See also: IRC log
<scribe> Scribe: Hugo
<dorchard> I'll call in for the serialization part of the agenda.
April 7 telcon minutes approved
Jonathan: A lot of action items were due today, and unfortunately, quite a few weren't done
[ Chair calling names of people not on the call ]
scribe: Jacek, how about the RDF mapping?
Jacek: I think that this action item can be retired
<Marsh> ACTION: Marsh to put RDF mapping on the agenda for next week [recorded in http://www.w3.org/2005/04/14-ws-desc-minutes.html#action01]
Jonathan: I'll ask Bijan for a status update next week
<Marsh> ACTION: WG to despair over the RDF mapping WG. [recorded in http://www.w3.org/2005/04/14-ws-desc-minutes.html#action02]
Jonathan: Microsoft is not planning on support the HTTP binding in Indigo (as a response to "DaveO to query MS whether they would support part of the HTTP binding if divided, due 2005-04-13.")
[ Chair continues going through AIs ]
<scribe> ACTION: Arthur to evaluate text from Amy on schemaLocation; due 2005-04-21 [recorded in http://www.w3.org/2005/04/14-ws-desc-minutes.html#action03]
Jonathan: w.r.t. the Berlin F2F,
we're not meeting on the Monday anymore
... that should be helping people to get there in time
<dbooth> s/zakim,/zakim/ (dbooth has no interest in 420)
Jonathan: BTW, HP has resigned from the WG
Jonathan: I sent out a draft
agenda
...
http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0063.html
<DaveO> could we move the HTTP issues earlier? I have a 2pm flight on Friday and we usually slip the schedule..
Jonathan: I'm going to try to
limit discussion on issues to say 30 minutes
... if we can't reach a conclusion, I'll send it back to the
list
Jonathan: I added LC119 & LC120
WG agrees to close issues listed in the agenda
WG agrees to refer issues listed in the agenda to the editors
Jonathan: Arthur, any comments about LC113?
Arthur: I think that Glen has to work on this
Jonathan: is Arthur's solution OK with the WG?
Asir: is my suggestion in there?
Arthur: yes
RESOLUTION: LC113's resolution accepted
Agreement to do it next week at the F2F
Jonathan: we have 5 proposals on the table (see agenda)
Tom: I like the idea of ignoring
the namespace prefix, but I'm worried about the risks
... otherwise, disallowing them seems good to me
DaveO: not allowing ns's is going to prevent a large number of use cases
Jonathan: the issues come up when you're mixing ns's
Asir: status quo is not broken,
but needs clarification
... we can live with 2 or 4
<Zakim> dbooth, you wanted to say disallowing is safer than ignoring
David: I'm very concerned about option 2, it can bring undetected errors
Hugo: I think we should do it well or not at all, i.e. 5 or 4
Arthur I like option 5
<uyalcina> +1 to Hugo
<Zakim> JacekK, you wanted to note how option 4 is unnecessary
JacekK: you said that we only
allow local elements
... it makes option 4 unnecessary
Asir: local elements can have a namespace
JacekK: yes, but they're all the same
Asir: oh, you're right
... but you may be able to derive types, in which case you're
not right anymore
JacekK: then I think that we should limit to only one namespace then
<Zakim> DaveO, you wanted to respond to the "guarantee" issue
Jonathan: it's kind of a variation on option 4
DaveO: I didn't get Hugo's comments about guaranties
Hugo: schema doesn't know about prefixes at the abstract level
DaveO: in the simple scenario, you use the schema prefixes from the serialization of an instance
Arthur: it's unworkable
DaveO: I like option 5 to all
other options
... however, I'd like to keep it as simple as possible
Roberto: 5 is fine but too complex
<Zakim> Marsh, you wanted to propose 2a, (2) + require unique localNames
Roberto: 3 is nice from the matter
<Marsh> 2b: (2) + single namespace?
Roberto: I'd like to propose (2) + require unique localNames
Umit: does that restrict it to a single ns?
<DaveO> local names must be unique, kind of like "ID" values..
Roberto: no, you can have bar:foo and foo:bar serialized as foo and bar
<DaveO> <bar:foo><any> means <any> can't be <*:foo>
<Marsh> 2c: (2) + require names to be uniquely mappable to QNames (unique names unless order can disambiguate)
Dave: that would disallow wildcards
Roberto: wildcards are bad anyway
Jonathan: presents 2c
<Zakim> asir, you wanted to introduce something related
Asir: we serialize the HTTP
headers by ignoring the namespaces
... if we want to keep it consistant, then we should go with
option 2
Umit: we didn't know what we were doing
Asir: I think we had a similar discussion
<Zakim> dbooth, you wanted to say Uniqueness can have subtleties: Uniqueness within what context? E.g., if schema SA allows B as child of A, but schema SB does not permit B as child of A.
