13:58:15 <RRSAgent> RRSAgent has joined #i18n
13:58:15 <RRSAgent> logging to http://www.w3.org/2015/03/26-i18n-irc
13:58:20 <Zakim> Zakim has joined #i18n
13:58:26 <aphillip> trackbot, prepare teleconference
13:58:28 <trackbot> RRSAgent, make logs world
13:58:30 <trackbot> Zakim, this will be 4186
13:58:30 <Zakim> ok, trackbot; I see I18N_WG()10:00AM scheduled to start in 2 minutes
13:58:31 <trackbot> Meeting: Internationalization Working Group Teleconference
13:58:31 <trackbot> Date: 26 March 2015
13:58:38 <aphillip> Agenda: https://lists.w3.org/Archives/Member/member-i18n-core/2015Mar/0018.html
13:58:45 <aphillip> Chair: Addison Phillips
13:58:47 <matial> matial has joined #i18n
13:58:50 <aphillip> ScribeNick: aphillip
13:59:04 <RRSAgent> I have made the request to generate http://www.w3.org/2015/03/26-i18n-minutes.html aphillip
14:00:05 <aphillip> zakim, who is here?
14:00:05 <Zakim> I18N_WG()10:00AM has not yet started, aphillip
14:00:07 <Zakim> On IRC I see matial, Zakim, RRSAgent, aphillip, Steve_Atkin, r12a, koji, fantasai, xiaoqian, trackbot
14:01:05 <aphillip> zakim, this is I18N
14:01:05 <Zakim> ok, aphillip; that matches I18N_WG()10:00AM
14:01:06 <r12a> zakim, dial richard please
14:01:06 <Zakim> ok, r12a; the call is being made
14:01:07 <Zakim> +Richard
14:01:12 <aphillip> zakim, who is here?
14:01:12 <Zakim> On the phone I see Steve_Atkin, aphillip, Richard (muted)
14:01:14 <Zakim> On IRC I see matial, Zakim, RRSAgent, aphillip, Steve_Atkin, r12a, koji, fantasai, xiaoqian, trackbot
14:03:44 <aphillip> zakim, who is here?
14:03:44 <Zakim> On the phone I see Steve_Atkin, aphillip, Richard
14:03:46 <Zakim> On IRC I see matial, Zakim, RRSAgent, aphillip, Steve_Atkin, r12a, koji, fantasai, xiaoqian, trackbot
14:04:09 <aphillip> Topic: Agenda
14:04:30 <aphillip> Topic: Info Share
14:04:39 <aphillip> Topic: Action Items
14:05:33 <aphillip> action-113?
14:05:33 <trackbot> action-113 -- Richard Ishida to Work on tidying up Additional Requirements for Bidi in HTML & CSS over coming couple of months -- due 2015-02-05 -- OPEN
14:05:33 <trackbot> http://www.w3.org/International/track/actions/113
14:05:55 <aphillip> action: addison: ping Aharon with email about additional requirements document
14:05:55 <trackbot> Created ACTION-421 - Ping aharon with email about additional requirements document [on Addison Phillips - due 2015-04-02].
14:08:37 <aphillip> action-412?
14:08:37 <trackbot> action-412 -- Felix Sasaki to Publish multilingual web usage scenarios -- due 2015-03-12 -- OPEN
14:08:37 <trackbot> http://www.w3.org/International/track/actions/412
14:09:32 <aphillip> close action-418
14:09:32 <trackbot> Closed action-418.
14:10:10 <aphillip> close action-420
14:10:10 <trackbot> Closed action-420.
14:10:16 <aphillip> Topic: Info Share
14:10:53 <aphillip> Topic: RADAR and Active Work Review
14:11:04 <aphillip>     https://www.w3.org/International/wiki/Review_radar
14:11:16 <r12a-limechat> r12a-limechat has joined #i18n
14:12:37 <aphillip> steve: will pick up svg a11y spec?
