Technical Architecture Group Teleconference -- 20 Dec 2012

<scribe> scribe: Yves

<trackbot> Date: 20 December 2012

Minutes of previous teleconference

RESOLUTION: minutes approved

<noah> http://www.w3.org/2001/tag/2012/12/13-minutes are approved

Publishing and linking on the Web

<noah> http://www.w3.org/2001/tag/doc/PublishingAndLinkingOnTheWeb-20121015.html

<rigo> W3C Working Draft 25 October 2012

Ashok: I wanted to thank Rigo, Thomas for their comments, I'm quite open to comments; if it means that we need to rescind the document, so be it.

<masinter> i'm not convinced we can't make a consensus document which is useful to multiple audiences

tlr: as I wrote, the document tries to do lots of things for lots of different people, there is value in exploring the topics of this documents

<masinter> i think we're at a point of trying to improve the document along multiple dimensions; i want thomas to say why he thinks that's impossible

what I would recommend is to find an author for the specific goal of rewriting the document with one audience (policy makers has my preference)

<masinter> i'd like more specific details as examples

larry: I'm not convinced yet that we can't create a document for multiple audiences

I would like Thomas to elaborate on why it's not possible

wendy: putting several audiences in the same document bring some techno-determinism.

<masinter> "code is law"

<masinter> http://harvardmagazine.com/2000/01/code-is-law-html

<noah> I think the intention is not to say: "the policy must follow the technology". I think the intention IS to say: "please be aware of how hundreds of millions of copies of software that people use today actually work"

<masinter> Wendy, are you're advising us not to say things that you think are true but just shouldn't be said?

it is stronger to not allude to the law but demonstrate technically some outcomes and reach policy conclusions that might trigger clarification of the law

<wseltzer> masinter, I'm suggesting that the document would be stronger if it omitted the references to law

<masinter> Wendy, why would it be stronger? for whom? Why is what it says weak?

<wseltzer> masinter, because once it starts down the advocacy path, it's easily dismissed as *mere* advocacy

<noah> NM: I don't want to completely avoid terms like "copying". I think it's OK to say: "here are the things the Web actually does as it operates. If some law were to outlaw "XXX" then we would likely have to turn off all proxies on the Web, and performance would tank."

<noah> I thought our goal was to set out terminology that would be useful to those working in the legal domain.

Larry: I don't understand why it would be stronger if laws are not referenced, as the goal is to remain technical

<masinter> i just think the results wouldn't be as useful

tlr: I wasn't saying that writing a document for multiple audience is impossible, just that this document is not doing that effectively

<masinter> i can believe it isn't there

<noah> tlr: ...and I think addressing one audience will be easier than addressing many

tlr: the question is where the TAG can make a difference for the good on that topic and for which audience
... using simpler language and being away from legal implications
... right now the document is untargeted and doesn't work for any audience

Ashok: I'm puzzled by those comments. THe audience is people who make laws and make policies. There are no legal analysis in the document, just wording saying that there are legal disputes, for examples

timbl: embedding or linking is different and socially different. Clarifying that for the press, policy makers, etc... will help not making mistakes with using "linking"

<wseltzer> *how* are embedding and linking different?

tlr: the document identifies an audience, it doesn't work for that audience, and some parts of the documents are clearly written for different audiences

like the definition of linking in HTML is way too complicated for lawyers, it seems written for technologists

<Zakim> noah, you wanted to talk about working through sections

<masinter> i like having motivating examples, though. i'd want to leave the background section because most people don't understand there are such issues

noah: the background section might go deeper than necessary for legal issues, while the intro stays at a fairly high level, could we do something with Thomas' comments?

squeeze section 2

<masinter> i'd rather elaborate or preface what we have

tlr: your foundation seems to be explaining difference between linking and embedding to lawyers

<rigo> https://en.wikipedia.org/wiki/Indecs_Content_Model

<rigo> for reuse

noah: the comments about technical details depends on the technicality of the audience, which vary

wendy: the premise is that linking and embedding are different, in which sense they are different? it might be difficult to infer from the markup the intent of the user

<noah> NM: I'm asking how much savy we should assume on the part of our audience. Clearly some lawyers have very little technical knowledge, yet Judge Posner apparently hacks Java. I'm assuming that lawyers who work on Web policy can be assumed to have at least a vague notion of the role of HTML on the Web.

timbl: this have nothing to do with intent, as you can tell what you want is your intent. If you use an iframe, and make a link there is an underlying intent

<wseltzer> but not everyone who uses the protocol knows or agrees with those intents

<noah> The HTML5 spec says of iframe "The iframe element represents a nested browsing context." Seems pretty vague to me.

