13:32:40 <RRSAgent> RRSAgent has joined #csvw
13:32:40 <RRSAgent> logging to http://www.w3.org/2015/07/08-csvw-irc
13:32:42 <trackbot> RRSAgent, make logs public
13:32:42 <Zakim> Zakim has joined #csvw
13:32:44 <trackbot> Zakim, this will be CSVW
13:32:44 <Zakim> I do not see a conference matching that name scheduled within the next hour, trackbot
13:32:45 <trackbot> Meeting: CSV on the Web Working Group Teleconference
13:32:45 <trackbot> Date: 08 July 2015
14:00:17 <JeniT_> JeniT_ has joined #csvw
14:03:08 <danbri1> (I'm in)
14:03:17 <danbri1> shall i scribe?
14:03:22 <danbri1> jeni/ivan re checklists
14:03:37 <danbri1> ivan: only thing I can't do, … if one of you could draft a more detailed blog that would be helpful
14:03:41 <danbri1> jeni: when?
14:03:51 <danbri1> ivan: after the publication, i.e. after a week from tomorrow
14:03:55 <danbri1> … would be good to have it then
14:04:54 <danbri1> jeni: quick review of issues…
14:05:07 <danbri1> … those labelled for LCCR
14:05:26 <danbri1> -changes for UC doc … but since a Note not REC not required
14:05:43 <danbri1> - test reqs; how important is it to have test suite finalized pre-cr?
14:05:48 <danbri1> ivan: good to have it when publishing goes through
14:05:54 <danbri1> https://github.com/w3c/csvw/labels/For%20LCCR
14:06:03 <danbri1> … i.e. between now and next week wouldn't be a problem
14:06:17 <danbri1> but when the test formally begins after the publication, it would be good if they're done
14:06:33 <danbri1> gkellogg: regarding #545 that's where i've kept a task list for testing
14:06:40 <danbri1> … handled except for one minor point
14:06:47 <danbri1> also points that arose in last ~48h or so
14:07:00 <danbri1> ivan: close as much as you can, gregg; if you think #545 is done, please close it
14:07:14 <danbri1> … on that, you never told me what to put into the wellknown file …
14:07:18 <danbri1> gkellogg: … but you did it
14:07:23 <danbri1> ivan: [checking]
14:07:31 <danbri1> gkellogg: we have tests for this and they pass :)
14:07:35 <JeniT> {+url}-metadata.json
14:07:35 <JeniT> csv-metadata.json
14:07:36 <JeniT> {+url}.json
14:07:37 <JeniT> csvm.json
14:07:41 <ivan> {+url}-metadata.json
14:07:41 <ivan> csv-metadata.json
14:07:41 <ivan> {+url}.json
14:07:41 <ivan> csvm.json
14:07:42 <danbri1> ivan: what I have now, … —^^
14:07:53 <danbri1> gkellogg: that is correct
14:07:58 <danbri1> ivan: then we can close #545
14:08:07 <JeniT> https://github.com/w3c/csvw/issues/617
14:08:10 <danbri1> jenit: last one then is #617 … number formats.
14:08:27 <danbri1> Ivan: we essentially took what Richard put there. I asked him yesterday if he could close it, but he hasn't yet.
14:08:36 <danbri1> ivan: we have included exactly what we said he should conclude
14:08:44 <danbri1> jenit: I believe he said it was ok in the pull req
14:09:08 <danbri1> jenit: there are changes to specs and to tests in pull reqs
14:09:21 <danbri1> ivan: ah if he agreed with pull req that's great
14:10:05 <JeniT> https://github.com/w3c/csvw/pull/651
14:10:46 <danbri1> what is his github id?
14:10:53 <danbri1> (e.g. commenter:defunkt can be used to search)
14:11:37 <JeniT> PROPOSE: We close #617 re number format changes (subject to test changes) given we haven’t heard otherwise from @r12a
14:11:43 <ivan> +1
14:11:43 <gkellogg> +1
14:11:45 <JeniT> +1
14:11:50 <danbri1> +1
14:11:59 <JeniT> RESOLVED: We will close #617 re number format changes (subject to test changes) given we haven’t heard otherwise from @r12a
14:12:00 <ivan> rrsagent, pointer?