14:13:07 <aphillip> data on the web bp: make felix shepherd
14:13:34 <aphillip> richard: new tweet about new version of latin layout req
14:13:53 <aphillip> https://www.w3.org/International/wiki/Project_radar
14:15:08 <JcK> JcK has joined #i18n
14:15:27 <aphillip> action: addison: warn WG members to look at LTLI ED for discussion next week as a new WD
14:15:27 <trackbot> Created ACTION-422 - Warn wg members to look at ltli ed for discussion next week as a new wd [on Addison Phillips - due 2015-04-02].
14:16:28 <aphillip> richard: cl req: not much activity this week, except that discussing with PLH what process is for getting people into TF
14:16:31 <Zakim> +JcK
14:16:41 <aphillip> ... don't need to make them WG members per-se
14:16:53 <aphillip> ... process is to make them public invited expert
14:17:07 <aphillip> ... put them in the IG
14:17:38 <aphillip> richard: they get a W3 account and then fill in IE form
14:18:48 <aphillip> richard: predefined counter styles
14:18:54 <aphillip> ... if we have extra time, could brainstorm
14:19:43 <aphillip> mlweb workshop
14:19:51 <r12a-limechat> http://www.multilingualweb.eu/documents/2015-riga-workshop/2015-riga-program
14:20:03 <r12a-limechat> program announced
14:21:49 <aphillip> Topic: TPAC attendance
14:22:02 <aphillip> addison: several people seem to be coming
14:22:06 <aphillip> ... do we need a room?
14:22:31 <aphillip> richard: tpac in china wasn't bad... had seats to sit on
14:23:04 <aphillip> addison: no room, wandering minstrel mode?
14:23:14 <aphillip> richard: think that might be best approach
14:23:31 <aphillip> addison: what about cl req folks?
14:23:58 <aphillip> action: richard: ping beihan folks about tpac logistics
14:23:58 <trackbot> Created ACTION-423 - Ping beihan folks about tpac logistics [on Richard Ishida - due 2015-04-02].
14:24:23 <aphillip>  Topic: Best practices with the W3C process
14:24:31 <aphillip>    https://www.w3.org/wiki/DocumentReview
14:24:34 <Zakim> +koji
14:26:10 <aphillip>          BP: if in doubt, before formally asking a horizontal group to review a document, talk to them informally (f.ex. via e-mail).  (comment): I agree that it may not be obvious that a document needs horizontal review, but I tend to take a stronger line: even if you have not identified a horizontal group dependency, you should always discuss whether a review is necessary or even request a review (saying if you think there are no issues of interest to t[CUT]
14:27:03 <aphillip>          BP: avoid cross-posting to multiple lists (and if this must be done, use Bcc:).      Alternate instructions for providing commentary (if needed). Some groups use bugzilla or github as a means of communication and for tracking issues. --Addison Phillips (talk)
14:28:32 <aphillip> richard: should add cross-posting as needed
14:28:37 <aphillip> ... "get a decent mail client"
14:29:23 <aphillip> addison: will edit the above
14:29:54 <aphillip> ... actively want cross-posting
14:30:21 <aphillip> action: addison: solicit comments from WG on w3c document review process wiki page
14:30:21 <trackbot> Created ACTION-424 - Solicit comments from wg on w3c document review process wiki page [on Addison Phillips - due 2015-04-02].
14:30:39 <aphillip> Topic: Best Practices when writing Specifications
14:34:23 <aphillip> richard: don't use bpXXX number. use the gray character off to the left (the id for the best practice) when referring to bps
14:34:34 <RRSAgent> I have made the request to generate http://www.w3.org/2015/03/26-i18n-minutes.html aphillip
14:34:42 <r12a-limechat> use http://w3c.github.io/bp-i18n-specdev/#lang_neg or #lang_neg NOT BP71 to refer to recommendations
14:38:05 <matial> matial has joined #I18N
14:39:07 <matial> matial has joined #i18n
14:39:23 <aphillip> action: richard: figure out means of making BP number and/or identifier properly emphasized
14:39:24 <trackbot> Created ACTION-425 - Figure out means of making bp number and/or identifier properly emphasized [on Richard Ishida - due 2015-04-02].
