Setting TAG Priorities - Review of discussion and planning to date

Shortly I will send out a note suggesting next steps for setting TAG 
priorities.  As background to that, I have tried to gather here pretty 
much everything I could find regarding progress so far, I.e. links to or 
copies of F2F notes, e-mails TAG members have sent with suggestions on 
organizing our work, etc.  So, although this is a very long email, it's 
primarily intended as reference material.  Read it all if like or else 
refer to it as needed while working on the items in the note to follow.  I 
do suggest you at least skim all of this so you'll know what's here if you 
need it.

This note is intended primarily for members of the TAG.  Nothing here is 
member-only, so I'm sending it to www-tag in case anyone else is 
interested in following our efforts to organize or work.

F2F Review
----------

Looking back over our sessions at the F2F, we made progress in the three 
areas I hoped we would, that is:

* Reviewing ongoing work and clarifying the status of open issues

* Gathering and discussing ideas for specific work we might do in the 
future

* Thinking about organizing our work under broad themes vs. more isolated 
initiatives

Each of those efforts is at this point incomplete.  We only reviewed in 
detail and updated descriptions for about 1/3 of our open issues, we took 
only a first pass at identifying larger themes, and we haven't yet shown 
that organizing some or all of our work under these themes will actually 
focus us on high value deliverables.  Indeed, in parallel with the above 
activities, we have had at least one short term item jump to the head of 
our agenda (discussing HTML in time for the AC meeting), and we've 
received input from members who have indicated preferences for working on 
particular ongoing activities [1].

FWIW, I'm relatively pleased with where we stand, unsettled as it is.  I'm 
not prejudging what the right mix of top down or bottom up will be, and 
I'm not surprised that one round of work at a F2F left is at best half 
done in this round of planning.

Open Issues
-----------

Open issues are listed in tracker at [2].  Tracker doesn't let users 
create queries and tailor reports on the fly, so I've manually updated my 
own table of open issues including pertinent status information from the 
F2F.  It's available at [3], and for easy reference by those who use HTML 
email readers, it's also copied inline below (Appendix I).  I've indicated 
which issues have had their descriptions updated in tracker, who the 
shepherds are, and most importantly, what the proposed priority (rank) is. 
 Of the ten issues that were considered in detail, two (errorHandling-20 
and contentTypeOverride-24) were given proposed ranks of high.  Three more 
were medium.  We also opened a new issue, UniformAccesstoMetadata-62, but 
did not give it a rank.

Information Gathered at the F2F
-------------------------------

At one point early in the meeting, we went around the room asking for 
ideas on what would constitute success for the TAG, who our audience might 
be, and what might be good things for the TAG to work on.  That list is 
available at [4], and for your convenience, a copy is included following 
the issues list at the end of this note (Appendix II).

Later we prioritized some of those possible work items and gathered a 
table.  That was edited in a spreadsheet and saved as a csv text file [5]. 
 The file includes indications of which TAG members were interested in 
each issue. (Appendix III)

Our last planning exercise was to try and identify key themes, with ideas 
under them for what we might actually do.  Those notes were gathered on 
flip charts.  Transcriptions are in reference  [6] and in Appendix IV.


Suggestions received in email, meetings, etc.
---------------------------------------------

Some TAG members have sent emails with suggestions on what we should do or 
how we should set priorities.  It's quite likely that I've lost track of 
some, but these are the ones I've noted.  If there are some I've missed, 
please let me know.  I generally list only the first in a thread.  Most of 
what I found were in threads started by Larry:

* Endorses "operational approach" [7]
* Priorities and topics for discussion (some of these did come up at the 
F2F - initial member-only copy at [8], later publicly archived at [9])
* Followup on metadata in particular [10] (public copy at [11])

In addition to email, some suggestions have been made in meetings.  I 
haven't been through all the recent minutes in detail, but I notice in 
particular from Tues at the F2F [12]:

HT: What I want us to do is: "produce a document which specifies a 
mechanism for bridging the gap between namespaced and non-namespaced forms 
of languages conveying the same things"

If I've missed or misrepresented anything, please let me know.  I hope 
that collecting all this will give a sense of the different angles we've 
explored so far, and that it will be useful as reference material.  Now 
the challenge will be to turn this into an operational agenda for what we 
actually want to do in the coming months.  Please see email to follow on 
what I think we should do next.  Thank you.