14:12:00 <RRSAgent> See http://www.w3.org/2015/07/08-csvw-irc#T14-12-00
14:12:29 <danbri1> btw comment search: https://github.com/w3c/csvw/issues?utf8=%E2%9C%93&q=commenter%3Ar12a
14:13:24 <danbri1> closed …
14:13:34 <danbri1> ivan: that was important - the last open technical issue that we had (is now closed).
14:13:45 <danbri1> jenit: aside from that Ivan, is there anything we should formally resolve or do now?
14:14:04 <ivan> "The goal is to have at least two, mutually independent implementation for each test case in the official test suite"
14:14:07 <danbri1> ivan: we never formally described what the success criteria for the CR would be. What I put into the CR call is a standard text.
14:14:14 <danbri1> ^^--
14:14:32 <danbri1> … this is the minimum we should do. It would be tough to have 2 impls that formally impl every test.
14:14:44 <danbri1> jenit: that seems fair
14:14:56 <danbri1> …the only other one that I was thinking was whether we should have a requirement around publication
14:15:08 <danbri1> … should we require that 3 separate sites have published CSV with metadata
14:15:41 <danbri1> ivan: could be shooting ourselves in foot?
14:15:48 <danbri1> dan: do they need to be WG independent?
14:16:11 <danbri1> ivan: only indep from each other
14:17:36 <danbri1> dan: I lean towards omitting. we'll talk to publishers naturally as we're close to those kinds of people
14:17:51 <danbri1> … discussion of met office, GDS registries, …
14:19:00 <danbri1> gkellogg: would be nice to state as a nice-to-have
14:19:11 <danbri1> jeni: am happy w/ formal req being around impl
14:19:14 <danbri1> gkellogg: fine
14:19:18 <danbri1> danbri: fine [no audio]
14:19:40 <JeniT> PROPOSED: We will have "The goal is to have at least two, mutually independent implementation for each test case in the official test suite" as our exit criteria from CR
14:19:43 <JeniT> +1
14:19:53 <ivan> +1
14:19:57 <danbri1> +1
14:19:57 <gkellogg> +1
14:20:00 <JeniT> RESOLVED: We will have "The goal is to have at least two, mutually independent implementation for each test case in the official test suite" as our exit criteria from CR
14:20:09 <danbri1> jenit: anything else?
14:20:27 <danbri1> ivan: we now need a formal resolution to move to CR. I'll list here in IRC  those who answered by email.
14:20:31 <danbri1> jenit: can you phrase it?
14:21:32 <ivan> PROPOSED: The WG officially requests a Candidate Recommendation Transition to the W3C Director
14:21:38 <JeniT> +1
14:21:42 <gkellogg> +1
14:22:01 <ivan> Note, votes were gathered by email https://lists.w3.org/Archives/Public/public-csv-wg/2015Jul/0037.html
14:22:02 <danbri1> +1
14:22:38 <ivan> Positive answers by email: Gregg, Yakov, Jeni, DanBri, Ivan Stasinos, Davide, Jürgen, and Jeremy
14:22:59 <danbri1> invalid comma-separation
14:23:04 <ivan> +1
14:23:16 <ivan> RESOLVED: The WG officially requests a Candidate Recommendation Transition to the W3C Director
14:23:24 <danbri1> :))
14:25:09 <danbri1> jenit: 5pm UK time for call?
14:25:13 <danbri1> ivan: yes
14:25:21 <danbri1> … you have webex etc? yup
14:25:35 <danbri1> ivan: we have everything prep'd.
14:26:29 <danbri1> ivan: see admin on repo
14:26:33 <ivan> https://github.com/w3c/csvw/tree/gh-pages/admin
14:29:31 <JeniT> https://github.com/w3c/csvw/issues/664
14:29:33 <danbri1> —
14:29:48 <danbri1> gkellogg: plan to gather non-normative tests in one manifest
14:29:54 <danbri1> … mark it as optional / non-norm
14:30:14 <danbri1> ivan: for my understanding (this may come up for tomorrow(?fri) ) … what do we mean non-normative?
14:30:35 <danbri1> gkellogg: the test aspects for our specs that is not normative, e.g. testing of a CSV processor isn't normative in the specs
14:30:44 <danbri1> jenit: mostly around interpretation of the dialect and the parsing of CSV
14:31:02 <danbri1> … to skip header rows, do right thing w/ commented rows, … all of which is spec'd in non-normative sections of the spec.