14:40:07 <aphillip> richard: this document
14:40:33 <aphillip> richard: orginally this document was a home for homeless information
14:40:47 <aphillip> ... but became apparent that people expected this to be everything they needed
14:40:51 <aphillip> ... or at least pointers
14:40:59 <aphillip> ... and confusing if they had to go to techniques page
14:41:07 <aphillip> ... so spent time rearchitecting it
14:41:22 <aphillip> ... all BPs in techniques page are now here too
14:41:39 <aphillip> ... and the techniques page would like to "more info" locations
14:41:54 <aphillip> ... changed techniques to allow pointer to more than one thing
14:42:15 <aphillip> ... and wrote a script that takes BP document and generates techniques
14:42:54 <aphillip> richard: two versions of techniques (one collapsable, one not)
14:43:09 <aphillip> ... one version is generated from github document
14:43:20 <aphillip> ... reduces need for duplicating information
14:43:29 <RRSAgent> I have made the request to generate http://www.w3.org/2015/03/26-i18n-minutes.html aphillip
14:43:55 <Zakim> -Steve_Atkin
14:44:20 <Steve_Atkin> My line dropped calling back
14:45:01 <Zakim> +Steve_Atkin
14:46:59 <r12a-limechat> http://w3c.github.io/bp-i18n-specdev/
14:47:19 <r12a-limechat> http://www.w3.org/International/techniques/developing-specs
14:47:37 <r12a-limechat> http://www.w3.org/International/techniques/developing-specs-dynamic
14:47:57 <RRSAgent> I have made the request to generate http://www.w3.org/2015/03/26-i18n-minutes.html aphillip
14:49:22 <aphillip> Topic: WebVTT and App-Manifest comment review
14:51:17 <aphillip> https://lists.w3.org/Archives/Public/www-international/2015JanMar/
14:52:04 <aphillip> https://lists.w3.org/Archives/Public/www-international/2015JanMar/0181.html
14:52:12 <aphillip> https://www.w3.org/Bugs/Public/show_bug.cgi?id=28262
14:53:30 <aphillip> https://lists.w3.org/Archives/Public/www-international/2015JanMar/0167.html
14:53:59 <aphillip> https://lists.w3.org/Archives/Public/www-international/2015JanMar/0168.html
14:54:14 <aphillip> https://www.w3.org/Bugs/Public/show_bug.cgi?id=28266
14:58:12 <aphillip> https://www.w3.org/Bugs/Public/show_bug.cgi?id=28265
14:58:12 <Steve_Atkin> I need to drop I have another call.
14:58:22 <Zakim> -Steve_Atkin
14:58:35 <aphillip> W3C HTML5 also has rb, rtc  That said, I think it is a non-goal to match the expressiveness of HTML in WebVTT. WebVTT only has elements to cater for specific use cases that apply to it.
15:00:00 <aphillip> koji: don't know where the boundaries are?
15:00:16 <aphillip> ... quite hard to say that
15:00:40 <aphillip> ... nobody from japan is helping with this?
15:00:50 <aphillip> addison: don't see anyone per-se
15:03:30 <aphillip> koji: common when only two lines to have first line ruby on top and second line ruby on bottom, for example
15:03:40 <aphillip> ... not sure if TV wants double-sided
15:03:57 <aphillip> ... maybe not so important to have group, etc
15:04:57 <aphillip> richard: no ruby styling provided? maybe new comment?
15:05:52 <aphillip> action: richard: ask at domain meeting about whether webvtt will be used for tv and cinema
15:05:53 <trackbot> Created ACTION-426 - Ask at domain meeting about whether webvtt will be used for tv and cinema [on Richard Ishida - due 2015-04-02].
15:09:44 <aphillip> action: addison: tag all threads based on our comments in appmanifest's github so we can track
15:09:44 <trackbot> Created ACTION-427 - Tag all threads based on our comments in appmanifest's github so we can track [on Addison Phillips - due 2015-04-02].
15:09:56 <aphillip> Topic: AOB?
15:10:51 <aphillip> zakim, who is here?
