State description

• Raised: Initial state; The group has not accepted the issue yet.
• Accepted: Initial state; The group has agreed to address the issue.
• Deferred: The Group has not reached a decision on this issue, but does not intend to work on it further without new input.
• agreed: The Group has made a decision and decided to follow the suggestion.
• Declined: The Group has decided not to address this issue.
• Moved: The issue was moved in an other group. The Group does not intent to resolve this issue.
• Subsumed: The issue has been subsumed by an other issue. A link to the other issue is given.

Decision cycle description

• Announced: Initial decision cycle; The group has announced his decision.
• Agreement: The group received an agreement to refine his decision.
• Proposal: The group received a proposal to refine his decision.
• Objection: The group received an objection to his decision.
• Accepted: The group reopened an issue to reconsider his decision after received an objection or proposal. A link to the new issue is given.
• Maintained: The group maintained its decision after received an objection or proposal.

Categories description

w3cMediaType-1: Should W3C WGs define their own media types? [link to this issue]

What are the general guidelines or policies(if any) for W3C working groups in definingtheir own media types? Should they bedefining them at all?

Request concerning

W3C Data Formats

Discussion history

22 Apr 2002,29 Apr 2002,20 May 2002,17 Jun 2002,8 Jul 2002,25 Jul 2006

Categories

Transition history

raised on 9 Jan 2002 by Mark Baker, on behalf ofXML Protocol WG

accepted on 21 Jan 2002

Background, proposals, threads, notes

• Tim Bray comments on thearchitecture docabout namespaces and media types.
• Message from Joseph Reagle aboutthree bits on MediaTypes andIANA.
• Comments from Mark Baker onXSLT and namespace dispatch.
• Forward from Simon St. Laurentabout ID: MediaFeature xmlns.
• Heads-up from Mark Nottinghamabout XForms document in last calland relation to media types issues.
• How to Register an IANA MediaType, Joseph Reagle
• 18 Dec 2004:IETF approves new media typeregistration process.
• 25 Feb 2004:New registration processdescriptionfrom Martin Duerst

customMediaType-2: What commonality should there be among W3C media types? [link to this issue]

For example, should all these custom XMLtypes being registered be required to usethe RFC 3023 +xml convention? If so, shouldall the SHOULDs of section 7.1 be followed?etc.. The question isn't restricted to RFC3023 issues though. There may be value toother common features between types.

Request concerning

W3C Data Formats

Categories

Transition history

raised on 9 Jan 2002 by Mark Baker, on behalf ofXML Protocol WG

accepted on 21 Jan 2002

agreed on 3 Jun 2002

See resolution forw3cMediaType-1. Seechanges from Chris Lilley regarding RFC3023.

Acknowledgment cycle

announced by group on 3 Jun 2002

nsMediaType-3: Relationship between media types and namespaces? [link to this issue]

What is, or what should be, the relationshipbetween a media type and an XML namespace?

Request concerning

• Namespaces in XML
• RFC3023

Categories

Transition history

raised on 9 Jan 2002 by Mark Baker, on behalf ofXML Protocol WG

accepted on 21 Jan 2002

Background, proposals, threads, notes

• ID: xmlns Media Feature Tag
• message from Larry Masinter
• SeeTB's proposal

Subsumed by issue(s)mixedNamespaceMeaning-13 on 22 Apr 2002

Cf. issuemixedNamespaceMeaning-13.

xformsReview-4: Request to review XForms Last Call document [link to this issue]

Broad request to review XForms Last Calldocument.

Request concerning

XForms 1.0

Categories

Transition history

raised on 18 Jan 2002 by Art Barstow, on behalf ofXForms WG

declined on 28 Jan 2002

In general, the TAG does not expect toreview documents on the W3C RecommendationTrack, unless specific issues are brought totheir attention.

Acknowledgment cycle

announced by group on 28 Jan 2002

uncefactLiaison-5: Invitation to create liaison with UN/CEFACT ebTWGArchitecture Group [link to this issue]

There are several architectural issues inUN/CEFACT and ebXML which should probably besolved by the W3C group. The needs are notspecific to ebXML and several other"Registry" and XML vocabularygroups may have similar requirements.

Request concerning

• UN/CEFACT
• ebXML

Categories

Transition history

raised on 3 Jan 2002 by Duane Nickull, on behalf ofUN/CEFACT ebTWG Architecture Group

declined on 8 Feb 2002

TAG suggests that request be redirected tonew Web Services Architecture Working Group

Acknowledgment cycle

announced by group on 8 Feb 2002

rdfmsQnameUriMapping-6: Algorithm for creating a URI from a QName? [link to this issue]

"It seems to me that the RDFCore andXMLSchema WGs (at the very least) ought todevelop a common, reasonably acceptableconvention as to the mapping between QNamesand URIs. Perhaps this is an issue that theTAG ought to consider (because it is areally basic architectural issue)."

Request concerning

• RDF Model Theory
• XML Schema Part 2: Datatypes

Discussion history

5 May 2002,24 Sep 2002,25 Nov 2002,6 Feb 2003,23 Jun 2003,15 Nov 2003,5 Dec 2002,5 Jan 2004,26 Jan 2004,9 Feb 2004,2 Mar 2004

Categories

Transition history

raised on 22 Jan 2002 by Jonathan Borden

accepted on 29 Jan 2002

Background, proposals, threads, notes

• URIs for terms: motivation [was:Requirements Document]
• Discussion started by TB
• Thread within XML Schema WG
• rdfmsQnameUriMapping-6 affectsWSDL toofrom DanC. See input from JonathanMarsh (forwarded by Paul Cotton)

agreed on 15 Nov 2003

The use of Qnames as identifiers withoutproviding a mapping to URIs is inconsistentwith Web Architecture. See the TAG findingUsing Qualified Names (QNames) asIdentifiers in Content.

Acknowledgment cycle

announced by group on 2 Dec 2003

agreement by reviewer on 16 Dec 2003

Accepts the situation

Action history

whenToUseGet-7: (1) GET should be encouraged, not deprecated, in XForms(2) How to handle safe queries (New POST-like method?GET plus a body?) [link to this issue]

Seecomments from Paul PrescodtoForms WG"I know you've recently been askedabout PUT. During that discussion it arosethat HTTP GET is deprecated in thespecification. Does this mean that XFormswould be incompatible with an applicationlike Google that uses a form to generate aGET URL?"

Request concerning

• XForms 1.0
• Design Issues

Discussion history

4 Feb 2002,8 Apr 2002,15 Apr 2002,22 Apr 2002,29 Apr 2002,5 May 2002,20 May 2002,3 Jun 2002,10 Jun 2002,8 Jul 2002,25 Nov 2002,6 Feb 2003,5 May 2003,23 Jun 2003,7 Jul 2003,15 Sep 2003,8 Oct 2003,5 Dec 2002,5 Jan 2004,9 Feb 2004,2 Mar 2004,10 Oct 2006,17 Oct 2006,24 Oct 2006,11 Dec 2006,11 Dec 2006,9 Jan 2007,26 Feb 2007

Categories

Transition history

raised on 23 Jan 2002 by Dan Connolly, on behalf ofTAG

accepted on 29 Jan 2002

Background, proposals, threads, notes

• Background information fromMicha Dubinko
• Martin Duerst commentson using GET with a message body.
• Draft SOAP HTTP GET binding(DO)
• Additional comments from DanConnollyregarding some security issues (toTAG).
• Seecomments from Noah, however.
• WS-Transfer and HTTP, re TAGIssues whenToUseGet-7 &endPointRefs-47

agreed on 25 Nov 2002

URIs, Addressability, and the use ofHTTP GET.

Acknowledgment cycle

announced by group on 22 Sep 2003

agreement by reviewer on 22 Sep 2003

Acknowledged by DanC by virtue of WGagreement

raised on 10 Oct 2006

Reopening the issue as an umbrella (togetherwith issueendPointRefs-47) for discussingsubmission WS-Transfer.

Acknowledgment cycle

Not started

accepted on 10 Oct 2006

Acknowledgment cycle

Not started

Action history

namespaceDocument-8: What should a "namespace document" look like? [link to this issue]

Thesection on namespacesinWeb Architecture from 50,000 feetstates: "The namespace document (withthe namespace URI) is a place for thelanguage publisher to keep definitivematerial about a namespace. Schema languagesare ideal for this." Tim Braydisagrees.

