| 02:26 | Hixie | is half-way through 14000 lines' worth of text on microdata in html5, and has so far found 296 lines' worth of use cases, requirements, scenarios, and problem descriptions |
| 02:26 | <Hixie> | (of which i wrote about half myself) |
| 02:27 | <Hixie> | (and 71 of which are blank lines) |
| 07:57 | gsnedders | has interview any minute… |
| 07:59 | <MikeSmith> | NicolasRaoul: http://hsivonen.iki.fi/testing-html5-parsing/ |
| 08:02 | <MikeSmith> | hsivonen: is there any way I can tell Minefield to automatically look for updates from the try server instead of wherever it normally looks? |
| 08:04 | <hsivonen> | MikeSmith: I don't know. Does it look for updates at all? |
| 08:04 | <hsivonen> | AFAIK, tryserver builds don't look for updates on the tryserver. |
| 08:04 | <MikeSmith> | OK |
| 08:05 | <MikeSmith> | I think Minefield itself normally checks for updates |
| 08:06 | <hsivonen> | nightly builds do update to later nightly builds |
| 08:12 | <MikeSmith> | hsivonen: about the statement "Other HTML parsing operations still use the old parser unconditionally." |
| 08:12 | <MikeSmith> | what other parsing operations are there? |
| 08:12 | <hsivonen> | off the top of my head: |
| 08:13 | <hsivonen> | feed payload sanitizing |
| 08:13 | <hsivonen> | Netscape bookmarks import |
| 08:13 | <hsivonen> | HTML clipboard import |
| 08:13 | <hsivonen> | text/plain loading |
| 08:13 | <hsivonen> | view source syntax highlighting |
| 08:14 | <hsivonen> | I wouldn't be at all surprised if there were other cases lurking somewhere |
| 08:14 | <hsivonen> | DOMParser and XHR2 will introduce new cases |
| 08:15 | <MikeSmith> | OK, I see |
| 08:17 | <MikeSmith> | hsivonen: so in my about:config, html5.enable is not showing up as an option |
| 08:18 | <MikeSmith> | I just installed from hsivonen⊙ifd |
| 08:19 | <MikeSmith> | hsivonen: is there any test I can do to make sure it's actually using the HTML5 parse? |
| 08:19 | <MikeSmith> | parser |
| 08:19 | gsnedders | takes a deep breath |
| 08:19 | <gsnedders> | My hand is sore from holding the phone so tight during the interview… |
| 08:19 | <MikeSmith> | gsnedders: interview? |
| 08:20 | <gsnedders> | MikeSmith: Summer internship, Opera |
| 08:20 | <hsivonen> | MikeSmith: http://hsivonen.iki.fi/test/moz/html5-parsing.html |
| 08:20 | <hsivonen> | MikeSmith: if you see proper math and the video has a clip path |
| 08:22 | MikeSmith | checks |
| 08:22 | <MikeSmith> | wow, cool |
| 08:23 | <MikeSmith> | hsivonen: thanks |
| 08:23 | <MikeSmith> | btw, I notice that the blur also blurs the default video controls |
| 08:24 | <hsivonen> | I guess it's a feature :-) |
| 08:24 | <MikeSmith> | heh |
| 08:24 | <MikeSmith> | I wonder if it's also a feature that it doesn't let me actually use the controls |
| 08:25 | <MikeSmith> | if I mouse over the video, the controls pop up, but then I when I most down to try to click, they disappear again |
| 08:26 | <hsivonen> | hmm. that used to work some builds ago |
| 08:26 | <MikeSmith> | hsivonen: anyway, fwiw, html5.enable pref is definitely not showing up in the OSX environment I'm trying it on |
| 08:26 | <hsivonen> | I wonder if it is a bug or a clickjacking prevention feature |
| 08:27 | <hsivonen> | MikeSmith: you need to add the pref |
| 08:27 | <MikeSmith> | d'oh |
| 08:27 | <MikeSmith> | OK |
| 08:27 | <MikeSmith> | gsnedders: about my tweet, I don7t think thy are using Unicode at all, but whatever the normal mainland Chinese encoding is |
| 08:28 | <MikeSmith> | or some subset of it |
| 08:28 | <MikeSmith> | or some font that doesn't have all the glyphs |
| 08:28 | <MikeSmith> | or something |
| 08:28 | <hsivonen> | MikeSmith: PRC national standard maps to Unicode, IIRC |
| 08:28 | <hsivonen> | GB18030 IIRC |
| 08:28 | <MikeSmith> | OK |
| 08:29 | <hsivonen> | it would have been madness to make it not fully mappable |
| 08:29 | <hsivonen> | fortunately that was situation was avoided |
| 08:30 | <MikeSmith> | it may be that the characters in question are neither in Unicode nor in GB18030 |
