD on Publishing and Linking [on Larry Masinter - due 2012-08-21].
noah: Ashok has pointed us to a discussion about linked data and RDF where there is some controversy around whether linked data has to use RDF or could be in other formats
<noah> http://lists.w3.org/Archives/Public/www-tag/2012Aug/0027.html
<noah> From David Booth:
<noah> http://lists.w3.org/Archives/Public/www-tag/2012Aug/0028.html
<noah> Discussion starts approximately here:
<noah> http://lists.w3.org/Archives/Public/public-ldp-wg/2012Aug/0066.html
<noah> Overview of messages:
<noah> http://lists.w3.org/Archives/Public/public-ldp-wg/2012Aug/
Ashok: we had a workshop late last year, and started the Linked Data Platform WG
... the idea as I understood it was to use REST
... which doesn't tell you how to deal with collections, or what you need to do when you have large quantities of data that you want to paginate
... and updates and stuff like that
... and we ought to update REST, possibly not with standards, but with usage guidelines
... then the WG started and they are focused on RDF
... a couple of the people that are part of the WG started saying that it was just data pointing to other pieces of data using URIs
... can't you do this in XML, or JSON?
... and we've had a spirited discussion on this
... and I was wondering what the wise people on the TAG think about it
... and I'm glad Tim is with us, as he has the principles of linked data
... I'm asking what people think about this
timbl: we already have a word -- structured data -- for CSV and JSON
... that's invaluable
... the new word -- linked data -- for stuff that links together
... it's not just CSV that holds URIs with no explanation
... it has to be RDF to be interoperable
... if you make up your own format, or use hypertext
... we need to push for interop
... there are masses of cases where you can keep working in CSV up until the point you have to map into RDF
... and once it's there then people can link it
... so I think we should keep 'linked data' for stuff that's in RDF
<masinter> communication: multiple modes, content negotiation, but "least common denominator", need a lingua franca
ashok: and 'structured data' for other formats
timbl: and REST can use structured data that isn't linked data, and that's fine, but it's not linked
masinter: for interoperability, you need a lingua franca, which everyone speaks
... right now linked data's lingua franca is RDF
... other systems can use other languages, that's ok
... you need a lingua franca or a common gateway
noah: I see a circularity here which I want to ask about, how much is this the 'linked data' brand?
... this reminds me of the HTTP scheme and protocol on the web
... the architecture includes FTP, for example, and there's fuzziness about whether that's on the web
... when you use RDF you get powerful properties because of self-description
... and sometimes the links will go to things other than RDF
... and depending on the context, that might feel like you're on the edges on the graph
... when you do that, is it still linked data? that sounds like a branding question
... there are good reasons to do it
... you could define the 'linked data' more widely, or arbitrarily tighter
timbl: what you're missing is going back to the time when everything was FTP
... when people ask if a FTP server is a web server, the answer is 'no'
... if an RDF document mentions a CSV file
... the CSV file is mentioned in the linked data graph, but none of the data in the CSV is linked
... it's very important that people don't claim to use linked data when they have just published CSV
<jar_> Whether a CSV contains a link (URI), depends on whether it contains a link (URI). SOme do, some don't
noah: when I use a FTP URI I don't get a media type back, isn't that analogous?
timbl: how does that help?
noah: it just seems like a matter of terminology
timbl: it is a brand, yes, but it's an important brand and we should not dilute the brand with structured data
noah: Ashok, do you want to go further?
ashok: branding is very powerful, but what we're really doing is extending REST in different ways
... and once you extend REST, you can use it with XML data for example
... you might say you can use extensions with structured data and with linked data
<noah> I think the point is that you are using REST, but you are not buying strongly enough into the "Self-describing Web". Of the technologies mentioned, only RDF does tthat well.
<Zakim> ht, you wanted to try to say what TimBL said in a different way. . .
ht: ashok, that's true but it doesn't change the crucial core of Tim's point
... that without a way of knowing that a particular sequence of characters is a link, you don't have linked data
... you may have data that contains links, but you have no way of knowing that
... CSV by definition, well-formed XML by definition
... gives you no way of detecting links at all
... there's real blue water between HTML and RDF on one hand and everything else on the other
... in terms of its architectural standing
... that legitimises reserving 'linked data' to HTML+RDFa and RDF
<noah> http://www.w3.org/2001/tag/doc/selfDescribingDocuments.html
<masinter> there's no xml:base in json
<noah> I think the TAG's writing on RDF and the Self-describing Web are pertinent: http://www.w3.org/2001/tag/doc/selfDescribingDocuments.html#RDFSection
<jar_> xml schema provides a way to distinguish link from non-link
ashok: but XML Schema can indicate where URIs are used
ht: yes, but the scale of deployment is orders of magnitude different
noah: in HTML when you have a link, the link text is marked up reliably
... there's nothing in XML Schema that tells you anything other than 'there is a URI here'
... anything you infer about relationship is not standardised
... so you have to know the schema
... you might guess at a relationship based on position in the XML tree
<jar_> grumble
<ht> Two levels: identifying URIs (outside RDF/HTML) is hard enough, identifying _links_ is even harder
noah: but you'd be guessing, whereas in RDF you know
... we wrote about this in the Self Describing Web finding
... RDF is more strongly self-describing than other formats
timbl: you could argue that you could use other technologies, but what we're trying to promote is to use the brand, to get interoperability between linked data publishers
... one of the things which the read/write linked data is going to do is not only provide those protocols, but to provide minimum requirements like understanding Turtle
... with the aim of getting better interoperability
... interoperability is what we're after
<noah> http://www.w3.org/2001/tag/doc/selfDescribingDocuments.html#standards
<noah> Good practice: Web resource representations should be published using widely deployed standards and formats.