Request concerning

Web Architecture from 50,000 feet

Discussion history

25 Mar 2002,1 Apr 2002,8 Apr 2002,5 May 2002,24 Sep 2002,25 Sep 2002,18 Nov 2002,9 Dec 2002,16 Dec 2002,6 Jan 2003,13 Jan 2003,6 Feb 2003,17 Feb 2003,24 Feb 2003,24 Mar 2003,7 Apr 2003,14 Apr 2003,23 Jun 2003,8 Sep 2003,15 Sep 2003,5 Dec 2002,26 Jan 2004,2 Mar 2004,14 May 2004,5 Apr 2005,16 Jun 2005,12 Jul 2005,30 Aug 2005,22 Sep 2005,11 Oct 2005,1 Nov 2005,8 Nov 2005,6 Dec 2005,10 Jan 2006,21 Feb 2006,2 May 2006,14 Jun 2006,5 Oct 2006,14 Nov 2006,13 Dec 2006,6 Mar 2007

Transition history

raised on 14 Jan 2002 by Tim Bray, on behalf ofTAG

accepted on 4 Feb 2002

Background, proposals, threads, notes

• XML Schema Part 1: Structures
• Seecomments from Patrick Sticklerthat "namespaces are strictlypunctuation." andreply from Larry Masinter.
• See14 Thesesfrom Tim Bray
• Seethread on RDF and RDDLfrom Stuart Williams.
• SeeRDF in HTMLfrom Tim Berners-Lee
• Emails (1,2) from James Clark to xml-dev aboutschemas, assuggested by Tim Bray. See alsoEmail from James about Relax NGand W3C XML Schema
• Guidelines for the Use of XMLwithin IETF Protocols, section 4.9. See alsoietf-xml-use mailing listin case the document disappears.
• RDDL Challenge
• RDDL Proposal from Tim Bray
• RDDL Proposal from Chris Wilper
• RDDL Proposal from JonathanBorden
• Examples of RDDL in RDFfrom TBL
• RDDL Proposal from Micah Dubinko
• RDDL proposal from Sandro Hawke
• Summary by Normof RDDL Proposals.
• RDDL Proposal from GarrettWilson
• 1 Jun 2003 draft of RDDL
• Namespace URIs should bedereferencable (to find usefulexplanatory material) (12 Feb ftf meeting)
• Refer to draft TAGopinion from Tim Brayon the use of URNs for namespacenames.
• RDDL2 Backgroundfrom Tim Bray.
• grokRDDL.xsl mapping to RDFfrom Dan Connolly.
• namespaceDocument-8 Notesfrom Norm Walsh (24 June 2005)
• Associating Resources withNamespacesfrom Norm Walsh
• Gleaning Resource Descriptionsfrom Dialects of Languages(GRDDL)

Action history

uriMediaType-9: Why does the Web use mime types and not URIs? [link to this issue]

Media types are not first-class objects onthe Web, or are they?

Request concerning

General

Discussion history

8 Apr 2002,15 Apr 2002,22 Jul 2002,12 Aug 2002,30 Aug 2002,9 Dec 2002,6 Feb 2003,5 May 2003,23 Jun 2003,15 Dec 2003,14 May 2004,12 Jul 2005

Transition history

raised on 17 Dec 2001 by Aaron Swartz

accepted on 4 Feb 2002

Background, proposals, threads, notes

• RFC2046
• Proposal by Don Eastlake:draft-eastlake-cturi-03andcommentary from Larry Masinter
• Email from TBL to IETF
• Response from Larry Masinter
• Layout of MIME Media Types pages
• IANA appears to have responded tothe spirit of this draft (seeemail from Chris Lilley).
• XMLP WG request for revision ofuriMediaType-9 issue and relatedfinding

agreed on 21 Oct 2002

TAG Finding: Mapping between URIs andInternet Media Types. The TAG has not resolved this issue sincethe loop has not been closed with the IETF.See Internet DraftA Registry of Assignments usingUbiquitous Technologies and CarefulPoliciesby D. Connolly and M. Baker.

Acknowledgment cycle

announced by group on 21 Oct 2002

Action history

xmlSW-10: Should next version of XML be XML 1.0 - DTDs +namespaces + xml:base + the infoset? [link to this issue]

Should next version of XML be XML 1.0 - DTDs+ namespaces + xml:base + the infoset?

Request concerning

• XML 1.0 Second Edition
• XBase
• XML Namespaces
• XML Information set

Categories

Transition history

raised on 6 Feb 2002 by Tim Bray, on behalf ofTAG

declined on 12 Feb 2002

Forward this to the XML Coordination Group

Acknowledgment cycle

announced by group on 12 Feb 2002

soapRPCURI-11: What is the appropriate relationship between SOAP RPCand the Web's reliance on URIs? [link to this issue]

What is the appropriate relationship betweenSOAP RPC and the Web's reliance on URIs?

Request concerning

SOAP 1.2

Categories

Transition history

raised on 8 Feb 2002 by Paul Prescod

declined on 12 Feb 2002

Forward to XML Protocol WG

Acknowledgment cycle

announced by group on 12 Feb 2002

xmlAsText-12: Do proposed changes to XML 1.1 ignore Unicodeconstraints? [link to this issue]

Do proposed changes to XML 1.1 ignoreUnicode constraints?

Request concerning

XML 1.1

Categories

Transition history

raised on 11 Feb 2002 by Rick Jelliffe

declined on 12 Feb 2002

XML Core WG is aware of these issues. Refertodraft response from David Orchard

Acknowledgment cycle

announced by group on 12 Feb 2002

mixedNamespaceMeaning-13: What is the meaning of a document composed of content inmixed namespaces? [link to this issue]

This was raised in the light of lack ofconsensus result from the workshop, andspecifically prompted by a question,occurring as XEncryption made its way toCandidate Recommendation status in W3C,about the relationship of XEncryption toother specs, and TAG discussion of XSLT"templates" as an apparent cornercase in XML processing.

Second issue: namespace-based dispatching.From TAG draft finding on issues *-{1,2,3},the following draft text was removed fordiscussion as part of this issue:

When processing XML documents, it isappropriate for Web applications to dispatchelements to modules for processing based onthe namespace of the element type.

Correct dispatching and processing requirescontext - in general it is not reasonablenor safe to do namespace-based processingwithout knowledge of the namespace ofancestor elements. Because of this, thenamespace of the root element of an XMLdocument has special status and servesnaturally as a basis for top-level softwaredispatching in the case where the dispatchinformation is not externally supplied.

It is acknowledged that there are exceptionsto this rule, for example XSLT documentswhose root element's namespace depends onthe desired output from application of theXSLT.

It should be noticed that in the case ofcertain sort of element including some inXSLT, XInclude, XEncryption namespaces, thata system conforming to the specificationwill regcognize them at any point in adocument and elaborate them in place,typically producing more XML which replacesthe element instance in the tree.

Request concerning

• XML
• XSLT

Discussion history

11 Mar 2002,22 Apr 2002,5 May 2002,25 Sep 2002,6 Feb 2003,9 Feb 2004,5 Apr 2005

Categories

Transition history

raised on 22 Apr 2002 by Tim Berners-Lee, on behalf ofTAG

accepted on 22 Apr 2002

Background, proposals, threads, notes

• Discussion by Tim Berners-Lee
• Background of the XHTML MediaTypes Notefrom Masayasu Ishikawa

Subsumed by issue(s)mixedUIXMLNamespace-33,xmlFunctions-34,RDFinXHTML-35 on 6 Feb 2003

Split into three smaller issues:mixedUIXMLNamespace-33,xmlFunctions-34, andRDFinXHTML-35

httpRange-14: What is the range of the HTTP dereference function? [link to this issue]

TBL's argument the HTTP URIs (without"#") should be understood asreferring to documents, not cars.