| 08:31 | <MikeSmith> | that's probably it, actually |
| 08:31 | <MikeSmith> | I know that's the case for some name characters in use in Japan |
| 08:31 | <MikeSmith> | variant kanji |
| 08:31 | <MikeSmith> | so-called gaiji |
| 08:32 | <hsivonen> | "Like UTF-8, GB18030 is a superset of ASCII and can represent the whole range of Unicode code points; in addition, it is also a superset of GB2312." says Wikipedia |
| 08:32 | <MikeSmith> | OK |
| 08:34 | hsivonen | still hasn't found out the technical specs for personal names in Finland |
| 08:35 | <MikeSmith> | I've tried to find an estimated of how many not-in-Unicode gaiji for personal names there are in use in Japan, but nobody seems to know |
| 08:36 | <MikeSmith> | estimate of |
| 08:36 | <MikeSmith> | it's sort of a long tail, I guess |
| 08:39 | <hsivonen> | I think the tone of the NYT article is a bit hypocritical considering that the U.S. requires European to coerce their names to ASCII on DHS forms |
| 08:39 | <hsivonen> | *Europeans |
| 08:41 | <MikeSmith> | I didn't know the US required that |
| 08:41 | <MikeSmith> | can't say that I'm surprised, though |
| 08:45 | <MikeSmith> | hsivonen: the problem with that is, I seem to remember reading something a while back that said US was requiring people to also have boarding passes that exactly match the name as it it on their passports |
| 08:46 | <hsivonen> | fortunately, I have an ASCII-compatible name |
| 08:47 | <MikeSmith> | me too |
| 08:47 | <MikeSmith> | even including the parenthesis |
| 08:52 | <MikeSmith> | hsivonen: so are there any pages that currently have problem behavior if I use the HTML5-enabled build with them? |
| 08:52 | <hsivonen> | MikeSmith: I think there may be problems around meta refresh and HTML5 application manifests |
| 08:53 | <MikeSmith> | OK |
| 09:19 | <hsivonen> | sigh. aria-dropeffect now has even more possible tokens |
| 09:20 | <hsivonen> | one would think the tokens were mutually exclusive... |
| 09:34 | <jgraham> | hsivonen: AFAICT the only choice that makes sense to combine with anything else is execute since that seems to be a different way of saying "other" |
| 09:37 | <hsivonen> | it seems to me that the principle applied to ARIA is the exact reverse of what the PFWG applied to alt: |
| 09:37 | <hsivonen> | if there is a semiconceivable possibility of a combination maybe making sense, allow it |
| 09:45 | <Hixie> | hsivonen: it's different people, isn't it? |
| 09:46 | <hsivonen> | Hixie: I don't know. I haven't done Member-only mailing list archeology to find out who wanted what. |
| 09:46 | <Hixie> | ah ok |
| 09:48 | <hsivonen> | what's the point of having values "true", "false", "undefined" where "undefined" is equivalent to not specifying the attribute? |
| 09:53 | <annevk2> | that's what you get when you define things in terms of markup rather than states I suppose |
| 09:54 | <annevk2> | it would be nice if for consistency the empty string mapped to true as well |
| 09:56 | <hsivonen> | as far as I can tell, the empty string is equivalent to "undefined" |
| 10:00 | <jgraham> | <span aria-checked> == <span aria-checked="undefined">? Nice... |
| 10:05 | <annevk2> | aria-checked=false actually being false is somewhat inconsistent already |
| 10:07 | <hsivonen> | whoa! the new values for aria-invalid are, umm, interesting |
| 10:07 | <hsivonen> | "true", "false", "grammar", "spelling" |
| 10:07 | <zcorpan> | o_O |
| 10:08 | <MikeSmith> | beautiful |
| 10:08 | <MikeSmith> | I'm pretty sure it's some kind of surrealist put-on |
| 10:09 | <jgraham> | Submit it to the Daily WTF? |
| 10:11 | <hsivonen> | hmm. "rude" is no longer supported for aria-live :-( |
| 10:11 | <annevk2> | I was hoping I did not have to review the actual vocabulary |
| 10:13 | <annevk2> | "User agents may refuse to submit the form as long as there is an element for which aria-invalid is true." |
| 10:13 | <hsivonen> | the authoring requirements of aria-posinset and aria-level still depend on things that aren't well-defined |