15:10:51 <Zakim> On the phone I see aphillip, Richard, JcK, koji
15:10:53 <Zakim> On IRC I see matial, JcK, r12a-limechat, Zakim, RRSAgent, aphillip, koji, fantasai, xiaoqian, trackbot
15:11:11 <Zakim> -aphillip
15:11:15 <Zakim> -JcK
15:11:19 <aphillip> Regrets: Felix
15:11:25 <JcK> JcK has left #i18n
15:11:30 <aphillip> Present: Addison, Richard, Steve, JcK, Koji
15:11:36 <aphillip> rrsagent, make minutes
15:11:36 <RRSAgent> I have made the request to generate http://www.w3.org/2015/03/26-i18n-minutes.html aphillip
15:15:04 <r12a-limechat> http://www.w3.org/International/tests/repository/css3-writing-modes/bidi/results-bidi
15:15:47 <koji> http://test.csswg.org/harness/results/css-writing-modes-3_dev/grouped/
15:18:43 <r12a-limechat> http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/bidi-plaintext-002/format/html5/
15:19:06 <r12a-limechat> http://www.w3.org/International/tests/repository/run?manifest=css3-writing-modes/bidi&test=block-plaintext-002
15:25:08 <koji> lost voice?
15:25:38 <r12a-limechat> yes, apparently
15:25:51 <koji> I'm seeing here https://github.com/w3c/csswg-test/pulls?page=1&pulls_only=true&q=is%3Apr+is%3Aopen
15:25:52 <r12a-limechat> basically, i was saying i'll investigate and send you an email
15:26:04 <koji> do not see much bi-di tests PR are open
15:26:46 <r12a-limechat> it may be that they were merged into github, but not copied over to shepherd's mercurial db ??
15:27:08 <r12a-limechat> which was the case for the counter-styles tests i was discussing recenlty with ChrisL
15:27:31 <koji> oh really, I thought that should be automatic
15:27:40 <koji> maybe system problem?
15:27:52 <r12a-limechat> give me a short while to investigate off-line and double-check the files, and i'll send you what i find, ok?
15:28:01 <koji> that'd be great
15:28:03 <r12a-limechat> it's not automatic, no :(
15:28:16 <koji> oh...
15:28:21 <r12a-limechat> ok, speak to you later
15:28:29 <koji> thank you!
15:34:11 <Zakim> -koji
15:39:12 <Zakim> disconnecting the lone participant, Richard, in I18N_WG()10:00AM
15:39:13 <Zakim> I18N_WG()10:00AM has ended
15:39:13 <Zakim> Attendees were Steve_Atkin, aphillip, Richard, JcK, koji
16:46:44 <r12a-limechat> koji, still there?
16:47:11 <koji> yeah
16:47:14 <r12a> wow
16:47:22 <r12a> so i've been investigating...
16:47:31 <r12a> turns out i was comparing the wrong files
16:48:25 <r12a> if i look at bidi-plaintext-001, this test fails on my current version of firefox
16:48:37 <r12a> looking at http://test.csswg.org/harness/results/css-writing-modes-3_dev/grouped/
16:48:48 <r12a> it shows
16:49:06 <r12a> one pass and 6 fails for gecko
16:49:28 <r12a> the pass and one of the fails is for the same version of firefox
16:49:30 <r12a> :(
16:49:57 <r12a> this is one of the reasons i created my own test framework, rather than continue with the w3c one
16:50:04 <koji> ah
16:50:14 <r12a> i was finding that people were adding spurious results quite often
16:50:22 <koji> yeah, it's not surprising to see wrong results submitted
16:50:24 <r12a> also, there's no column for blink
16:50:41 <koji> I actually did one by mistake, but Shepherd did not give me a chance to edit/delete records
16:50:54 <koji> yeah :-(
16:51:03 <r12a> yeah, there is a method of deleting incorrect stuff, but its a little painful
16:51:08 <r12a> and only admins can do it
16:51:38 <koji> oh I see
16:51:50 <r12a> btw, the bidi-plaintext-002 file no longer appears in my results at http://www.w3.org/International/tests/repository/css3-writing-modes/bidi/results-bidi#plaintext
16:52:26 <r12a> i've been trying to figure out why, but i think its because the test produces a positive result without the unicode-bidi property, ie. not a good test
16:52:38 <r12a> i'm not sure how to remove it from the shepherd environment
16:52:39 <koji> I see
16:53:12 <r12a> i was thinking that i could run all these tests again on the latest versions of the browsers and update as necessary
16:53:19 <r12a> in my results page
16:53:33 <r12a> though i'm not sure there'll be many, if any changes
16:53:41 <koji> ok, so for bi-di, I should trust your results than w3c ;)
16:53:49 <koji> yeah
16:53:50 <r12a> :)
16:53:59 <r12a> or run them yourself ;-)
16:54:09 <koji> so, from your page
16:54:17 <koji> it looks like browsers are quite consistent for pass/fail
16:54:33 <r12a> i could also create a set of tests for prefixed versions of the properties, since this would show a lot more green
16:54:53 <r12a> in fact, i think it would be nearly all green
16:55:12 <r12a> should i do that?