Request concerning

• HTTP
• URI

Discussion history

1 Jul 2002,15 Jul 2002,22 Jul 2002,29 Jul 2002,16 Sep 2002,24 Sep 2002,6 Jan 2003,27 Jan 2003,6 Feb 2003,7 Jun 2003,23 Jun 2003,22 Jul 2003,28 Jul 2003,12 May 2004,7 Feb 2005,15 Mar 2005,29 Mar 2005,3 May 2005,31 May 2005,15 Jun 2005,27 Feb 2006,27 Feb 2006,26 Feb 2007,30 May 2007,31 May 2007,11 Jun 2007,18 Jun 2007,2 Jul 2007

Transition history

raised on 25 Mar 2002 by Tim Berners-Lee, on behalf ofTAG

accepted on 6 Feb 2003

Background, proposals, threads, notes

• Axioms of Web Architecture, by Tim Berners-Lee
• What do HTTP URIs Identify?by TimBL
• History of Fragment Identifiersby Roy Fielding.
• Moby Dick example and summaryfrom TB.
• Namespace v. Namespace Document isan example of this issue asevidenced bythis thread.
• See discussion ofinformation resources
• See discussion of phrase "onthe Web" at22 Jul 2003 teleconferenceand subsequent threads
• Summary and proposal from NW
• GoodURIs

agreed on 15 Jun 2005

The TAG provides advice to the communitythat they may mint "http" URIs for anyresource provided that they follow thissimple rule for the sake of removingambiguity:

• If an "http" resource responds to aGET request with a 2xx response,then the resource identified by thatURI is an information resource;
• If an "http" resource responds to aGET request with a 303 (See Other)response, then the resourceidentified by that URI could be anyresource;
• If an "http" resource responds to aGET request with a 4xx (error)response, then the nature of theresource is unknown.

Acknowledgment cycle

announced by group on 18 Jun 2005

Action history

URIEquivalence-15: When are two URI variants considered equivalent? [link to this issue]

From Joseph Reagle:

Stephen [Farrell] has asked an interestingquestion below that I expect will beimportant to any activity that uses URIs asidentifiers in the context of asemantic/security application: when are twoURI variants considered identical?

Request concerning

• URI
• XML Namespaces

Discussion history

22 Jul 2002,29 Jul 2002,30 Aug 2002,18 Nov 2002,16 Dec 2002,20 Jan 2003,7 Feb 2003,24 Mar 2003,31 Mar 2003,14 Apr 2003,28 Apr 2003,30 Jun 2003,5 Dec 2002,5 Jan 2004,2 Mar 2004,22 Mar 2004,14 May 2004

Categories

Transition history

raised on 19 Feb 2002 by Joseph Reagle

accepted on 1 Apr 2002

Background, proposals, threads, notes

• XML Namespaces
• Internationalized ResourceIdentifiers (IRI)
• Character Model for the Web, notablyChapter 4andChapter 8
• Email from Misha Wolf
• Input from Martin Duerston IRIs
• See text from TimBL onURI canonicalization.
• Assigned initially to Dan Connolly(at30 Aug 2002 teleconf). SeeProposal from DanC
• Email from Larry Masinter proposingA simpler solution to %7e vs %7Evs ~ in namespace comparison
• Proposal from Noah Mto say "Hex-escapes aren'tallowed [in xml namespace names]Don't use them."
• Email from TBL aboutRationalizing the term URI
• Email from Larry MasinteraboutURI BOF minutes from IETF56

agreed on 14 Apr 2003

Draft finding:URI Comparison(link not maintained but see RFC3986).. This has been integrated into RFC2396bis (CVS repository); the TAG expects to follow the progress ofRFC2396bis. Commentary and resolution shouldhappen through the IETF process.

Acknowledgment cycle

announced by group on 14 Apr 2002

agreement by reviewer on 12 Dec 2002

Action history

HTTPSubstrate-16: Should HTTP be used as a substrate protocol? Does W3Cagree with RFC 3205? [link to this issue]

From Mark Nottingham:

The IETF has recently published RFC3205,"On the use of HTTP as aSubstrate" [1] as Best CurrentPractice.

This document makes a number ofrecommendations regarding the use of HTTP.Some are reasonable, such as guidelinesabout what kinds of scenarios the HTTP ismost useful in, how to use media types andmethods to extend the HTTP, etc. However, italso bases a number of recommendations on afuzzily-defined concept of 'traditional use'of the HTTP. These directives may seriouslylimit the future potential of the Web,effectively freezing its capability tocommon practice in 2001."

Request concerning

RFC3205

Discussion history

30 Jun 2003,12 May 2004,7 Feb 2005,21 Sep 2005

Transition history

raised on 24 Mar 2002 by Mark Nottingham

accepted on 1 Apr 2002

Background, proposals, threads, notes

• Response from Randy Hall to IETF
• Message from Roy Fielding
• Propoposed criticsm of RFC3205from TBL.
• Email from Mark Bakeron tunneling as a property of theapplication.
• Seemessage from Larry Masinterw.r.t. Web services.

deferred on 12 May 2004

The TAG decided to defer this issue pendingany attempt to enforce RFC3205.

Action history

charmodReview-17: Request to review "Character Model for theWeb" Last Call document [link to this issue]

Request to review "Character Model forthe Web" Last Call document

Request concerning

Character Model for the Web 1.0

Discussion history

3 Jun 2002,24 Jun 2002,14 Jul 2003,8 Sep 2003,8 Oct 2003,2 Feb 2004,15 Mar 2004,22 Mar 2004

Transition history

raised on 16 Apr 2002 by Misha Wolf, on behalf ofI18N WG

accepted on 29 Apr 2002

agreed on 24 Jun 2002

Comments sent by Norm to the I18Ncomments listandreminder from Dan Connolly. See alsoComments from CL. See other TAG resolutions regarding thisissue in3 Jun minutes.

Acknowledgment cycle

announced by group on 24 Jun 2002

Action history

qnameAsId-18: Is it ok to use Qnames as Identifiers? [link to this issue]

Is it ok to use Qnames as Identifiers?

Request concerning

Namespaces in XML, section 3

Discussion history

20 May 2002,10 Jun 2002,17 Jun 2002,24 Jun 2002,15 Jul 2002,15 Dec 2003,12 Jan 2004,9 Feb 2004,23 Feb 2004,2 Mar 2004,15 Mar 2004

Categories

Transition history

raised on 19 Apr 2002 by Joseph Reagle, on behalf ofXKMS WG

accepted on 29 Apr 2002

Background, proposals, threads, notes

• A radical finding on UsingQualified Names (QNames) asIdentifiers in Contentfrom Norm Walsh

agreed on 22 Jul 2002

Finding:Using QNames as Identifiers

Acknowledgment cycle

announced by group on 22 Jul 2002

agreement by reviewer on 12 Dec 2002

Action history

formattingProperties-19: Reuse existing formatting properties/names, coordinatenew ones [link to this issue]

Reuse existing formatting properties/names,coordinate new ones

Request concerning

CSS3

Discussion history

17 Jun 2002,8 Jul 2002,15 Jul 2002,5 Dec 2002

Categories

Transition history

raised on 8 May 2002 by Steve Zilles

accepted on 20 May 2002

Background, proposals, threads, notes

• CSS2
• XSLT
• HTML CG Work
• Proposed changes from RickJeliffe

agreed on 22 Jul 2002

Finding:Consistency of Formatting PropertyNames, Values, and Semantics

Acknowledgment cycle

announced by group on 22 Jul 2002

errorHandling-20: What should specifications say about error handling? [link to this issue]

What should specifications say about errorhandling?

Request concerning

W3C specifications

Discussion history

27 May 2002,3 Jun 2002,6 Feb 2003,30 Jun 2003,15 Nov 2003,5 Jan 2004,2 Mar 2004

Categories

Transition history

raised on 22 May 2002 by Rob Lanphier

accepted on 3 Jun 2002

Background, proposals, threads, notes

• QAWG issue 51

agreed on 15 Nov 2003

The TAG believes it has addressed a majorityof points about the issue in the 11 Nov 2003draft, with pointers to relevant sections3.4 and 1.2.2, as well as the section onversioning and extensibility. The TAGdeclines at this time to handle thefollowing questions raised by the reviewer:(1) Extension of XML. Answer: Applicationdependent. (2) Handling of deprecatedelements.

Acknowledgment cycle

announced by group on 2 Dec 2003

agreement by reviewer on 12 Dec 2003

Action history

RFC3023Charset-21: Do all "shoulds" of RFC 3023 section 7.1apply? [link to this issue]

Do all "shoulds" of RFC 3023section 7.1 apply?