<noah> http://www.w3.org/html/wg/drafts/html/master/embedded-content-0.html#the-iframe-element

rigo: if you look at the document, and compare to what Tim just said, the conclusion is that what Tim said is useful, but the document doesn't say that.
... also we need to assess the intent between user clicking, a robot following a link.
... I have identified 8 goals in that document, one of them is the copyright issue. This issue alone is at the root of the creation in policital groups

<Zakim> masinter, you wanted to talk about changing the target audience

Larry: the goal was to try to represent the consensus of the web community

if the W3C endorse a document that explains how the web works, that has more value than any random statement.

Also the document is not explaining the meaning of life, so rigo's comment is strange

<masinter> is the statement wrong, or just 'not enough' ?

<masinter> Rigo "Legislation that governs the possession and distribution of unlawful material (such as child pornography, information that is under copyright or material that is legally suppressed through a gag order) often needs to exempt certain types of services, "

<tlr> "often needs"

<tlr> on what basis does it need to?

<noah> FWIW, I'm inclined to let this discussion go for awhile, but pretty soon we'll have to start thinking about how we move forward.

<Ashok> Thomas, would more detail help in the example sentence?

<rigo> 5.1.6 Licensing

<noah> http://www.w3.org/2001/tag/doc/PublishingAndLinkingOnTheWeb-20121015.html#licensing

<rigo> Websites indicate a license that describes how the information within the website can be reused by others.

<rigo> Just as with HTTP headers, robots.txt and sitemaps, there can be no technical guarantees that crawlers will honor license information within a site.

<noah> " Websites indicate a license that describes how the information within the website can be reused by others.

<noah> Just as with HTTP headers, robots.txt and sitemaps, there can be no technical guarantees that crawlers will honor license information within a site. However, to give well behaved crawlers a chance of identifying the license under which a page is published, websites should: "

timbl: rigo, can you give a concrete example?

<noah> Tim, I think I've quoted the licensing bit that Rigo's talking about in the IRC log just above

rigo: in the publishing area, there can be business models attached to it, and some business models may requires some tehcnologies and obstacles

<masinter> what does 5.1.6 say about policy?

<masinter> there seems to be some concern about things being out of scope "for the TAG to say"

tlr: in 5.1.6 the "should" list could be the subject of debates

another example is offline applications, where it's not really about obscuring following links

<Zakim> noah, you wanted to talk about long term thrashing

<masinter> i needed detailed examples to understand the high-level comments

noah: we receive comments, iterate, and produce new versions. at long as it converge it is good. Looks like we are thrashing here

<masinter> we removed all of the "best practices", although they're still there if you view source

it is likely that even after processing all those comments we would receive the same kind of comments

<masinter> view-source:http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb.html and look for <div class="practice">

rigo: some people, including in the TAG underestimate the complexity, difficulty, etc... of explaining even a single issue to policy makers

<noah> RW: The main critique from me is: trying to do too many things, each of which if done right would take much more work than you probably would think.

<masinter> i think the detailed description of the interaction of policy and technology probably belongs in a working group, and that this is more of a survey of issues

<noah> RW: e.g. could focus on bits such as "difference between embedding and linking". That alone would be substantial work if done right.

<tlr> masinter, a survey of issues doesn't belong on the recommendation track

<noah> LM: I think we intended more of a survey...hearing you say a survey doesn't belong on REC track. Not sure, maybe.

larry: we started more with a survey in mind. I hear that the survey does not belong to the REC track

<masinter> i think we have a couple of choices of where to go

ashok: we expect to have more reviews of that document, so we shouldn't hurry up

tlr: it would be very useful to give clear expectations "what the TAG want to achieve with that document"

the low-level point is that the LC announcement was not sent to the chairs ML

<masinter> I think we got down this road because we were asked if we could bring some clarity with respect to screwy legal decisions.

