See also: IRC log
… if we had had more WG interest/expertise in this area, we might have gone deeper. Can put it on the list for 'if we get rechartered…'
jtandy: but we can discuss the union datatypes
… he says union datatypes don't address his problem
<JeniT> https://github.com/w3c/csvw/issues/223
#223 also
gkellogg: there's a complexity on the implementations that is at this point a major barrier. Adding this right now would be troubling for fledgling implementations.
… there are other ways to consider getting in multi-typing info
… the union types does address the same problem but not in his preferred manner
jenit: I agree. I'll go back to him on the xsd schemas aspect, to try to get to the bottom of it. But I don't think we need a lot more discussion here - we have some consensus.
jtandy: he could flesh it out and try to show how it would work
gkellogg: there are 2 aspects, … one is requirement that you implement the default xsd mechanism. The other is providing space to allow an unanticipated datatype.
Where the base could be say an IRI or a compact IRI or URI instead of one of our predefined tokens. That presumably could be added without too much pain.
jtandy: that doesn't support validation though?
gkellogg: it supports it to degree we support it for anything beyond time/numbers/booleans.
…(regex something)
jenit: if we want to support that, and it is quite a nice idea, having specialized datatypes which have names (URIs), … useful in both XML and RDF. If so, the way we should do that is an @id on the datatype …
gkellogg: but we have a field, it is the baes
base
jenit: that is different
… base tells you what kind of datatype, eg. Integer -> age
… min/max etc
jenit: having @id on datatype and using that within RDF (and or XML) conversion could be quite a nice thing
gkellogg: if the datatype had an ID, would be used in output otherwise use base
jtandy: and in json, we'd ignore it / use base
gkellogg: seems a reasonable extension point
jenit: shall i make a proposal? separate issue? ok...
(… will write that out after the call)
jenit: that would allow a named xsd datatype to be used, carried through, validated etc
<jtandy> [example: "datatype": {"base": "boolean", "format": "#|-", "@id": "http://example.org/mydatatype" }
<JeniT> https://github.com/w3c/csvw/issues/534
jenit: also: issue #534 from Gregg
gkellogg: seems to have been shot down
…. doesn't change any processing
… moves responsibility for specifics of metadata into metadata doc rather than datamodel
… but late in the game.
… maybe we need a status 'deferring for v2' label [in github]?
(ivan … would need to do this)
(label vs milestone)
<JeniT> ivan, please could you add a “vNext” label to Github so that we can label issues that we want to defer with it?
(schema.org has 'dim and distant future, possible never' milestone: https://github.com/schemaorg/schemaorg/issues?q=is%3Aopen+is%3Aissue+milestone%3A%22The+distant%2C+hazy+future+%28or+possibly+never%29%22 )
<gkellogg> And then set #534 to vNext
jenit: other open issues not flagged for discussion:
<ivan> I have created a "Deferred to Version 2" label to github
… waiting for I18n. Unions discussed. Changes - we know needs doing. Registration templ gone in. Data Cube 509:
https://github.com/w3c/csvw/issues/509
jtandy: done all needed doing. waited for a response, fine to close it.
+1 to close it
… could easily mark it as resolved. he's been emailed.
case insensitivity - close.
xsd - discussed
allowing metadata in csv files - …
gkellogg: we do talk about it, but don't say how
… just that it is a processor specific thing
[...]
gkellogg; we have made the ability to express arbitrary rdf from a csv rather general purpose
eg. you could have a csv that could be turned to json-ld to be a csvw metadata file
jtandy: net result from whatever arrangement needs to be an annotated table
see #534
gkellogg: someone could do this and be compatible
danbri: json-ld inside html via <script>?
gkellogg: "CSV" is fuzzy as needn't be actual CSV
tables … smiilar, considering html tables
jenit: afaik there is nothing in spec that stops you from saying an HTML file is a tabular data file, from which you can extract (via some uknown process) some json metadata AND an annotated table model
gkellogg: if metadata is script tag w/ json-ld vfersion of our metadata
and table uri can be resolved to a table fragment
moving on,
#528:
document requirement for conneg for retrieving the context document
gkellogg: why to do this?
… an actual processor needn't have to do this
json-ld tool chains if used will handle this
not sure why we'll need to do something here
jenti: i think the issue just articulaltes reader's surprise.. .add a note?
gkellogg: ok, will do
jenit marks as editor action
access control allow origin — ivan is doing this
or done
… specifically for this context file
gkellogg: some q of whether it negotiates over anything but the json-ld
<jtandy> +1
jenit: can close?
+1
<scribe> closed.
jenit: descr of line terminators. -> editor action.
… on the ns doc
gkellogg: I don't think we can close that one above. re accss control.
csvw vs csvw.jsonld
ah
gkellogg: sorry, it is done. Close!
<JeniT> https://github.com/w3c/csvw/issues/532
jenit: ok, closed.
#532 - editor action on gregg, update property definitions, line terminator, ...
jtandy: i took a look, don't consider it garbled. Name and then a title for that element.
not confusing, … but may look weird
jenit: description says […]
#534 - discussed.
… anything else needing progress e.g. use case doc?
gkellogg: also tests. I stalled out on generating a lot of new tests.
… probably another 100 tests, easily, if not more.
… a q of when to say "enough!"?
jenit: gkellogg, can you add issues like "add tests around section n" in github? to break up the work?
gkellogg: will do
jenit: that would be useful. Also usecase doc. Nothing back from Davide around reworking the requirements. Jeremy: can you take that on?
jtandy: struggling with bandwidth currently. Not a giant job to cross-ref but I would struggle to make time.
danbri; what exactly needs doing?
<jtandy> http://w3c.github.io/csvw/use-cases-and-requirements/#req
jtandy: 3.1 accepted requirements is blank. 3.2 candidate requirements. Consider 1st one: well formed csv check, …
… with most of these we have 1 liners. Felt to me as though now that we understand better what that requirement is as we've written the code, … can we flesh out meaning
jtandy: also a bunch that never
made it out of candidate
... could try splitting it by the 4 sections
jenit: to split requirements out
… people then move them into either accepted, rejected, deferred
jtandy: lumpy distribution across sections
doc vs github?
jtandy: just do in the doc
jenit work in the doc directly.
jtandy: as a default position, all requirements listed as candidate should go into the Accepted list
if exception we should flag it in github for review.
[discussion of implementations]
looking at start of meeting before rrsagent, the following text was typed:
JeniT https://github.com/w3c/csvw/issues/524
jenit: 2 issues require teleconf discussion. irc://irc.w3.org:6667/#524 re xsd schema datatypes, … …which I have tried to get some clarity on. Needs more discussion on github.
jtandy: seems we have someone who finds it simple to use xml schema, and perhaps doesn't appreciate how difficult others find it. But I don't want to have to make an xsd parser in e.g. Javascript.
rragent, please draft minutes
This is scribe.perl Revision: 1.140 of Date: 2014-11-06 18:16:30 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/viua/via/ No ScribeNick specified. Guessing ScribeNick: danbri-user1 Inferring Scribes: danbri-user1 WARNING: No "Topic:" lines found. WARNING: No "Present: ... " found! Possibly Present: JeniT danbri gkellogg https ivan jenti jtandy You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 06 May 2015 Guessing minutes URL: http://www.w3.org/2015/05/06-csvw-minutes.html People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]