Request concerning

• W3C Data Formats
• RFC3023

Discussion history

3 Jun 2002,26 Aug 2002,9 Sep 2002

Categories

Transition history

raised on 9 Jan 2002 by Mark Baker, on behalf ofXML Protocol WG

accepted on 3 Jun 2002

Background, proposals, threads, notes

• This issue was split off from issuew3cMediaType-1

agreed on 7 Oct 2002

TAG Finding: Internet Media Typeregistration, consistency of use

Acknowledgment cycle

announced by group on 8 Oct 2003

augmentedInfoset-22: Infoset augmentation outside of PSVI? [link to this issue]

So I recommend a TAG finding along thefollowing lines:

1. Type-augmented XML is a good thingand a recommendation should beprepared describing it both at theinfoset and syntax level. (I gatherthere is already some work alongthese lines in XML Schema?). Seriousconsideration should be given to80/20 points rather than simplyre-using the plethora of primitivetypes from XML Schema.
2. Type-augmented XML has nothing tosay about default values created inany schema.
3. Any software can create and/or usetype-augmented XML, whether or notany validation is being performed.
4. Work on XQuery and other things thatrequire a Type-Augmented Infosetmust not depend on schemaprocessing, and should not havenormative linkages to any schemalanguage specifications.

Request concerning

XML Schema Part 1: Structures

Discussion history

17 Jun 2002,8 Jul 2002,26 Aug 2002

Categories

Transition history

raised on 12 Jun 2002 by Tim Bray, on behalf ofTAG

accepted on 17 Jun 2002

Background, proposals, threads, notes

• Email from Noah Mendelsohn
• Email from Mary Holstege
• Summary, proposals from Tim Bray
• Proposal from Rick Jeliffe
• Request from Tim Bray to"unraise". See thread for pushback.

agreed on 26 Aug 2002

For now, the TAG has decided the issue bywithdrawing it. From TB: "I learnedthat while there are linkages between xqueryand xml schema, they are non-normative; youcan implement xquery with other schemalanguages; so I don't see an architectureissue at the moment. I submitted a largecomment to the xquery process that theredoes remain too much intermingling with xmlschema that could easily go away. If the twospecs aren't made sufficiently independent,I expect to come back to the TAG."

Acknowledgment cycle

announced by group on 26 Aug 2002

agreement by reviewer on 26 Aug 2003

Acknowledged by TB by virtue of WGagreement

xlinkScope-23: What is the scope of using XLink? [link to this issue]

For me this questions depends on whether thedocument type is a human-readable hypertextdocument, when generic hypertext xml toolswould benefit from knowing what is a link,and whether significance of the URI inquestion is a hypertext link or somethingdifferent.

Request concerning

XML Linking Language (XLINK)

Discussion history

17 Jun 2002,1 Jul 2002,26 Aug 2002,30 Aug 2002,16 Sep 2002,24 Sep 2002,7 Oct 2002,21 Oct 2002,11 Nov 2002,18 Nov 2002,16 Jan 2003,20 Jan 2003,6 Feb 2003,17 Mar 2003,30 Jun 2003,2 Mar 2003,14 May 2004,6 Oct 2004,7 Feb 2005,22 Mar 2005,21 Sep 2005

Transition history

raised on 14 Jun 2002 by Tim Berners-Lee, on behalf ofTAG

accepted on 17 Jun 2002

Background, proposals, threads, notes

• Steven Pemberton (for HTML WG)
• HLink Working Draft(Member-only)
• Background from Mimasa(TAG-only)
• SVG and non-XLink attributes.
• Tech arguments from StevenPemberton
• Notes from 22 Oct 2002 XML CGcall(Member-only).
• When to use XLinkfrom TBL.
• Summary of technical discussionfrom Stuart Williams.
• SkunkLink: a skunkworks XMLlinking proposalfrom Micah Dubinko
• Summary of reqsfrom Lloyd Rutledge.
• Proposal for kind of xlink basicfrom TB.
• Seedraft, andSW messageto CG chairs.
• XML Linking Language (XLink)Version 1.1

agreed on 21 Sep 2005

Paragraph4.5.2 of Web Architecturecloses the issue. See also draft ofXML Linking Language (XLink) Version 1.1.

Acknowledgment cycle

Not started

Action history

contentTypeOverride-24: Can a specification include rules for overriding HTTPcontent type parameters? [link to this issue]

For me this questions depends on whether thedocument type is a human-readable hypertextdocument, when generic hypertext xml toolswould benefit from knowing what is a link,and whether significance of the URI inquestion is a hypertext link or somethingdifferent.

Maybe a compromise is to only allow the linkto specify the content-type when the serveris FTP (or something else with nocontent-type control) or the HTTP serverreturns text/plain or octet-steam, whichseem to be used for "don't know"types.

Request concerning

• Speech Recognition Grammar Specification Version1.0
• HTTP/1.1

Discussion history

6 Feb 2003,24 Mar 2003,7 Apr 2003,5 May 2003,12 May 2003,16 Jun 2003,23 Jun 2003,30 Jun 2003,7 Jul 2003,8 Oct 2003,5 Dec 2002,15 Dec 2003,26 Jan 2004

Categories

Transition history

raised on 14 Jun 2002 by Tim Berners-Lee, on behalf ofTAG

accepted on 22 Jul 2002

Background, proposals, threads, notes

• Link metadata cannot overrideserver media typefrom DanC
• Email from Christo TAG about SMIL 2.0, andrationale
• Email from Chris Lilleyon MSDN documentation onMIME-sniffing in Internet Explorer(FindMimeFromData method)
• Email from Chris on MIME typesand encoding
• Scott McGlashan email to TAGwith links to Voice WG points.
• Bug filed by Martin Duerstregarding default charset setting inApache conf file
• Draft Finding"Client Handling ofAuthoritative metadata"
• Approved FindingAuthoritative Metadata

agreed on 4 Dec 2003

The3 Dec 2003 Editor's Draft of theArchitecture Documentaccurately represents the TAG's position onthe authoritative nature of server messages.

Acknowledgment cycle

announced by group on 4 Dec 2002

agreement by reviewer on 4 Dec 2003

Acknowledged by TBL by virtue of WGagreement

Action history

deepLinking-25: What to say in defense of principle that deep linking isnot an illegal act? [link to this issue]

Strawman from Tim Bray:

The architecture of the World Wide Web doesnot support the notion of a "homepage" or a "gateway page",and any effort in law to pretend otherwiseis therefore bad policy. The publication ofa Uniform Resource Identifier is, in thearchitecture of the Web, a statement that aresource is available for retrieval. Thetechnical protocols which are used for Webinteraction provide a variety of means forsite operators to control access, includingpassword protection and the requirement thatusers take a particular route to a page. Itwould be appropriate to bring the law tobear against those who violate theseprotocols. It is not appropriate to use itin the case where information consumers areusing the Web according to its publishedrules of operation.

Request concerning

Slashdot article on court ruling in Denmark

Discussion history

30 Aug 2002,9 Sep 2002,7 Feb 2003,17 Feb 2003,15 Sep 2003,8 Oct 2003,3 Nov 2003,15 Nov 2003

Categories

Transition history

raised on 5 Jul 2002 by Tim Bray, on behalf ofTAG

accepted on 22 Jul 2002

Background, proposals, threads, notes

• Summary from Henrik FrystykNielsenof Danish Court case (TAG only)

agreed on 7 Feb 2003

AcceptedDraft finding from TB

Acknowledgment cycle

announced by group on 7 Feb 2003

agreement by reviewer on 7 Feb 2003

Acknowledged by TB by virtue of WGagreement

Action history

contentPresentation-26: Separation of semantic and presentational markup, to theextent possible, is architecturally sound. [link to this issue]

Strawman from Tim Bray:

I would however, support an assertion in thearchitecture document that importantinformation SHOULD be stored and(optionally) delivered with markup that isas semantically rich as achievable, and thatseparation of semantic and presentationalmarkup, to the extent possible, isarchitecturally sound.

Request concerning

• XML
• CSS
• XSL-FO
• XML Accessibility Guidelines

Discussion history

24 Sep 2002,6 Feb 2003,21 Jul 2003,8 Oct 2003,14 May 2004,7 Feb 2005,21 Sep 2005

Transition history

raised on 15 Aug 2002 by Dan Connolly, on behalf ofTAG

accepted on 26 Aug 2002

Background, proposals, threads, notes

• Email from Kynn Bartletton separation of content frompresentation.
• Email from Sean Palmerabout WAI PF work in this area.
• Draft finding from CL:Separation of semantic andpresentational markup, to theextent possible, isarchitecturally sound(Historical document)

agreed on 21 Sep 2005

Section4.3 of Web Architecturecloses the issue.