16:55:24 <koji> are they prefixed?
16:55:34 <koji> I know writing-mode is
16:55:39 <koji> but is unicode-bidi?
16:55:55 <r12a> yes, that's why our css fix works for html
16:56:07 <r12a> for isolation
16:56:32 <koji> that I didn't know...the properties are from CSS2, right?
16:56:37 <koji> so values are prefixed?
16:56:46 <r12a> see the shim at http://www.w3.org/International/articles/inline-bidi-markup/Overview#html5markup
16:57:01 <r12a> yes, values
16:58:32 <koji> I see...
16:58:45 <koji> I have a bookmarklet to test writing-mode on Chrome https://github.com/kojiishi/tweak-css-tests
17:01:19 <koji> I can add them and try to support other browsers if it helps
17:01:27 <koji> or you can send me PR ;-)
17:01:47 <r12a> this is for the prefixed tests?
17:02:27 <koji> It modifies DOM to add prefixes and old syntaxes
17:02:33 <r12a> i see
17:02:58 <r12a> actually i'll probably just copy all the files and globally replace the necessary parts
17:03:06 <koji> or you can include <script> to do so
17:03:23 <koji> ok
17:03:34 <r12a> however, i'm not sure whether prefix tests are welcome at the ttwf repos ?
17:04:06 <koji> yeah, that's what troubled me
17:04:20 <r12a> but i can do it in my repo ;-)
17:04:38 <koji> and that's why I wrote the bookmarklet
17:06:02 <koji> back to priorities,
17:06:38 <koji> so all new things -- isolate, isolate-override, and plaintext are not supported at all
17:06:57 <r12a> unless you use prefixes
17:07:03 <r12a> i think
17:07:14 <r12a> though i'm not sure about plaintext
17:07:27 <r12a> i'd need to check
17:07:37 <koji> oh, so your results do not include "supported but prefixed"?
17:07:42 <r12a> correct
17:07:53 <r12a> only standard usage
17:08:00 <koji> ah ha
17:08:25 <r12a> but if i make a separate set of prefixed tests, i can include results for those - but separately
17:08:37 <koji> that'd be helpful
17:08:46 <r12a> for example, see the hyphens tests at http://www.w3.org/International/tests/repo/results/hyphens
17:08:53 <koji> to understand which features were supported how much
17:08:57 <r12a> first section of results = standard
17:09:03 <r12a> next section = prefixed
17:09:09 <koji> and to measure if we're stable enough to consider unprefix
17:09:10 <r12a> so it's easy to see what's what
17:09:19 <r12a> yes
17:09:54 <koji> can you do that?
17:09:56 <r12a> ok, well i'll try to do that
17:10:02 <r12a> it may take a few days
17:10:08 <koji> superb
17:10:34 <koji> are all these 3 values equally important?
17:11:32 <koji> do you have any sense of which one should be supported earlier?