Noah

References:

[1] http://lists.w3.org/Archives/Member/tag/2009Mar/0030.html 
[2] http://www.w3.org/2001/tag/group/track/issues/open 
[3] http://www.w3.org/2001/tag/2009/03/03-TAG-issue-status.html 
[4] http://www.w3.org/2001/tag/coordination/groupPriority.txt
[5] http://www.w3.org/2001/tag/coordination/groupPriority.csv
[6] http://www.w3.org/2001/tag/2009/03/05-whiteboard-priorities.txt
[7] http://lists.w3.org/Archives/Public/www-tag/2009Jan/0113.html
[8] http://lists.w3.org/Archives/Member/tag/2009Feb/0016.html
[9] http://lists.w3.org/Archives/Public/www-archive/2009Feb/0105.html
[10] http://lists.w3.org/Archives/Member/tag/2009Feb/0036.html
[11] http://lists.w3.org/Archives/Public/www-archive/2009Feb/0106.html 
[12] http://www.w3.org/2001/tag/2009/03/03-tagmem-minutes.html#item04


--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------

Appendix I:

TAG Open Issue Status as of 20 March 2009
The following list was extracted from Tracker's list of open issues. The 
"F2F" column indicates issues that appeared as explicit agenda items for 
the 3 March 2009 F2F.

ISSUE #
F2F
Tracker
Updated
F2F
Rank
Shepherd
Old Issue Name
Description
Raised on
Number of 
Open Actions
7

Y
Background
Kemp
whenToUseGet-7
(1) GET should be encouraged, not deprecated, in XForms(2) How to handle 
safe queries
23-Jan-02
0
16

Y
PENDING REVIEW
Masinter
HTTPSubstrate-16
Should HTTP be used as a substrate protocol? Does W3Cagree with RFC 3205?
24-Mar-02
0
20
X
Y (except shepherd)
High
No shepherd - raised by Berners-Lee
errorHandling-20
What should specifications say about error handling?
22-May-02
0
24

Y
High
Masinter
contentTypeOverride-24
Can a specification include rules for overriding HTTPcontent type 
parameters?
14-Jun-02
1
27

Y
Medium
None - Raised by Berners-Lee
IRIEverywhere-27
Should W3C specifications start promoting IRIs?
9-Oct-02
2
30

Y
Medium
Mendelsohn
binaryXML-30
Standardize a "binary XML" format?
9-Oct-02
1
33

Y - a bit sketchy
Background
None - Raised by Berners-Lee
mixedUIXMLNamespace-33
Composability for user interface-oriented XML namespaces
6-Feb-03
0
34
X
Y
Medium
Thompson
xmlFunctions-34
XML Transformation and composability (e.g., XSLT,XInclude, Encryption)
6-Feb-03
1
36

Y
Background
None - Raised by Connolly
siteData-36
Web site metadata improving on robots.txt, w3c/p3p and favicon etc.
10-Feb-03
0
37

Y
Background
None - Raised by Berners-Lee
abstractComponentRefs-37
Definition of abstract components with namespace names and frag ids
3-Feb-03
0
39




rdfURIMeaning-39
Meaning of URIs in RDF documents
13-Jul-03
0
40




URIGoodPractice-40
What are good practices for URI construction?
16-Oct-03
0
41




XMLVersioning-41
What are good practices for designing extensible XMLlanguages and for 
handling versioni
27-Jun-03
3
42




ultimateQuestion-42
What is the answer to life, the universe, and everything.
15-Nov-03
4
43




DerivedResources-43
How are secondary resources derived?
15-Nov-03
0
45




mediaTypeManagement-45
What is the appropriate level of granularity of the media type mechanism?
14-May-04
0
46




xml11Names-46
Impact of changes to XML 1.1 on other XML Specifications
19-May-04
0
47




endPointRefs-47
WS-Addressing SOAP binding & app protocols
3-Jan-05
0
49




schemeProtocols-49
Relationship of URI schemes to protocols and operations
7-Feb-05
0
50
X



URNsAndRegistries-50
URIs, URNs, "location independent" naming systems and associated 
registries for namin
15-Mar-05
2
53
X



genericResources-53
Generic resources
04-May-06, reopened at 3 March 2009 F2F
2
54
X



TagSoupIntegration-54
Tag soup integration
17-Oct-06
0
55




utf7Encoding-55
Security issues with incorrect metadata
14-Dec-06
0
56




abbreviatedURIs-56
Abbreviating URIs in Web Languages
30-Mar-07
0
57
X



HttpRedirections-57
The use of HTTP Redirection
22-Aug-07
2
58




scalabilityOfURIAccess-58
Scalability of URI Access to Resources
22-Aug-07
1
60
X



webApplicationState-60
Web Application State Management
14-Dec-07
3
61




uriBasedPackageAccess-61
URI Based Access to Packaged Items
8-Jul-08
0
62



Malhotra
UniformAcessToMetadata-62
Uniform Access to Metadata
04-Mar-09
0


Appendix II:

These notes were generated during discussion by the W3C Technical
Architecture group during its F2F meeting of 3-5 March 2009.  These
are working notes, intended primarily for internal use by the TAG.
They do not necessarily represent the actual opinions of the TAG as to
success criteria, group priorities, etc.

