See also: IRC log
<scribe> Scribe: DanC
<DanC_> minutes 7 Feb
<DanC_> minutes 7 Feb good enough for me
RESOLUTION: to accept minutes 7 Feb
RESOLUTION: to meet again 21 Feb, NDW to scribe
regrets timbl 21 Feb
<DanC_> meeting page
DO: I'm working on the state finding... how about that for the agenda?
DC: suggest moving metadataInURI-31 after the other 3 technical things
NM: if we don't finish least power, it might merit ftf discussion. leave it off for now, if the agenda is fluid.
"Monday 27 February: 13:30 - 17:30 @@@"
some sentiment for 2p, some for 1:30
RESOLUTION: to start at 13:30 on Monday, 27 Feb
NM: did we end up with any liaison meetings scheduled?
VQ: not at this time
Ed: previously we had a "what's important for the coming year" session... shall we do that again?
HT: I prefer the current contents of the agenda to that sort of thing
DC: me too
NM: perhaps make some time to chat with TV, but otherwise, yes, technical topics
TBL: hmm.. indeed, looking forward would be good... do we have a social time scheduled? it's hard to swap between technical topics and looking ahead
DC: perhaps the "what did we learn this week?" session will be sufficient?
HT: I'm constrained to Monday evening for an evening thing
NM: I'll be on US east coast time, so not too late
<scribe> ACTION: VQ organize a monday evening quiet social event [recorded in http://www.w3.org/2006/02/14-tagmem-irc]
VQ: around 7pm
DC: there's a convention of publishing on /TR/ at least every 3 months. We haven't done it in over a year. I'm inclined to take something and publish it. Maybe the next one we approve... say, didn't we just approve one recently?
NDW: yes, the ns48 finding is approved
TimBL: how about concatenating the approved findings?
DC: that's more work than I'm offering now
<timbl> namespaceState finding
<Ed> list of findings
(norm, I'm inclined to work from the .html only and not bother with the xml)
<noah> Speaking of which, the approved finding link at http://www.w3.org/2001/tag/issues.html?type=1 is to the xml
<timbl> This page contains the following errors:
<timbl> error on line 17 at column 140: Entity 'http-ident' not defined
<timbl> error on line 19 at column 199: Entity 'draft.day' not defined
<timbl> error on line 20 at column 226: Entity 'draft.monthname' not defined
<timbl> error on line 21 at column 247: Entity 'draft.year' not defined
<timbl> error on line 24 at column 283: Entity 'http-ident' not defined
<timbl> error on line 27 at column 370: Entity 'http-ident' not defined
<timbl> error on line 30 at column 431: Entity 'http-ident' not defined
<timbl> error on line 33 at column 488: Entity 'http-ident' not defined
<timbl> Below is a rendering of the page up to the first error.
<Norm> What page was that timbl?
TimbL: good to put all this in the SOTD: (1) it's approved by the tag (2) this is one of many issues in the TAG's list (3) The eventual disposition of this text is not cler, but one possibility is it being integrated wioth other finids into a new AWWW or a second volume AWWW
PROPOSED: to publish http://www.w3.org/2001/tag/doc/namespaceState.html as a W3C Working Draft
<Norm> Uhm, with what shortname?
<timbl> TAG-namespaceState
namespaceState
RESOLUTION: to publish http://www.w3.org/2001/tag/doc/namespaceState.html as a W3C Working Draft
<scribe> ACTION: NDW to with DanC, publish WD of ns48 finding [recorded in http://www.w3.org/2006/02/14-tagmem-irc]
VQ: this is just one finding; we'll see what we learn from this
<Norm> What publication date should we use for namespaceState, DanC ?
<DanC_> dunno
<Norm> Returning to the publication of namespaceState, I chose 23 Feb as the publication date because that's the last day before the moritorium.
<DanC_> least power finding, latest version
<DanC_> 13 Feb draft
NM: there has been much www-tag discussion of chomsky hierarchies and complexity...
<noah> There are many dimensions to language power and complexity that should be considered when publishing information. For example, a language with a straightforward syntax may be easier to analyze than an otherwise equivalent one with more complex structure. A language that wraps simple computations in unnecessary mechanics, such as object creation or thread management, may similarly inhibit information extraction. The intention of this finding is neither to rigor
<noah> necessarily interferes with information reuse. Rather, this finding observes that a variety of characteristics that make languages powerful can complicate or prevent analysis of programs or information conveyed in those languages, and it suggests that such risks be weighed seriously when publishing information on the Web.
<noah> Indeed, on the Web, the least powerful language that's suitable should usually be chosen. This is The Rule of Least Power:
<noah> Good Practice: Use the least powerful language suitable for expressing information, constraints or programs on the World Wide Web.
HT: why not just say "occam's razor applies to computers too"?