14:31:17 <danbri1> ivan: In the CR transition req I said there are over 600+ tests available
14:31:24 <danbri1> but from a CR pov the non-norm tests are not req'd
14:31:37 <danbri1> …therefore the exit criteria is also such that we don't req those to be handled
14:31:46 <danbri1> we need a very clear separation of what is normative vs not
14:31:56 <danbri1> so would be good if you can give me an approx notion of the number of normative tests
14:32:06 <danbri1> gkellogg: easiest way is to do the updates and see what doc says
14:32:09 <danbri1> ivan: needed v soon
14:32:19 <danbri1> jenit: i have 15 tests identified that are non-normative
14:32:30 <danbri1> … 16
14:33:05 <danbri1> ivan: if in the req I say there are "over 600 tests" that statement is correct?
14:33:33 <danbri1> gkellogg: 205 each for rdf and json, 255 for validation
14:33:49 <danbri1> gkellogg: jeni also identified duplicative tests
14:33:53 <danbri1> eg. no.28
14:33:59 <danbri1> jenit: it is ok for them to be in there
14:34:03 <danbri1> not problematic as such
14:35:05 <danbri1> ivan: if you can separate norm from non-norm in the description that would be great, ideally before cr call
14:35:33 <danbri1> gkellogg: jeni, in your survey did you notice any areas lacking test coverage?
14:35:40 <danbri1> jenit: no
14:35:59 <danbri1> … a little worried around foreign keys, … having just focussed on the json ones, but it was ok when i looked at the validation
14:36:08 <danbri1> gkellogg:I tried to get every 'must' from the specs
14:36:11 <danbri1> jenit: amazing job!
14:36:15 <danbri1> +1
14:37:07 <danbri1> ivan: compares well with eg. rdfa (which was also well tested, and more complex)
14:37:19 <danbri1> jenit: anything more from those issues to discuss today?
14:37:42 <danbri1> gkellogg: didn't look into #667 enough… but can't say without diving into them
14:37:56 <danbri1> … if tests are inconsistent w/ what's in the specs and my impl is off, i'll just correct that
14:38:09 <danbri1> sometimes it isn't clear whether something should be a negative valuation vs a warning
14:38:18 <danbri1> … when should everything stop?
14:38:24 <ivan> rrsagent, draft minutes
14:38:24 <RRSAgent> I have made the request to generate http://www.w3.org/2015/07/08-csvw-minutes.html ivan
14:38:28 <danbri1> what counts for being invalid metadata is something as a judgement call
14:39:05 <danbri1> jenit: if i have any fear it would be around those kind of changes, as we can't do them during CR without needing another CR
14:39:21 <danbri1> gkellogg: yes, so solution is to make sure tests are consistent with whatever the specs say
14:39:35 <danbri1> … don't anticipate any wording change needed
14:39:54 <danbri1> jenit: if you're planning to get thru those today, flag as issues anything you find
14:40:07 <danbri1> … let's make sure we have in the tests an agreed take on what the spec is saying
14:40:29 <danbri1> gkellogg: there were cases in the format tests, where you flagged things as errors, maybe should've  been warnings
14:41:07 <danbri1> ...
14:41:35 <danbri1> eg see test 153
14:41:52 <danbri1> … a bad format string, a warning but not invalid
14:41:58 <danbri1> i took it to be an ommission
14:42:18 <danbri1> only other change was some fields where we didn't match on the format, where it would still be interpreted as an int not a string
14:42:26 <danbri1> i made it an int in the turtle too
14:42:34 <danbri1> jenit: I didn't go thru all the turtle outputs
14:42:48 <danbri1> anything else?
14:44:14 <ivan> rrsagent, draft minutes
14:44:14 <RRSAgent> I have made the request to generate http://www.w3.org/2015/07/08-csvw-minutes.html ivan
14:44:29 <ivan> trackbot, stop telcon
14:44:29 <trackbot> Sorry, ivan, I don't understand 'trackbot, stop telcon'. Please refer to <http://www.w3.org/2005/06/tracker/irc> for help.
14:44:35 <ivan> trackbot, end telcon
14:44:35 <trackbot> Zakim, list attendees
14:44:35 <Zakim> sorry, trackbot, I don't know what conference this is
14:44:43 <trackbot> RRSAgent, please draft minutes
14:44:44 <RRSAgent> I have made the request to generate http://www.w3.org/2015/07/08-csvw-minutes.html trackbot