What is success for the TAG?
============================

* Helping communities be more effective.  Help where perspective is
  missing, or might lack background to solve.

* Coordinating review of documents within W3C and outside, noticing
  mismatches between what groups need from each other and provide each
  other.

* See evidence of communities who have responsibility for parts of the
  Web that they recognize having that responsibility.

* Cause new activites, with good value, to spring pu in W3C

* Good coverage of the architecture without bits out. The Web works.

* Happy ending to XML/HTML.

* Necessary and sufficient: TAG to evidence leadership in topics of
  most concern to the Web community (especially the W3C).  Something
  recognized as causing people (especially W3C WG participants) to
  follow.  Can measure effectiveness using TPAC surveys as to whether
  TAG was more help this year than last.

* Win back technical credibility.

* Open, Web-standard technologies continnue to win for Web-content.

* Making the Web better -- be an advocate for the greater
  good. Effectively resolve tensions between working group short term
  interests and serving greater good.

* Javascript security less of a black art.

* XRI TC publishes drafts based on directions emerging from our
  discussions (I.e. no new URI scheme)

* We have an integrated architecture for Web and Semantic Web.

* Explain Web architecture in such a way that a random architect can
  discover which specs are pertinent and can succeed in using them
  effectively.  We need a natural language explanation of >why< things
  are the way they are.  (Integrating both in one WebArch document has
  proven confusing.)

* Demonstrate that the architecture we propose leads to good
  engineering long term.

* Lead in areas where Web is going: e.g. Web 2.0 and successors,
  video, mobile, new devices.

* Ensure that Web technologies provide near state of the art
  capabilities for animation and other modern UI.

* Dealing with services web as opposed to endpoint Web.  Internet
  scale services is a new phenomenon.

* Watch for and document cost/benefit changes in what have been core
  assumptions and technologies: e.g. # replacing ? for state
  management.

* Spec writers use vocabulary we define

What audiences should we address?
=================================

* Anyone who creates or uses Web technologies. Spec authors, web site
  admins, content creators, ....

* People who are building core technologies (e.g. object capability
  security, and semantic web semantics). Language, API and protocol
  designers.

* Spec writers

* W3C working groups, possibly OASIS TC's

* W3C, and peers (IETF, OASIS, etc.)

* Formal technical liaison with IETF & OASIS, etc. where
  possible. Should have new liaisons as necessary.

* 3 parts: whom you want to reach; the channel you use; community you
  survey to guage effectiveness

* anyone who writes or reads a Web spec.

What should we do next year? What should we produce?
====================================================

* Cluster of related issues to write about or delegate to some group
  (or actively decide not to deal with):
   - Metadata
   - Understand relation between HTTP and RDF
   - Scheme/protocols
   - Naming systems in relation to HTTP and other protocols

* Push information resource / resource debate to a conclusion

* Convert architecture document into foil set and notes.

* Harmominization of XHTML and HTML from DOM layer up.  (We don't
  necessarily produce it, but help the community to succeed.)

* Say something about versioning and/or error handing, if only to
  document history.

* Avoid doing things we can't complete

* Put as low priority things that have only short term value.  High
  priority to things that produce documents that have long term value.

* We should emphasize activities that produce "artifacts" of long term
  value (findings, etc.)

* Web app state

* Javascript

* Mobile

* xmlFunctions-34 and semantics of recursively structured documents

* Web security - details TBD

* Use semantic Web technologies such as ontologies to describe Web
  architecture.

* Use formal rules to document TAG conclusions

* Widget packaging

* Architectural issues that affect HTML 5
   - Versioning
   - Error handling
   - Extensibility
   - Tag soup
   - XML vs. XML 5
   - Technical specifications vs. applicability statements
   - HTML as universal Web content vs. extensibilty via plugins.

* Resource description access & redirections

Stuart preferences:
  uniform access to metadata
  webApplicationstate
  distributed extensibility vs. monolithic specs.