<DaveO> is there combination of #2 and #5, where the client can send the prefix OR the namespace name?
Arthur: we would need a computable restriction in any case for uniqueness checking
David: I agree, uniqueness can be
tricky
... but the direction is reasonable
Jonathan: it seems that we have
people unhappy with a lot of our options
... we're looking at variations of option 2, option 5 is still
on the table
Dave: I'd like to summarize the proposals and we could talk about it next week
<scribe> ACTION: DaveO to summarize LC77a options; due date 2005-04-21 [recorded in http://www.w3.org/2005/04/14-ws-desc-minutes.html#action04]
http://lists.w3.org/Archives/Public/www-ws-desc/2005Mar/0115.html
Arthur: I think it should be
required in the component model
... then there's the schema: either required or sensible
default
Roberto: the {message content model} only makes sense if you're using an XML-based data model
Jonathan: so you're arguing that we just need a clarification
Arthur: so I think that the spec should say what the meaning of empty is: other
Tom: I think that empty should be equivalent to #none
<dbooth> Roberto, so you are saying that empty == unspecified?
Jonathan: if it's missing, it means that it's using a different type system
Arthur: I'd like to withdraw my proposal and take Roberto's suggestion as an AI
<charlton> +1 to empty == different type system
DavidB: so empty is unspecified
Arthur: we should add a #other value; we're overloading the meaning of empty here
<scribe> ACTION: Arthur to present a new proposal for LC99; due 2005-04-21 [recorded in http://www.w3.org/2005/04/14-ws-desc-minutes.html#action05]
Jonathan: Asir, can you clarify anyUri's issues with IRI?
Asir: there some differences
between 2396 and 3896
... it's possible that Schema will issue an erratum for Schema
1.0
... or will do it for 1.1
Jonathan: can we be ignorant?
Asir: I think so; we could go to
the XML CG
... I've seen a proposal from XML Core about this
Jonathan: we should be able to
make a decision then
... I had some comments about where IRIs are correct
... let's finish that one up at the F2F
...
http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0057.html
http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0056.html
Jonathan: what does this give us?
Arthur: extensions are more
composable with each other
... also, all the extensions that we described follow this
pattern
... so it seems general enough
<Tomj> +1 on making the language uniform
Hugo: what's the impact on Part 2?
Arthur: be more explicit on the
types of extension components
... specify where they're extensible
Jonathan: let's decide on this next week
http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0049.html
Proposal: close LC69b with no action
RESOLUTION: LC69b closed with no action
http://lists.w3.org/Archives/Public/www-ws-desc/2005Mar/0038.html
Roberto: I want to allow element wildcard on input only
Jonathan: anybody objecting?
Asir: should we allow it on output too?
Roberto: this proposal only deals with input
Tom: this will make a lot of work for the implements
Jonathan: then please bring the issue up
This is scribe.perl Revision: 1.122 of Date: 2005/03/31 04:43:41 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/schemaLocation/schemaLocation; due 2005-04-21/ WARNING: Bad s/// command: s/zakim,/zakim/ (dbooth has no interest in 420) Succeeded: s/Asir:/Arthur/ Succeeded: s/Dave:/DaveO:/g Succeeded: s/to/with/ Succeeded: s/AU/AI/ Succeeded: s/David/DavidB/ Succeeded: s/uniformed/uniform/ Found Scribe: Hugo Inferring ScribeNick: hugo Default Present: Amelia_Lewis, T_Jordahl, Dbooth, Hugo, Allen_Brookes, Ugo_Corda, D_Moberg, Jonathan_Marsh, Roberto, Jacek, A_Ryman, Charlton_Barreto, asir, Adi_Sakala, Canon, +1.947.778.aaaa, Dave_Orchard, Anish Present: Amelia_Lewis T_Jordahl Dbooth Hugo Allen_Brookes Ugo_Corda D_Moberg Jonathan_Marsh Roberto Jacek A_Ryman Charlton_Barreto asir Adi_Sakala Canon +1.947.778.aaaa Dave_Orchard Anish Agenda: http://lists.w3.org/Archives/Public/www-ws-desc/2005Apr/0062.html Got date from IRC log name: 14 Apr 2005 Guessing minutes URL: http://www.w3.org/2005/04/14-ws-desc-minutes.html People with action items: arthur daveo marsh wg[End of scribe.perl diagnostic output]