noah: that's also in the Self-Describing Web finding
<noah> Indeed, that was my point;.
ashok: that was very useful, I have an idea about how various people are thinking about this, and that's what I wanted
<noah> I think the GPN is trying to capture the spirit of Tim's comment: I.e. having one way of doing each thing is better than having everyone having to implement multiple ways
<noah> ACTION-719?
<trackbot> ACTION-719 -- Larry Masinter to reply to Hannes pointing to the minutes, summarizing the discussion, and asking him if he has any more specific questions. -- due 2012-06-20 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/719
<noah> Open actions by person: http://www.w3.org/2001/tag/group/track/actions/open?sort=owner
<noah> close ACTION-719
<trackbot> ACTION-719 Reply to Hannes pointing to the minutes, summarizing the discussion, and asking him if he has any more specific questions. closed
<noah> http://www.w3.org/2001/tag/group/track/actions/open?sort=owner
<jar_> I think 704 needs attention
noah: which of these need attention for F2F?
<noah> ACTION-704?
<trackbot> ACTION-704 -- Jonathan Rees to with help from Jeni and Henry to try to identify next steps for moving forward on httpRange-14 -- due 2012-08-14 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/704
<noah> Fragids?
<masinter> action-731?
<trackbot> ACTION-731 -- Larry Masinter to and Ashok to remove the best practises parts of the P&L document -- due 2012-08-09 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/731
<masinter> close action-731
<trackbot> ACTION-731 And Ashok to remove the best practises parts of the P&L document closed
<noah> JT: Let's try for a followup draft for F2F
<noah> ACTION-704?
<trackbot> ACTION-704 -- Jonathan Rees to with help from Jeni and Henry to try to identify next steps for moving forward on httpRange-14 -- due 2012-08-14 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/704
jar: I know people have been on vacation, I'm ready to get going again when Jeni & Henry are
... further progress was delegated to this group
<ht> I'm ready as of today
<masinter> i want to tell Jonathan that there is no http URI for the magna carta
jar: we did a fair amount of that up until beginning of July
... then Henry and Jeni had vacations or other things to do
noah: will we have something to discuss at F2F
http://www.legislation.gov.uk/id/aep/Edw1cc1929/25/9
<timbl_> https://www.w3.org/2007/awwsw/tracker/actions/40
masinter, there is a URI for the Magna Carta
<jar_> I want to ask Masinter to give me an example of something for which there *is* an http URI.
<timbl_> http://www.w3.org/wiki/AwwswDboothsRules
timbl: I need to write mapping rules between the vocabularies
<masinter> every http URI identifies a resource that has an http URI (by definition)
<noah> TBL: I want to do this, but think I'll write the mapping rules rather than changing the tabulator
<masinter> there's a 1-1 correspondence between resources-with-http-URIs and http-URIs
<noah> ACTION-40?
<trackbot> ACTION-40 -- Stuart Williams to send MEZ email asking for a joint meeting with the Security WGduring the Plenary -- due 2007-10-25 -- CLOSED
<trackbot> http://www.w3.org/2001/tag/group/track/actions/40
<noah> ACTION-40: Align the tabulator internal vocabulary with the vocabulary in the rules http://esw.w3.org/topic/AwwswDboothsRules, getting changes to either as needed.
<trackbot> ACTION-40 Send MEZ email asking for a joint meeting with the Security WGduring the Plenary notes added
<noah> ACTION-116?
<trackbot> ACTION-116 -- Tim Berners-Lee to align the tabulator internal vocabulary with the vocabulary in the rules http://esw.w3.org/topic/AwwswDboothsRules, getting changes to either as needed. -- due 2011-02-11 -- CLOSED
<trackbot> http://www.w3.org/2001/tag/group/track/actions/116
<noah> Hmm, my phone just dropped.
<noah> I suppose the easiest is to just say: WE ARE ADJOURNED.
<timbl_> . @prefix http: <http://example/httpspec#> .
timbl: the HTTP vocabulary David has used is not dereferenceable
<noah> Sorry for the abrupt departure.
<noah> We'll have call next week.
<noah> Thank you all.
timbl: I'll discharge the action by mapping the rules
thank you, noah
<timbl_> He uses http: as a prefix which doesn't help editing! :-))
<timbl_> nuf bout that
<jar_> masinter, that's fine, but I was asking for an example. Can you give *one* http: URI, and then describe the resource it identifies.