| 10:13 | <annevk2> | Same for aria-labelledby |
| 10:13 | <annevk2> | I thought the idea was that there would be no user agent requirements other than exposing information correctly to assistive technology? |
| 10:14 | <hsivonen> | annevk2: are you going to send feedback for that form submission thing? |
| 10:14 | annevk2 | sighs |
| 10:15 | jgraham | adores the note for aria-labelledby that explains that they have chosen a word with known ambiguous spelling to "minimize the difficulty for developers" |
| 10:16 | <annevk2> | ouch |
| 10:18 | <annevk2> | hsivonen, the deadline for comments was last Friday I believe; I wonder if any new comments would be in vain |
| 10:20 | <hsivonen> | annevk2: depends on whether they want to make to REC fast or whether they want a proper spec. |
| 10:21 | <hsivonen> | annevk2: also, does Process require anything about a DL or just about things happening before the next transition? |
| 10:21 | <hsivonen> | it's still unclear to me, how ARIA can make it to REC with two interoperable implementations |
| 10:22 | <hsivonen> | is there another criterion for transitioning to REC? |
| 10:22 | <hsivonen> | or do desks count as implementations |
| 10:22 | <annevk2> | Looking at PF minutes I'm not sure they're addressing my comments in a sensible way. Mu |
| 10:23 | <annevk2> | The XHTML2 WG uses schema "implementations" as implementations so I'm pretty sure anything goes. |
| 10:23 | <annevk2> | Of course, XHTML2 is not something that makes it into browsers so that is not really a concern, but WAI-ARIA is somewhat. (Although it remains to be seen how good it will work.) |
| 10:24 | <jgraham> | This is why I have given up taking the time to review aria. The process seems set up to produce bad results |
| 10:24 | <hsivonen> | huh? does Google Groups intentionally require login to browse? |
| 10:26 | <zcorpan> | at some point, the aria spec referenced html5 for the definition of "whitespace" |
| 10:26 | <zcorpan> | then, the reference was removed because aria was going to rec before html5 was |
| 10:26 | <zcorpan> | so whitespace is now undefined |
| 10:26 | <MikeSmith> | I think if the ARIA spec includes testable assertions about UA requirements, the normal criteria for it transitioning beyond CR would need to be that there are at least two interoperable implementations that meet those requirements |
| 10:27 | <MikeSmith> | but I don't think the process doc has any specific language stating that |
| 10:27 | <hsivonen> | MikeSmith: I think any browser since circa IE4 or Netscape 6 is a conforming implmentation, then |
| 10:29 | MikeSmith | looks back and sees annevk2's quote "User agents may refuse to submit the form as long as there is an element for which aria-invalid is true." |
| 10:29 | <MikeSmith> | but I see also that's a "may" |
| 10:31 | jgraham | wonders if the requirement "User agents MUST implement all supported states and properties for the role" is as silly as it sounds |
| 10:32 | <hsivonen> | MikeSmith: even exposure to AT is only a "should"! |
| 10:32 | <zcorpan> | http://lists.w3.org/Archives/Public/public-xml-core-wg/2009Apr/0031.html |
| 10:32 | <jgraham> | "Assistive technology must treat an article like a document in that article must be processed like an application" makes no sense |
| 10:34 | <jgraham> | "When a value is indicated as the default, the behavior prescribed by this value MUST be followed when the state or property is not provided" - sounds almost like a conformance criterion |
| 10:37 | <jgraham> | I wonder if I can claim that section 6.1 requires two host langauges to implement @role, both of which must presumably reach REC before aria :) |
| 10:40 | <zcorpan> | jgraham: svg 1.2 tiny could be one of them |
| 10:41 | <hsivonen> | can SVG 1.2 tiny be one of them the same way HTML 4.01 can be one of them? |
| 10:42 | <zcorpan> | hsivonen: no, i think svg 1.2 tiny actually includes role="" |
| 10:42 | <hsivonen> | oh |