Acknowledgment cycle

Not started

Action history

IRIEverywhere-27: Should W3C specifications start promoting IRIs? [link to this issue]

The XML Core WG would like TAG input onwhether the desirability of adopting IRIsinto the web infrastructure early outweighsthe anticipated disruption of legacysystems.

The XML Core WG would also like TAG input onthe wisdom of early adoption given the"Internet Draft" status of theIRI draft. So far adoption has relied on "copyand paste", but there is potential forthese definitions to get out of sync.

Request concerning

• XML
• XML 1.1
• [Internet Draft] IRI draft

Discussion history

28 Oct 2002,11 Nov 2002,18 Nov 2002,27 Jan 2003,6 Feb 2003,31 Mar 2003,7 Apr 2003,14 Apr 2003,28 Apr 2003,14 May 2004,21 Sep 2005,13 Dec 2005,23 Jan 2007,12 Feb 2007,26 Mar 2007

Transition history

raised on 9 Oct 2002 by Jonathan Marsh, on behalf ofXML Core WG

accepted on 28 Oct 2002

Background, proposals, threads, notes

• URIEquivalence-15
• Email from Larry MasinteraboutURI BOF minutes from IETF56
• Summary email from Tim Bray
• Comments from Ray Whitmerabout IRIs in the DOM (see sameemailwith inclusion. See1.3.2 DOM URIsin 7 May 2003 DOM 3 Core draft
• RFC3490: InternationalizingDomain Names in Applications(IDNA)
• See TB'sproposed step forward on IRI 27.
• RFC 3987is a Proposed Standard

Action history

fragmentInXML-28: Use of fragment identifiers in XML [link to this issue]

Do fragment identifiers refer to asyntactice element (at least for XMLcontent), or can they refer to abstractions?

Example from17.2.2 SVG fragment identifiers:

MyDrawing.svg#svgView(viewBox(0,200,1000,1000))

The SVG spec states "This form ofaddressing specifies the desired view of thedocument (e.g., the region of the documentto view, the initial zoom level) completelywithin the SVG fragment specification."

From Dan Connolly:

Do you consider the quoted paragraph abovein error?

Or do you disagree with my interpretation ofit, i.e. thatMyDrawing.svg#svgView(viewBox(0,200,1000,1000))identifes a view of the drawing, and not anyparticular XML element (nor other syntacticstructure) in the document.

Request concerning

XML

Discussion history

12 May 2004,7 Feb 2005,10 May 2005,31 May 2005,1 Nov 2005

Transition history

raised on 31 Oct 2002 by Dan Connolly, on behalf ofTAG

accepted on 4 Nov 2002

Background, proposals, threads, notes

• httpRange-14,rdfmsQnameUriMapping-6(since the WSDL WG is asking us howthey should do nearly exactly whatwe want the XML Schema WG to do e.g.provide URIs for "schema"components)
• Guidelines for the Use of XMLwithin IETF Protocols. See alsoietf-xml-use mailing listin case the document disappears.
• SemWeb use casefrom Sandro Hawke.
• Per13 Jan 2003 teleconf, note that the TAG considersXInclude issues raised by M.Muratato be related to this issue. Whichmedia type should be used forinterpreting fragment identifiers?Section 2.4 in the architecturedocument says the media type of theretrieval result, but Sections 4.2and 4.3 in the XInclude CR saystext/xml or text/plain.
• Seeemail from Paul Cottonabout WSDL component designators.
• Also related: Content negotiation?Opacity of URIs?

agreed on 12 May 2004

• In general, the fragment part of aURI may be used to refer toabstractions as well as syntacticfragments of a representation; themedia type identifies aspecification, which explains thesemantics.

Acknowledgment cycle

Not started

Action history

xmlProfiles-29: When, whither and how to profile W3C specifications inthe XML Family [link to this issue]

When, whither and how to profile W3Cspecifications in the XML Family

Request concerning

XML family of specifications

Discussion history

2 Dec 2002,9 Dec 2002,16 Dec 2002,6 Jan 2003,13 Jan 2003,27 Jan 2003,6 Feb 2003,15 Sep 2003,5 Dec 2002,5 Jul 2005

Categories

Transition history

raised on 25 Nov 2002 by Paul Grosso

accepted on 2 Dec 2002

Background, proposals, threads, notes

• SOAP 1.2, 26 Jun 2002 draftmakes use of an XML format that doesnot permit any internal subset,despite the fact that XML 1.0 doesnot define such a profile/subset ofXML. Seerationale from David Fallside
• Options for dealing with ID, from Chris Lilley
• Web services arch WG positionfrom Mike Champion
• Glenn Adams emailon the existence of a number ofstandards in the television domainthat: (1) disallow internaldeclaration subsets; (2) requirestandalone="no"; (3)require a document type declaration,with a specifically enumerated setof public FPIs to be supported;
• Extensible Messaging andPresence Protocol (XMPP) Core,draft-ietf-xmpp-core-06.From Chris Lilley: "XMPP dissalows PIs,commnents, and both internal andexternal TD subsets."
• Henry Thompson proposalfor "minimal" conformanceclass.

agreed on 27 Jan 2003

TAG recommendation for work on subset ofXML 1.1. Seefollowup to AC (Member-only). Work is being carried out in the XML CoreWG.

Acknowledgment cycle

announced by group on 30 Jan 2003

Action history

binaryXML-30: Standardize a "binary XML" format? [link to this issue]

Given that binary infosets (currently,binary PSVIs) is what I work on daily and that I amcurrently investigating ways in which theycould fit naturally into the web(content-coding registration for instance),I would be very interested in knowing what-- if anything at this point -- the TAGthinks of them and of how they could bestfit in.

Request concerning

XML

Discussion history

2 Dec 2002,13 Jan 2003,27 Jan 2003,6 Feb 2003,17 Feb 2003,12 May 2004,7 Feb 2005,28 Feb 2005,15 Mar 2005,5 Apr 2005,12 Apr 2005,26 Apr 2005,3 May 2005,10 May 2005

Transition history

raised on 9 Oct 2002 by Robin Berjon

accepted on 2 Dec 2002

Background, proposals, threads, notes

• Seeexplanation scenario from DC; a typical conversation aboutbinary XML.
• Rationale from Robin Berjon
• XML binary compression progress:X3D, XFSP projectfrom Don Brutzman.
• OGC is another customer for binaryXML.
• Binary XML Workshop Minutes
• Summary from Chris.
• XML Binary Characterization UseCases.
• TAG opinion on XML Binary Format, 24 May 2005.

deferred on 12 May 2004

W3C has chartered theXML Binary Characterization WorkingGroupto address this issue. The TAG anticipatesreviewing the WG's deliverables in thisarea.

Action history

metadataInURI-31: Should metadata (e.g., versioning information) beencoded in URIs? [link to this issue]

The TAG's preliminary response is that URIsshould not include metadata. The TAGaccepted this issue to provide guidance onaddressing the issues raised.

From Ossi:

To outline the following text, I'm actuallysuggesting (asking comments for) two ratherpractical things:

1. There should be a uniform way todeclare version history of webresources (recommended by W3C)?, andmore importantly
2. There should be a "clean",uniform way to refer to (and thusaccess) the metadata of webresources?

Request concerning

• Architecture Document
• URI

Discussion history

2 Dec 2002,6 Feb 2003,7 Jul 2003,21 Jul 2003,8 Oct 2003,14 May 2004,7 Feb 2005,21 Sep 2005,13 Dec 2005,21 Mar 2006,2 May 2006,16 May 2006,30 May 2006,14 Jun 2006,25 Jul 2006,8 Aug 2006,19 Sep 2006,4 Oct 2006,7 Nov 2006,14 Nov 2006,11 Dec 2006,2 Jan 2007

Transition history

raised on 25 Nov 2002 by Ossi Nykänen

accepted on 2 Dec 2002

Background, proposals, threads, notes

• SeeArch Doc
• SeeTim Bray mail on Apple MusicStore.
• Fnding
• Email from Tony HammondonThe OpenURL Framework(PDF)
• On the topic of URI fragility, seethe section entitledWhy Names Changein TBL'sThe Myth of Names and Addresses
• Comments from Larry Masinteron draft finding
• See comments fromMark Nottinghamandfollowup from Noah M.
• Guide to Feedback andOutstanding Issuesby Noah M.

agreed on 11 Dec 2006

Seeresolution.