17:13:12 <r12a> isolation is definitely important
17:13:36 <r12a> isolate-override perhaps less so
17:13:52 <r12a> but if you are doing one, it may be easy to do both together
17:14:05 <r12a> plaintext would also be especially useful
17:15:05 <r12a> note that unicode now recommends isolating control characters eg. RLI rather than the old embedding ones eg. RLE as the default (although unfortunately those don't yet work in most browsers)
17:15:23 <r12a> and we recommend isolation too to content authors
17:15:32 <r12a> also, dir in html is supposed to isolate
17:15:40 <r12a> in HTML5
17:16:08 <koji> as far as I browsed the related code a bit ago
17:16:16 <koji> we have some code for isolate
17:16:20 <r12a> though there's a little lag there too wrt adoption - but that may be because css isn't there yet - the implemnetation of dir isolation in html depends on the deafult stylesheet
17:16:29 <koji> little-to-no for isolate-override and plaintext
17:16:40 <r12a> yes, that's what i expected
17:18:37 <Zakim> Zakim has left #i18n
17:19:12 <koji> looks like we set "embed" for dir attribute https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit/Source/core/html/HTMLElement.cpp&q=htmlelement.cpp&sq=package:chromium&l=222
17:20:07 <koji> and "isolate" if dir="auto" https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit/Source/core/html/HTMLElement.cpp&sq=package:chromium&rcl=1427370675&l=119
17:21:30 <r12a> there's a bug for that somewhere...
17:22:37 <r12a> https://code.google.com/p/chromium/issues/detail?id=391260
17:24:07 <koji> ah, I know these guys...
17:24:55 <r12a> i have some tests for that too...
17:25:33 <r12a> http://www.w3.org/International/tests/repository/html5/the-dir-attribute/results-dir-isolation
17:26:01 <r12a> (note the results below for the CSS Shim)
17:26:18 <r12a> much better coverage - just a few corner case issues
17:26:45 <r12a> except, of course, for IE :(
17:32:37 <koji> so isolate and dir are working at some level?
17:32:45 <r12a> yes
17:34:26 <koji> depends on the test results of prefixed tests, maybe plaintext could be the higher priority?
17:35:29 <koji> but then what does the bug you mentioned mean?
17:35:48 <koji> it says Chrome doesn't set isolate, but your tests pass?
17:36:27 <r12a> the bug asks that the default stylesheet for the dir attribute in html be set to use the css properties for isolation
17:36:52 <r12a> but, as we saw, the standard version of those property values is not yet supported
17:36:56 <r12a> only prefixed
17:40:58 <koji> understood, your tests explicitly sets properteis using prefixed values
17:41:20 <r12a> the ones for the shim, yes
17:44:46 <koji> looks to me that the html one might be the first one to attack
17:45:10 <koji> I'll check with aharon too, but
17:45:18 <koji> what do you think?
17:46:19 <r12a> if we can convince folks to implement the standard prop value, rather than prefixed, then the html dir should be more straightforward, no?
17:46:30 <r12a> i'm not sure, just guessing
17:47:57 <koji> as far as i understand, whether it's prefixed or not does not matter much to support dir-to-CSS mapping
17:48:23 <r12a> it would certainly be very good to fix the dir
17:48:29 <r12a> very useful
17:48:37 <koji> there are two parts of code, I don't understand well yet the relationship, but
17:48:50 <r12a> in fact, in terms of author use, that's probably the main thing
17:48:54 <r12a> for html
17:49:15 <koji> Both UA default stylesheet CSS and code mapping
17:49:19 <r12a> we try to encourage people to use dir everywhere, rather than css
17:49:32 <koji> The former needs to use prefix until we unprefix, but unprefixing it is a tiny work we can do anytime
17:49:43 <koji> I see
17:50:11 <koji> lol
17:50:57 <koji> Ok, I'll contact aharon and two guys in the bug to see if there were any progress on it, and if not, I'll try to look into it.
17:51:38 <r12a> thanks
17:52:00 <koji> thank you for your support as always!
17:52:08 <r12a> and yours ! :)
19:34:10 <r12a> r12a has joined #i18n
21:23:11 <r12a> r12a has joined #i18n
22:00:04 <r12a> r12a has joined #i18n
23:18:44 <r12a> r12a has joined #i18n