| 10:42 | <zcorpan> | but not the other aria-* |
| 10:42 | <hsivonen> | also, IIRC, SVG 1.2 Tiny can time-warp the REC track :-) |
| 10:45 | <annevk2> | jgraham, are you sending in comments? |
| 10:46 | <hsivonen> | annevk2: I sent a comment about form submission |
| 10:46 | <jgraham> | annevk2: No. Should I? I don't have any faith that they will be dealt with in a sensible way but if you think it is a good idea I will |
| 10:47 | <annevk2> | jgraham, yeah, lets give it the benefit of the doubt |
| 10:47 | <jgraham> | OK |
| 10:51 | <jgraham> | Has someone already commented on the use of must vs MUST? |
| 10:53 | <MikeSmith> | zcorpan: that message from Paul Grosso about avoiding "must" is disappointing |
| 11:02 | annevk2 | isn't even sure whether the behavior of aria-invalid="false spelling" is defined or aria-invalid="false x" for that matter |
| 11:03 | <zcorpan> | is the behavior of aria-invalid="false" defined? :) |
| 11:03 | <annevk2> | yes |
| 11:03 | <zcorpan> | ok |
| 11:04 | <annevk2> | "If the attribute is not present, or its value is false, or its value is an empty string, the default value of false applies." |
| 11:04 | <jgraham> | Does section 4.2.2 make any sense to anyone? |
| 11:05 | <annevk2> | jgraham, not a whole lot |
| 11:06 | jgraham | is starting to understand little |
| 11:07 | <jgraham> | s//a/ |
| 11:11 | <zcorpan> | hsivonen: http://alfonsopina.eresmas.net/ looks weird in minefield-html5 |
| 11:24 | <hsivonen> | zcorpan: just that the heading isn't centered? |
| 11:24 | <zcorpan> | hsivonen: no, i get text like "�������t hesitate to contact the author (see end of page).', TEXTFONT, 'Comic Sans MS', TEXTCOLOR, '#FFFFFF', FGCOLOR, '#3366ff') " onMouseOut="nd();">" |
| 11:25 | <hsivonen> | hmm. that's bad |
| 11:26 | <zcorpan> | and "Libro de visitas Ver &n�����a) " |
| 11:26 | <hsivonen> | zcorpan: what's your character encoding setting? |
| 11:27 | <zcorpan> | hsivonen: UTF-8 |
| 11:27 | <hsivonen> | if I switch to UTF-8, the result is totally bogus, indeed |
| 11:27 | <hsivonen> | I think I don't understand the Gecko unicode converter API contract right |
| 11:27 | <hsivonen> | zcorpan: thanks |
| 11:28 | <hsivonen> | I still wonder why the heading/header isn't centered |
| 11:28 | <zcorpan> | hsivonen: maybe because of <p><table>? i tested a random url from the list |
| 11:29 | <zcorpan> | hsivonen: if i select another encoding or check autodetect, it crashes |
| 11:29 | <hsivonen> | zcorpan: indeed, it's a <p><table> issue. |
| 11:29 | <hsivonen> | <p align=center> |
| 11:41 | zcorpan | wonders if it's possible to fix the styling issues in the parser by copying over attributes to the table |
| 11:42 | <hsivonen> | I wonder if it's possible not to close <p> if it has the align attribute |
| 11:42 | <hsivonen> | and whether that's even worse than having one quirk |
| 11:42 | <zcorpan> | oh yeah that's probably better than copying attributes |
| 11:46 | <zcorpan> | <P ALIGN=RIGHT><TABLE - http://123relax.holisticwebdirectory.com/drhypnosis/ |
| 11:51 | <zcorpan> | thinking about it, i wonder whether <p><form> would benefit from the same treatment |
| 12:00 | <zcorpan> | http://64.8.148.68/webpage/highschool.htm - <p align=center>...<table |
| 12:01 | <annevk2> | is it just <p align>? |
| 12:02 | <zcorpan> | from what i've looked at so far, yeah |
| 12:03 | <zcorpan> | <p align="center"> http://www.shopin.co.il/ |
| 12:06 | <zcorpan> | <p class="topbana"><table border="0" width="" cellspacing="0" cellpadding="3" style="width: 583px"> http://www.ace-company.net/~coolcat/ |
| 12:07 | <zcorpan> | .topbana { margin-top: 5; margin-bottom: 12 } |
| 12:07 | <zcorpan> | nothing too serious i guess |
| 12:08 | <annevk2> | it would regress layout of the page |
| 12:08 | <hsivonen> | maybe we should just fix Acid2 instead |
| 12:10 | <annevk2> | we could do that I suppose, but all UAs interoperate on this now |
| 12:10 | <hsivonen> | annevk2: yeah. I blame Hixie. |
| 12:12 | <annevk2> | we almost had this with comments too |