Acknowledgment cycle

announced by group on 3 Jan 2007

Action history

xmlIDSemantics-32: How should the problem of identifying ID semantics inXML languages be addressed in the absence of a DTD? [link to this issue]

I would like to raise a new issue to theTAG. The issue is how to determine IDattributes in any new work on XML, such as anew profile or subset as dealt within issuexmlProfiles-29. I understand that this issue will benormatively referred to in anycommunications on issue #29.

Chris Lilley has started anexcellent discussionon the various options for ID attributes, soI won't duplicate that work. A number ofresponders have said they are quitesupportive of providing a definition of IDsas part of any new work on XMLProfiles, suchas the Web Services Architecture WorkingGroup. There is also some pushback, so itseems worthy to have a continued discussion,and the TAG should attempt to quickly reachconsensus.

Request concerning

XML 1.1

Discussion history

27 Jan 2003,6 Feb 2003,14 Apr 2003,30 Jun 2003,8 Oct 2003,12 Jan 2004,7 Feb 2004,12 May 2004,19 Apr 2005,21 Sep 2005

Transition history

raised on 30 Jan 2003 by David Orchard, on behalf ofTAG

accepted on 30 Jan 2003

Background, proposals, threads, notes

• Offshoot ofxmlProfiles-29. This issue was raised on requestfrom the TAG at their27 Jan 2003teleconference.
• Options for dealing with IDsfrom Chris Lilley
• Web Services Architecture WGposition on XML profiling/subsettingfrom Mike Champion
• Findingfrom Chris Lilley: Approved by TAG12 May 2004
• xml:id Requirementsfrom theW3C XML Core Working Group
• SeeComments from Henry Thompsonon the XPointer Framework definitionof Shorthand Pointer.

deferred on 12 May 2004

At their 12 May 2004 ftf meeting, the TAGaccepted the proposed finding "How shouldthe problem of identifying ID semantics inXML languages be addressed in the absence ofa DTD?". The issue is deferred while the XMLCore WG continues work on this issue.

agreed on 21 Sep 2005

xml:id Version 1.0is a Recommendation

Acknowledgment cycle

Not started

Action history

mixedUIXMLNamespace-33: Composability for user interface-oriented XML namespaces [link to this issue]

Raised by the TAG as an offshoot ofmixedNamespaceMeaning-13.

Request concerning

• XHTML
• MathML
• SVG
• XForms
• SMIL
• other user interface-oriented applications

Discussion history

6 Feb 2003,2 Mar 2004,7 Feb 2005,21 Sep 2005,1 Nov 2005,10 Jan 2006

Transition history

raised on 6 Feb 2003 by TAG, on behalf ofTAG

accepted on 6 Feb 2003

Background, proposals, threads, notes

• Offshoot ofmixedNamespaceMeaning-13
• An XHTML + MathML + SVG Profile
• W3C Workshop on Web Applicationsand Compound Documents
• W3C Compound Document FormatsWorking Group

deferred on 1 Nov 2005

Pend this issue until significant progressis made by theW3C Compound Document Formats WorkingGroupin a public working draft.

Action history

xmlFunctions-34: XML Transformation and composability (e.g., XSLT,XInclude, Encryption) [link to this issue]

Raised by the TAG as an offshoot ofmixedNamespaceMeaning-13.

Request concerning

• XSLT
• XInclude
• Encryption
• other specifications that involve transformations ofXML content

Discussion history

6 Feb 2003,21 Sep 2005,31 Jan 2006,7 Feb 2006,18 Apr 2006,25 Apr 2006,11 Dec 2006,30 Jan 2007,11 Jun 2007

Transition history

raised on 6 Feb 2003 by TAG, on behalf ofTAG

accepted on 6 Feb 2003

Background, proposals, threads, notes

• Offshoot ofmixedNamespaceMeaning-13
• The Interpretation of XMLdocumentsfrom TBL.
• The elaborated infoset: Aproposal

Action history

RDFinXHTML-35: Syntax and semantics for embedding RDF in XHTML [link to this issue]

Raised by the TAG as an offshoot ofmixedNamespaceMeaning-13.

Request concerning

• RDF
• XHTML

Discussion history

6 Feb 2003,14 May 2004,22 Feb 2005,28 Feb 2005,13 Dec 2005,14 Jun 2006,13 Dec 2006,12 Feb 2007

Transition history

raised on 6 Feb 2003 by TAG, on behalf ofTAG

accepted on 6 Feb 2003

Background, proposals, threads, notes

• Offshoot ofmixedNamespaceMeaning-13
• namespaceDocument-8
• Using RDF/XML with HTML andXHTML
• Some discussion about approaches
• "Crisp" statement ofthe problemfrom Dan Connolly
• Email from M. Ishikawaon using modular namespaces (MNS)
• RDF in XHTMLsummary by TBL.
• GRDDL: Gleaning ResourceDescriptions from Dialects ofLanguagesby Dan Connolly and DominiqueHazael-Massieux.

deferred on 13 Dec 2006

The TAG decided to defer this issue pendingwork by the GRDDL WG and/or theRDFa/HTML/SemWeb-deployment WGs.

Action history

siteData-36: Web site metadata improving on robots.txt, w3c/p3p andfavicon etc. [link to this issue]

The architecture of the web is that thespace of identifiers on an http web site isowned by the owner of the domain name. Theowner, "publisher", is free toallocate identifiers and define how they areserved.

Any variation from this breaks the web. Theproblem is that there are some conventionsfor the identifies on websites, that

• /robots.txt is a file controllingrobot access
• /w3c/p3p is where you put a privacypolicy
• /favico is an icon representative ofthe web site

and who knows what others. There is ofcourse no list available of the assumptionsdifferent groups and manufacturers haveused.

More in theoriginal message from TBL.

Request concerning

URI space

Discussion history

24 Feb 2003,8 Oct 2003,5 Jan 2004,12 Jan 2004,22 Feb 2005,13 Dec 2005,2 May 2006,9 Jan 2007

Transition history

raised on 10 Feb 2003 by Tim Berners-Lee, on behalf ofTAG

accepted on 24 Feb 2003

Background, proposals, threads, notes

• Note:Defining a Web Site.
• Sitemap protocol.

Action history

abstractComponentRefs-37: Definition of abstract components with namespace namesand frag ids [link to this issue]

Is it wise to use fragment IDs foridentifying abstract components within anamespace, even though it is the mostnatural and convenient mechanism? Is thereanother mechanism that would be preferable?

Request concerning

• Namespaces in XML
• WSDL

Discussion history

24 Mar 2003,14 Apr 2003,5 May 2003,23 Jun 2003,8 Oct 2003,20 Oct 2003,2 Mar 2004,22 Feb 2005,3 May 2005,16 Jun 2005,21 Sep 2005,25 Oct 2005,1 Nov 2005,16 May 2006,14 Jun 2006

Transition history

raised on 3 Feb 2003 by Jonathan Marsh, on behalf ofWSD WG

accepted on 24 Mar 2003

Background, proposals, threads, notes

• rdfmsQnameUriMapping-6,namespaceDocument-8
• Email from Tim Brayabout relation to RDDL.
• Email from Paul Cottonabout Schema WG response to the WSDWG questions about Schema ComponentDesignators
• Email from Noah Mendelsohnabout distinction between schemacomponents and schema documents.
• Email from Larry Masinteron relation to WebDAV to providingaccess to metadata.
• Summary of options from DavidOrchard
• Draft summary
• Comments from MSM on behalf ofSchema WG
• "XML Schema: ComponentDesignators"specification published by the W3CXML Schema WG.
• Draft findingAbstract Component References

Action history

putMediaType-38: Relation of HTTP PUT to GET, and whether client headersto server are authoritative [link to this issue]

Some scenarios that this issue concerns:

• Client PUTs representation to serverwithout content type information;what is proper server behavior?
• Client PUTs representation to serverwith content type information, butserver ignores. Is thisarchitecturally incorrect?
• Client PUTs representation to serverwith detectably inconsistent contenttype information. What is properserver behavior (e.g., signal errorand not silently ignoring)?