<masinter> we

<masinter> we've gotten several different suggestions on how to move forward

<masinter> saying less, fixing errors, pick one of several audiences

<wseltzer> [leaves, thanks]

noah: needs guidance on how to proceed, rewrite, rescind, etc...

<masinter> Want to hear from Jeni & Dan, who did most of the initial work

<masinter> yes

timbl: how about having thomas, rigo or wendy work with the editors?

<Ashok> +1 ti

<masinter> maybe a task force?

noah: if we have a focus that can attract their interest, then sure

<tlr> I'm happy to agree to being a reviewer.

<Ashok> great!

<masinter> 1+ to working with a small group to work on one section

<rigo> I identified the following goals in the document:

noah: does it include refocusing?

ashok: maybe

<rigo> - Daniel Glazman's issue: "you can't link to my site"

<rigo> - censorship and the web (various appearances thereof)/ illegal and harmful content

<rigo> - social meaning of linking

<rigo> - deep linking problems and access control

<rigo> - copyright, robots.txt and search engines

<rigo> - youth protection

<rigo> - linking

<masinter> . ACTION: tim to recruit task force to work on one section?

tlr: is one goal to explain the technicality of linking, or the social aspect of linking?

<noah> I think the TAG's competence is more toward explaining technical details that are in fact important. So, I'd err on the side of low level, with the goal that it be competent to lawyers, not programmers.

<rigo> Law enforcement might require that a site owner keep => eDiscovery (big issue in US at the moment)

tlr: I started to review section by section the document, but it didn't help as the main issues was the one I highlighted first, the goal and the audience

<masinter> what does the web community need? To remove uncertainty

tlr: I could send a detailed review, but the document should not be fixed by focusing on details

<Zakim> masinter, you wanted to get back to the fundamentals

larry: we were asked to address concerns about regulatory governance not matching the technology

<rigo> still more: reuse material => DRM

<tlr> you pick a specific audienc

<tlr> you pick specific topics

<tlr> you pick specific messages

<tlr> you write them up in a way that works for that audience

larry: I didn't hear from the commentor how to respond to the need of the web community

<masinter> this is a place where governance doesn't match the technology, or matches awkwardly, how can we help bring the web

<masinter> this isn't intended to be a statement by the TAG, it's intended to be a statement by the web community, as represented by W3C, and as originated by the TAG. That's the reason for going to REC rather than Finding

<tlr> side note, the material about "re-users" doesn't make much sense in the context it shows up in.

[re-statements of comments made earlier]

<rigo> I find actors and agents very useful!

<tlr> http://www.icann.org/en/groups/ssac/documents/sac-056-en.pdf

<masinter> what's the group that produced that?

tlr: the document above is descriptive on issues about content-blocking

not prescriptive

<tlr> (just as an example for a well-written technology meets policy document)

<Zakim> masinter, you wanted to suggest a task force

<masinter> Noah, you asked for discussion about next steps, so i gave a process answer

<noah> Process answer is fine, but I'm hoping the process will focus real soon on a set of goals that people believe are achievable either soon, or incrementally.

Noah: the TAG needs to assess if it is a good use of our time

<masinter> alternative next step: uncomment the "findings" and publish as "NOTE"

<masinter> suggest that W3C "should" take this up but not in TAG

<masinter> thanks indeed

<Ashok> Larry, did you mean uncomment the best practices?

noah: let me know if we need to discuss this on jan 3rd, it will be on the f2f agenda

<noah> ACTIOn-753?

<trackbot> ACTION-753 -- Larry Masinter to do first draft of technical issues list for Jeff -- due 2012-10-22 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/753

<noah> close ACTION-753

<trackbot> ACTION-753 do first draft of technical issues list for Jeff closed

actions

<noah> ACTION-766?

<trackbot> ACTION-766 -- Noah Mendelsohn to respond to e-mail on AWWW "typo" saying TAG did not agree that a change was needed -- due 2012-11-15 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/766

<noah> close ACTION-766

<trackbot> ACTION-766 Respond to e-mail on AWWW "typo" saying TAG did not agree that a change was needed closed

ADJOURNED

- DRAFT -

Technical Architecture Group Teleconference

20 Dec 2012

Attendees

Contents

Minutes of previous teleconference

Publishing and linking on the Web

actions

Summary of Action Items

Scribe.perl diagnostic output