| 12:12 | <annevk2> | http://ln.hixie.ch/?start=1137799947&count=1 |
| 12:13 | <hsivonen> | I was thinking of that blog post, too |
| 12:17 | <zcorpan> | <p> <table> http://adsg.syix.com/ - would get more space above the table |
| 12:18 | <zcorpan> | http://agri.astate.edu/Ag%20Education/education.htm too |
| 12:21 | <zcorpan> | http://alexey-savrasov.ru/ has <p align=justify> blah blah <table align=right>...</table> blah blah |
| 12:22 | <hsivonen> | ooh. I now see my character decoding bug |
| 12:29 | <zcorpan> | ok i've looked at the 50 first urls |
| 12:30 | <zcorpan> | i think our options are (1) having the quirk, and (2) changing acid2 |
| 12:30 | <hsivonen> | zcorpan: I agree |
| 12:31 | <zcorpan> | looking at align="" is not enough |
| 12:33 | <hsivonen> | I wonder if we could get rid of the quirks mode these days if CSS 1) defaulted to px for unitless numbers and 2) defaulted to almost standards cell height |
| 12:33 | <hsivonen> | perhaps inheriting into tables should have been author opt-in via CSS itself way back when |
| 12:34 | <annevk2> | case-insensitive class names are important too I think and there are probably a few more |
| 12:35 | <zcorpan> | hsivonen: i'm pretty sure people use unitlessness as a css filter |
| 12:35 | <annevk2> | case-insensitive class names are not scoped to HTML by the way at this point |
| 12:35 | <zcorpan> | hsivonen: to fix the ie5 box model |
| 12:35 | <annevk2> | though actually getting rid of quirks mode would be awesome |
| 12:36 | <hsivonen> | it seems to me that after IE3 none of the CSS quirks have been so horrible that CSS2 developed according to HTML5 principles couldn't have had backwards-compatible default behavior |
| 12:36 | <hsivonen> | too late now, I guess |
| 12:36 | <hsivonen> | hindsight 20/20 |
| 12:36 | <hsivonen> | the attitude towards 'standards' was very different back then |
| 12:37 | <annevk2> | https://twitter.com/erikdahlstrom/status/1574170590 sweet |
| 12:37 | <annevk2> | guess that's next on my list of things to change on my site |
| 12:38 | hsivonen | notes that ed doesn't even mention IE |
| 13:07 | <hsivonen> | zcorpan, Philip`: thanks for doing the <p><table> research |
| 13:14 | <Philip`> | annevk2: It's a shame that Opera doesn't support APNG favicons properly, and only Mozilla does |
| 13:17 | <zcorpan> | Philip`: have you filed a bug? |
| 13:19 | <Philip`> | zcorpan: No |
| 13:19 | <annevk2> | whoa, air greeland has instant customer service via email, what a treat |
| 13:19 | <Philip`> | since I'd probably just be told that it was a design decision to not animate favicons, or something |
| 13:19 | <annevk2> | what do you mean with "properly"? |
| 13:21 | <zcorpan> | Philip`: we animate gif favicons i think |
| 13:21 | <jgraham> | annevk2: You're going to greenland? |
| 13:22 | <annevk2> | yup |
| 13:22 | <jgraham> | Nice |
| 13:22 | <annevk2> | yup :) |
| 13:24 | <Philip`> | annevk2: Animatedly |
| 13:25 | <annevk2> | i doubt it's a design decision since we let you animate them through JavaScript |
| 13:29 | <zcorpan> | annevk2: we let pages animate things through javascript when the "disable animated images" option is set, too |
| 13:31 | Philip` | attempts to upload a less-subtly-animated test |
| 13:36 | hsivonen | predicts that in future studies, Philip` will find role=complimentary on the Web |
| 13:37 | <Philip`> | http://philip.html5.org/tests/apng/favicon.html |
| 13:38 | <Philip`> | hsivonen: That seems to be presupposing that people will use role at all |
| 13:38 | <Philip`> | Oh, some people do use it already |
| 13:38 | <Philip`> | Especially http://www.tee-rail-alliance.de/ |
| 13:39 | <Philip`> | which uses loads of different roles |
| 13:39 | <hsivonen> | Philip`: that one uses role="complementary" |
| 13:40 | <Philip`> | and then there's one site with a role="navigation" on each page, and two with a role="" |
| 13:40 | <hsivonen> | Philip`: I meant that complementary is one of the English words that people can't spell |