Request concerning

Authority of client headers

Discussion history

16 Jun 2003,22 Feb 2005,26 Apr 2005,3 May 2005,21 Sep 2005,6 Dec 2005,28 Mar 2006,11 Apr 2006

Transition history

raised on 6 May 2003 by Julian Reschke

accepted on 16 Jun 2003

Background, proposals, threads, notes

• contentTypeOverride-24
• "PUT semantics and MIMEheader handling"
• First version of approvedfinding "Authoritative Metadata"(25 Feb. 2004)
• Reopening discussion
• New version of approved findingAuthoritative Metadata(12 April 2006)

agreed on 18 Apr 2006

Approved TAG findingAuthoritative Metadata.

Acknowledgment cycle

announced by group on 19 Apr 2006

Action history

rdfURIMeaning-39: Meaning of URIs in RDF documents [link to this issue]

TBL: "The community needs:

1. A concise statement of the abovearchitectural elements fromdifferent specs in one place,written in terms which the ontologycommunity will understand, withpointers to the relevantspecifications.
2. Some outline guidance on specificquestions brought up in emailquestions.

This includes:

• Is a given inference engine expectedto take into account a givendocument under given circumstances?
• how does one avoid having to committo things one does not trust?

There may be some need to clarify frequentmisunderstandings by making some thingsclear."

Request concerning

• URI
• RDF
• OWL
• HTTP

Discussion history

18 Aug 2003,15 Sep 2003,22 Feb 2003,21 Sep 2005,13 Dec 2005

Transition history

raised on 13 Jul 2003 by Tim Berners-Lee, on behalf ofSemantic Web CG

accepted on 18 Aug 2003

Background, proposals, threads, notes

• Somebackground summarized by DanC
• Mailing list for discussions on thisissue: public-sw-meaning (archive)

Action history

URIGoodPractice-40: What are good practices for URI construction? [link to this issue]

Some issues:

• Fragment identifier syntax
• Related issue: URI Squatting?

Request concerning

URI

Discussion history

20 Oct 2003,14 May 2004,22 Feb 2005,21 Sep 2005,13 Dec 2005

Transition history

raised on 16 Oct 2003 by David Orchard, on behalf ofTAG

accepted on 20 Oct 2003

Background, proposals, threads, notes

• abstractComponentRefs-37
• Well Designed URLs areBeautiful!

Action history

XMLVersioning-41: What are good practices for designing extensible XMLlanguages and for handling versioning? [link to this issue]

What are good practices for designingextensible XML languages and for handlingversioning?

Request concerning

XML

Discussion history

3 Nov 2003,10 Nov 2003,15 Nov 2003,2 Mar 2003,14 May 2004,14 Feb 2005,21 Sep 2005,22 Sep 2005,8 Nov 2005,5 Dec 2005,14 Feb 2006,27 Feb 2006,3 Mar 2006,12 Jun 2006,18 Jul 2006,25 Jul 2006,8 Aug 2006,29 Aug 2006,5 Sep 2006,4 Oct 2006,5 Oct 2006,5 Oct 2006,12 Dec 2006,12 Dec 2006,16 Apr 2007,23 Apr 2007,30 Apr 2007,14 May 2007,30 May 2007,30 May 2007,31 May 2007,25 Jun 2007

Transition history

raised on 27 Jun 2003 by David Orchard, on behalf ofTAG

accepted on 3 Nov 2003

Background, proposals, threads, notes

• Draft finding from DO and NWpart 1,part 2
• Text proposed by DO and NW forArch Doc
• Instance diagram

Action history

ultimateQuestion-42: What is the answer to life, the universe, andeverything. [link to this issue]

This "issue" collects all discussionsrelevant to Web architecture that are notdirectly related to any other issue.

Request concerning

The Hitch Hiker's Guide to the Galaxy

Discussion history

15 Nov 2003

Transition history

raised on 15 Nov 2003 by Tim Berners-Lee, on behalf ofTAG

accepted on 15 Nov 2003

Background, proposals, threads, notes

• Wikipedia entry on issue 42
• Google Calculator; cf 6*9
• - - - - - - - - - -
• Semantic Web Architecture:16 Jun 2005,22 Sep 2005,13 Dec 2006
• Self-describing Web:06 Dec 2005,27 Feb 2006,6 Mar 2007,1 Jun 2007.
• State in Web application design:21 Feb 2006,28 Mar 2006,25 Apr 2006,9 May 2006,6 June 2006,13 June 2006,26 Sep 2006
• Security/Authentication:15 Jun 2005,5 Jul 2005,20 Sep 2005,8 Nov 2005,24 Jan 2006,21 Mar 2006,11 Apr 2006,25 Apr 2006,13 Jun 2006
• CURIEs:30 May 2006,13 Jun 2006,27 Jun 2006,26 Feb 2007,19 Mar 2007,2 Apr 2007.SeeabbreviatedURIs-56for continuing discussion.
• Device description repositories:6 Jun 2006,27 Jun 2006
• New and changing media types:18 Jul 2006,25 Jul 2006.
• Naming and Virtual Worlds18 June 2007

Action history

DerivedResources-43: How are secondary resources derived? [link to this issue]

How are secondary resources derived? The TAGdiscussed the case of parse='xml' andfragment identifiers in XInclude.

Request concerning

XInclude

Discussion history

15 Nov 2003,12 May 2004,21 Sep 2005,31 Jan 2006

Transition history

raised on 15 Nov 2003 by David Orchard, on behalf ofTAG

accepted on 15 Nov 2003

Background, proposals, threads, notes

• XInclude issues list

Action history

xmlChunk-44: Chunk of XML - Canonicalization and equality [link to this issue]

The XML architecture has tended to be builtaccording to a motto that all kinds ofthings are possible, and the application hasto be able to chose the features it needs.This is fine when there are simply the XMLtoolset and a single "application". However,real life is more complicated, and thingsare connected together in all kinds of ways.I think the XML design needs to be moreconstraining: to offer a consistent idea ofwhat a chunk of XML is across all thedesigns, so that the value of that chunk canbe preserved as invariant across a complexsystem. Digital Signature and RDF transportare just intermediate parts of the designwhich need to be transparent. This requireda notion of equality, and a relatedcanonical serialization.

Request concerning

XML

Discussion history

2 Feb 2004,2 Mar 2004,12 May 2004,18 Apr 2006,12 Feb 2007

Transition history

raised on 12 Jan 2004 by TBL, on behalf ofTAG

accepted on 2 Feb 2004

Background, proposals, threads, notes

• 1 Mar 2004: XML Core WG discussionof issue

agreed on 7 Mar 2007

xmlChunk-44 was an attempt to tackle deepequals for XML. TheTAG now thinkwe can't do better than XML Functions andOperators.

Acknowledgment cycle

announced by group on 12 Jul 2007

Action history

mediaTypeManagement-45: What is the appropriate level of granularity of themedia type mechanism? [link to this issue]

The TAG raised this issue to discuss issuessuch as paramterization of media typestrings to provide additional granularityfor different format versions.

Request concerning

RFC 2045

Discussion history

14 May 2004,4 Oct 2005

Transition history

raised on 14 May 2004 by TBL, on behalf ofTAG

accepted on 14 May 2004

Background, proposals, threads, notes

• See also issueuriMediaType-9
• See alsoCompound Document Formats WG.

Action history

xml11Names-46: Impact of changes to XML 1.1 on other XML Specifications [link to this issue]

XML 1.1 makes essentially four changes toXML 1.0:

• It increases the number ofcharacters that may legally appearin Names.
• Adds several new characters that mayappear in text if they are encodedas numeric character references (C0controls except NUL).
• Removes several characters so thatthey may not appear in text if theyare not encoded as numeric characterreferences (C1 controls).
• Adds as a line-end character.

XML Schema 1.0 normatively refers to XMLNamespaces 1.0 for the definition of QNameand XML Namespaces 1.0 normatively refers toXML 1.0 for the definition of Name and XML1.0 has fewer Name characters than XML 1.1.

That means that by a strict interpretationof the Recommendations, it is impossible towrite an XML Schema for a document that usesthe "new" Name characters. And by extension,it is impossible for an XPath expression ora protocol document to use XML 1.1.