DanC: yes, the principle is 2
lines, but what we add is to relate it to the history of web
technology development.
... e.g. how HTML is and why
TimBL: yes, examples. CSS vs
javascript.
... the fact that you can cascade to CSS stylesheets is a
result of a decision to make it declarative
<Zakim> DanC, you wanted to suggest going specific-to-general
<ht> HST doesn't understand why Turning-completeness is bad
<ht> Prolog is Turing-complete, and dead easy to analyze!
<DanC_> hmm... I thought validator.w3.org would be impossible/impractical if the web had used TeX rather than HTML
<ht> SQL is Turing-complete (or close), and probably more analyzed than almost any other language
<Zakim> noah, you wanted to discuss scope of this rewrite...are we thrashing?
the analysis of SQL is precicely on the bits that are *not* turning complete, no?
<raman> belated regrets -- I shamefully admit that I just plain forgot to call in...
<DanC_> VQ, I suggest a straw poll: how many think it's reasonable to approve as is.
<noah> If you have a Turing-complete program, you don't in general know whether it even gets done
<noah> If I have a table in a relational database, or a list of name/value pairs, I don't have that problem.
DC: due to the halting problem, Most of the other things you want to know follow from it.
PROPOSED: to approve "The Rule of Least Power" as 12 Feb draft, incorporating edits agreed by from NDW and NM.
HT: some discomfort, but I concur if others are OK
TBL: likewise, I concur.
RESOLUTION: to approve "The Rule of Least Power" as 12 Feb draft, incorporating edits agreed by from NDW and NM.
<scribe> ACTION: NM to announce approved least power finding, when discussion with NDW concludes [recorded in http://www.w3.org/2006/02/14-tagmem-irc]
DO: I did some work on
this...
... sent them to the schema WG a few weeks ago
... haven't seen [which?] draft posted as I expected
... I hope to talk with interested people at the TP in
France
... so I think this is done
DC: pointer?
HT: getting it public has taken a back seat to other things
NM: I think we have license to make this public already
HT: yes, if you can follow up, that would be fine
<scribe> ACTION: DO to contextualize his scenarios, such as more on what is happening with SOAP and WSDL [DONE] [recorded in http://www.w3.org/2006/02/14-tagmem-irc]
<scribe> ACTION: DO to with NM continue and extrapolate the versioning work DO et al have been doing already, updating the terminology section. [CONTINUES] [recorded in http://www.w3.org/2006/02/14-tagmem-irc]
<DanC_> terminology section update from DO 13 Feb
DO: I got some comments re
first/last name from Misha
... main list is public-xml-versioning
DC: hmm... public-xml-versioning... partial understanding isn't limited to xml
DO: public-xml-versioning was created at the suggestion of the TAG as a mechanism for collaboration with XML Schema WG.
NM: [... about broadening from xml-specific story to a story about strings, with markup as a special case]
<DanC_> (which appeals to me)
<Zakim> ht, you wanted to incline towards focussing on XML language
DO: broadening makes sense to some extent, but there's a limit, and we need to be sure to deliver for XML authors
<noah> Isn't URI an example of a non-QNamed namespace
<DanC_> (surely notation 3 is a webized language that's not XML)
<DanC_> (webized meaning: has its terms grounded in URI space)
<ht> DanC, remind me what N3's media type is?
<ht> I.e., can I follow-my-nose to find out about N3?
<DanC_> text/n3+rdf or some such; registration pending
<noah> I thought we set up in Edinburgh that versioning was about the conclusions drawn by a consumer and a producer for any particular document, where the two parties have imperfect agreement on the language they thought they were using.
<noah> I like that start a lot, and it's not XML-specific
HT: [...] XML gives us the "follow your nose" principle, with namespaces
<noah> Follow your nose seems to give you something very important, which is self description. I'm not convinced that versioning should be only about self-describing documents.
TBL: all stories about versioning
depend on a notion of semantics/meaning...
... at the level of XML, there is only a basic infrastructure.
At higher levels, e.g. HTML and RDF, there's more to say
DC: meanwhile, I have a new .violet file from DO that I intend to check against my changePolicy.n3 work
TBL: I wonder about a 4 part finding:
(1) at the level of representations
(2) at the level of namespaces in XML
(?) [...] in HTML and such
(4) an one about RDF
NM: about strings of characters?
TBL: that's what I meant by (1)
<noah> Cool.
DO: let's please have some discussion on public-xml-versioning of the new terminology section
<DanC_> +1
<Norm> +1
<DanC_> hmm... the archive cover page of http://lists.w3.org/Archives/Public/public-xml-versioning/ doesn't say that it's a joint tag/xml-schema thingy
VQ: with regret, it's time to
curtail this discussion
... maybe next time we'll get to ns8
<Norm> I'll try to get back to Jonathan and make progress on ns8 for next week
ADJOURN.