Member interests:
  - HTML
  - Web Services - (unclear whether this meant WS* or more broadly)
  - Semantic Web
  - Potholes: finishing things (e.g. file: URI scheme -- why is Web
              apps inventing whole new URI)


What else should we talk about wrt/
TAG priorities?
===================================



Appendix III: (from the csv file)

Number
Cluster
Who
Topic
27
*
LM, Dan
Liason with other organization
4
A
HT, Tim, JK, LM, TVR
Harmominization of XHTML and HTML from DOM layer up.  (We don't 
necessarily produce it, but help the community to succeed.)
6
A
Dan, Tim, LM
HTML
17
A
HT, LM, TVR
Architectural issues that affect HTML 5,- Versioning,- Error handling,- 
Extensibility,- Tag soup,- XML vs. XML 5,- Technical specifications vs. 
applicability statements,- HTML as universal Web content vs. extensibilty 
via plugins.
21
A
Tim, TVR, HT
distributed extensibility vs. monolithic specs.
25
A
HT
ErrorHandling-20
1
B
Jar, Tim, LM?
Cluster of related issues to write about or delegate to some group (or 
actively decide not to deal with):  ,- Metadata,- Understand relation 
between HTTP and RDF,- Scheme/protocols,- Naming systems in relation to 
HTTP and other protocols
2
B
Jar, HT
Push information resource / resource debate to a conclusion
18
B
Jar
Resource description access & redirections
19
B
Jar, Ashok, Tim
uniform access to metadata
28
B

Conneg
23
C

Scheme/protocols
24
C
HT, JAR
URNS and Registries
8
D
Tim
Semantic Web including integration of SemWeb with Web as a whole (e.g. 
vocabularies, linked data synchronization and update)
10
E
Jar?, Dan, JK,, LM?, TVR
Javascript security
13
E
Jar?, Ashok, JK
Web security - details TBD
3
TW
Ashok, JK
Convert architecture document into foil set and notes.
15
TW
Tim
Use formal rules to document TAG conclusions
5

HT, JK, LM?
Say something about versioning and/or error handing, if only to document 
history.
7

Ashok?
Web Services
9

LM
Potholes (e.g. file: URI)
11

LM?
Mobile
12

HT, Tim
xmlFunctions-34 and semantics of recursively structured documents
14

Jar, Tim, LM?
Use semantic Web technologies such as ontologies to describe Web 
architecture.
16

JK, LM?
Widget packaging
20

Dan, TVR
webApplicationstate
22


******
26


contentTypeOverride-24
27
D

Viewing AWWSW as TAG offshoot to set down HTTP semantics using OWL 
ontologies, etc.


Appendix IV: Themes from Flip Charts on last day of F2F

This is what we collected on the flip charts during our wrapup session on 
priorities at the F2F.  I suggest these be included by the scribe at a 
suitable point in the minutes.  Maybe or maybe not it makes sense to show 
each section at the head of the discussion of that topic.  For your 
reference, the original photos of the flipcharts are attached.

The capital letters identify the major topics, each of which had one flip 
chart:

---------------------------

A. HTML Theme

* Liaison: to reduce conflict
- IETF

* Findings:
- Error handling (Postel's law / exceptions)

* Distributed extensibility arguments (in ESW wiki)

* Harmonize XHTML + HTML 
- Understand both sides

* Media-type based namespace defulating, etc. (bridging the gap)

* Education for:
- Ourselves
- HTML Group
- (note, a bi-directional arrow connected this with the distributed 
extensibility bullet above)

* Specification layering: orthogonality
- finding

---------------------------

B. Metadata access and HTTP use

* With respect to conneg questions, respond "no" (chairs explanation:  I 
believe this was meant to convey "no, you should not use conneg except 
when the reprsentations convey substantially the same information)

* Firefighting
- conneg

* Set out core principles (Larrry) 
- or -
* Go bottom up (Tim)

* Liaison
- Help with the DRD draft
- Help with the link header draft
- Help with the site metadata draft

* Review Jonathan's metadata summary

* Check the architectural soundness of the 3 documents mentioned above

* Write down principles: e.g. can HTTP URI's "speak to" URI's such as 
"mailto"

---------------------------

C. URI's; Naming; Meaning of Names

* Note: HT + JAR commited to new drafts

* Members asked us:
- what are tradeoffs between using http vs. new naming system (e.g. URI 
schemes; URN namespaces)

* Scheme protocols: explain to community when new protocols do/don't merit 

new URI schemes

---------------------------

D. Semantic Web Architecture (e.g. combining vocabularies & linked data 
sync)

* Produce revised or augmented architecture document integrating Semantic 
Web and Linked data

---------------------------

E. Security

Javascript security

* Review capability-based systems: track them

* Track cross-site scripting

* Influence AC to invest in coverage of security - identify holes

* Survey applicability of drafts such as:
- Origin header (Barth / Hixie)
- Security context working group / UI work

* Try to greatly increase focus on security at W3C:
- update AWWW for security
- Liaison w/Web security organizations such as IETF security directorate
- Focus phishing

---------------------------

F. Mobile

* Special concerns of mobile security

* Survey workshop report (see F2F minutes for link)

* Hear from platform providers

* Device independence in era of mobile

* Liaison

---------------------------

Received on Sunday, 22 March 2009 15:01:15 UTC