Request concerning

XML 1.1

Discussion history

7 Jun 2004,28 Jun 2004,4 Oct 2005,31 Jan 2006,13 Dec 2006

Transition history

raised on 19 May 2004 by Mark Nottingham, on behalf ofXML Protocol WG

accepted on 7 Jun 2004

Action history

endPointRefs-47: WS-Addressing SOAP binding & app protocols [link to this issue]

From the commentersemail:

"In a nutshell, it [WS-Addressing - SOAP Binding] requires that the URI in the "Address"component of a WS-Addressing EPR beserialized into a wsa:To SOAP header,independent of the underlying protocol. IMO,a Web-architecture consistent means of doingthis would be to serialize it to theRequest-URI when using SOAP with HTTP, orthe "RCPT TO:" value when using SOAP withSMTP, etc.."

The issue has been raised with the relevantWGanddeclined.

TheWS-Addressing SOAP Binding CRof 17 Aug 2005 still has this problem.

Request concerning

Web Services Addressing - SOAP Binding

Discussion history

24 Jan 2005,29 Mar 2005,5 Apr 2005,19 Apr 2005,22 Sep 2005,4 Oct 2005,11 Oct 2005,18 Oct 2005,25 Oct 2005,1 Nov 2005,22 Nov 2005,6 Dec 2005,24 Jan 2006,28 Mar 2006,14 Jun 2006,24 Oct 2006,11 Dec 2006,12 Dec 2006,9 Jan 2007

Transition history

raised on 3 Jan 2005 by Mark Baker

accepted on 24 Jan 2005

Background, proposals, threads, notes

• decisionby the WS-Addressing WG
• A Worked Example, by H. Thompson
• WS-Transfer and HTTP, re TAGIssues whenToUseGet-7 &endPointRefs-47

Action history

nameSpaceState-48: Adding terms to a namespace [link to this issue]

The question is about the identity of anamespace, in particular, the xml:namespace. One perspective is that the xml:namespace consists of xml:space, xml:lang,and xml:base (and no other names) becausethere was a point in time in which thosewhere the only three names from thatnamespace that had a defined meaning.Another perspective is that the xml:namespace consists of all possible localnames and that only a finite (but flexible)number of them are defined at any givenpoint in time.

Request concerning

XML

Discussion history

22 Feb 2005,8 Mar 2005,22 Sep 2005,6 Dec 2005,13 Dec 2005,20 Dec 2005,18 Apr 2006,25 Apr 2006

Transition history

raised on 9 Feb 2005 by Norman Walsh, on behalf ofXML Core WG and XML Coordination Group

accepted on 22 Feb 2005

Background, proposals, threads, notes

• Draft finding from NW
• Approved finding
• TR versionof the approved finding

agreed on 25 Apr 2006

Approved TAG findingThe Disposition of Names in an XMLNamespace.

Acknowledgment cycle

announced by group on 5 May 2006

Action history

schemeProtocols-49: Relationship of URI schemes to protocols and operations [link to this issue]

There are many situations in which thereappeared to be confusion in the Webcommunity regarding the architecturalrelationship between URI schemes and networkprotocols.

Request concerning

URI

Discussion history

31 May 2005,15 Jun 2005,28 Jun 2005,4 Oct 2005,5 Dec 2005,26 Sep 2006,23 Jan 2007

Transition history

raised on 7 Feb 2005 by Noah Mendelsohn

accepted on 22 Feb 2005

Background, proposals, threads, notes

• URI Schemes and Web Protocolsfrom Noah Mendelsohn

Action history

URNsAndRegistries-50: URIs, URNs, "location independent" naming systems and associated registries for naming on the Web [link to this issue]

This issue covers a) URIs for namespacenames b) URNs and other proposed systems for"location independent" names c) XML andother registries, and perhaps centralizedvs. decentralized vocabulary tracking.

Request concerning

• URI
• RFC 3688

Discussion history

15 Mar 2005,22 Mar 2005,29 Mar 2005,5 Apr 2005,26 Apr 2005,10 May 2005,4 Oct 2005,11 Oct 2005,6 Dec 2005,18 Apr 2006,6 Jun 2006,12 Jun 2006,13 Jun 2006,25 Jul 2006,15 Aug 2006,29 Aug 2006,26 Sep 2006,5 Oct 2006,23 Jan 2007,30 Apr 2007,14 May 2007,30 May 2007

Transition history

raised on 15 Mar 2005 by Henry Thompson

accepted on 15 Mar 2005

Background, proposals, threads, notes

• 29 Mar 2005:Draft finding

Action history

standardizedFieldValues-51: Squatting on link relationship names, x-tokens,registries, and URI-based extensibility [link to this issue]

Since short strings are scarce resourcesshared by the global community, fair andopen processes should be used to managethem. A pattern that I'd like to see more ofis

1. start with a URI for a new term,
2. if it picks up steam, introduce asynonym that is a short string thrua fair/open process.

Lately I'm seeing quite the opposite.

Nearby issues:

• uriMediaType-9
• URNsAndRegistries-50
• XMLVersioning-41
• nameSpaceState-48

Request concerning

• URI
• namespace

Discussion history

10 May 2005,31 Jan 2006

Transition history

raised on 6 Apr 2005 by Dan Connolly

accepted on 19 Apr 2005

Action history

passwordsInTheClear-52: Sending passwords in the clear [link to this issue]

Many applications send passwords in theclear. This raises obvious security issues.The TAG should recommend not to sendpasswords in the clear and proposealternatives.

Request concerning

• Security
• Authentication

Discussion history

15 Jun 2005,20 Sep 2005,18 Apr 2006,13 Jun 2006,26 Sep 2006,4 Oct 2006,10 Oct 2006,21 Nov 2006,11 Dec 2006,9 Jan 2007,23 Jan 2007,25 Jun 2007

Transition history

raised on 18 Apr 2006 by Dan Connolly

accepted on 18 Apr 2006

Background, proposals, threads, notes

• Discussions atSept. 2005 TAG f2f
• W3C Workshop on Transparency andUsability of Web Authentication
• public-usable-authenticationmailing list
• Draft findingPasswords in the Clear

Action history

genericResources-53: Generic resources [link to this issue]

A generic resource is a conceptual resourcewhich may stand for something which hasdifferent versions over time, differenttranslations, and/or different content-typerepresentations. How should one indicate therelationship between these?

Request concerning

Web resources

Discussion history

9 May 2006,30 May 2006,12 Jun 2006,18 Jul 2006,29 Aug 2006,19 Sep 2006,4 Oct 2006,31 Oct 2006,7 Nov 2006

Transition history

raised on 4 May 2006 by T. V. Raman

accepted on 30 May 2006

Background, proposals, threads, notes

• Generic Resourcesin Design Issues
• On Linking Alternative FormatsTo Enable Discovery AndPublishing

agreed on 31 Oct 2006

See TAG findingOn Linking Alternative RepresentationsTo Enable Discovery And Publishing.

Acknowledgment cycle

announced by group on 31 Oct 2006

Action history

TagSoupIntegration-54: Tag soup integration [link to this issue]

Is the indefinite persistence of 'tag soup'HTML consistent with a sound architecturefor the Web? If so, what changes, if any, tofundamental Web technologies are necessaryto integrate 'tag soup' with SGML-valid HTMLand well-formed XML?

Request concerning

Web resources

Discussion history

24 Oct 2006,31 Oct 2006,7 Nov 2006,11 Dec 2006,12 Dec 2006,5 Feb 2007,7 Mar 2007,7 Mar 2007,19 Mar 2007,26 Mar 2007,16 Apr 2007,23 Apr 2007,31 May 2007

Transition history

raised on 17 Oct 2006 by TAG

accepted on 24 Oct 2006

Background, proposals, threads, notes

• Description of new TAG issue

Action history

utf7Encoding-55: Security issues with incorrect metadata [link to this issue]

Security concerns with browsers sniffingunlabelled UTF7 encoding

Request concerning

• Security
• Metadata

Discussion history

2 Jan 2007,9 Jan 2007,12 Feb 2007

Transition history

raised on 14 Dec 2006 by Roy Fielding

accepted on 2 Jan 2007

Background, proposals, threads, notes

• A description of the problem
• See alsoCVE-2006-5442

Action history

abbreviatedURIs-56: Abbreviating URIs in Web Languages [link to this issue]

Do the expected benefits ofCURIEsoutweigh the potential costs in introducingathirdsyntax for identifiers into the languages ofthe Web?

This issue continues a thread of discussionthat originated underultimateQuestion-42.

Request concerning

• CURIE WD 7th March 2007
• Henry Thompson'semailframing the issue.

Transition history

raised on 30 Mar 2007 by TAG

accepted on 2 Apr 2007

Action history