| 13:42 | <Philip`> | hsivonen: Indeed, I was just wondering whether anybody would bother with role at all, because that's necessary before they'll have an opportunity to mis-spell it |
| 13:44 | <zcorpan> | does a label in javascript need to be followed by anything? i.e. is onclick="http://example.org"; a valid script? |
| 13:46 | <jgraham> | zcorpan: a label must be followed by a statement |
| 13:46 | <jgraham> | So no, that's not valid |
| 13:46 | <zcorpan> | ok |
| 13:46 | <zcorpan> | i don't seem to get anything in the error console though |
| 13:47 | <zcorpan> | i get an error in firefox |
| 13:47 | <zcorpan> | bug in opera? |
| 13:48 | <Dashiva> | hmm |
| 13:58 | <Dashiva> | I've seen similar cases before. Error in sort comparison function didn't cause error in sorting, error in a statement that would've been a nop was optimized away. |
| 14:02 | <hsivonen> | why would an user want to read an accessibility conformance claim? http://www.brucelawson.co.uk/2009/rel-accessibility/ |
| 14:02 | <hsivonen> | or a conformance claim of any kind? |
| 14:12 | <jgraham> | hsivonen: http://diveintoaccessibility.org/day_30_creating_an_accessibility_statement.html although it seems rather tenuous to me |
| 14:14 | <jgraham> | In particular it seems to be predicated on the assumption that you are using a UA that does not provide clues about the avaliable functionality |
| 17:40 | <sicking> | Hixie, ping |
| 17:43 | <annevk5> | weinig, did you ping me the other day? still needed? |
| 17:55 | <annevk5> | hmm, OWL 2 |
| 20:09 | <gsnedders> | Ow. In Saf4 for a display:table the 'width' property isn't the content width: the content width is 'width' - 'border-left'/'border-right' when it has borders (padding/margin is untested…) |
| 20:13 | gsnedders | stops going through bugs that need a reduction |
| 20:14 | <annevk5> | sounds like box-sizing is applied to table frames rather than HTML table elements or some such |
| 20:15 | <gsnedders> | Yeah, that's what I was guessing |
| 20:52 | <Hixie> | gsnedders: sounds vaguely like what the spec says to do |
| 20:52 | <gsnedders> | Hixie: Really? I had a look and it seemed wrong. Sites break if you implement the spec then, and Firefox and Opera don't follow it. |
| 20:53 | <gsnedders> | https://bugs.webkit.org/show_bug.cgi?id=13519 |
| 20:53 | <Hixie> | i forget exactly what the spec says |
| 20:54 | <gsnedders> | I didn't see any special case for a table itself, only table cells |
| 21:44 | <jgraham> | gsnedders: BTW if you put 200px in the pass condition for the test it is really annoying because you have to verify that the shapre is really 200px to confirm if the test has passed |
| 21:45 | <jgraham> | Which doesn't really seem to be necessary here |
| 21:45 | <gsnedders> | jgraham: The test is the widths are calculated correctly, so it does seem relevant |
| 21:46 | <a-ja> | hi folks. i may be missing something something, but is anyone aware of a reason placeholder shouldn't be supported on input type=search ? |
| 21:47 | <jgraham> | gsnedders: If you really care you should put a 200px green image and a 200px green box and make them line up and say something like "you should see a green rectangle below" |
| 21:48 | <jgraham> | Basically you don't want the pass condition to be anything that you can't verify in a second or two unless it is really super necessary |
| 21:49 | <jgraham> | (or hope that you have reftests so that you can say "this box should look exacly like this other box") |
| 21:50 | gsnedders | shrugs, and goes back to trying to shorten his English dissertation, as he thinks it is a bit more important |
| 21:50 | <jgraham> | a-ja: The only reason I can think of is some convention that input type=search should always have "search" as a placeholder |
| 21:51 | <jgraham> | gsnedders: Yeah that is more important :) Just a tip that may make some poor QA's life easier someday |
| 21:51 | <gsnedders> | Oh wow. I don't have the spec in my history |
| 21:52 | <gsnedders> | a-ja: placeholder is supported on input@type='search' |
| 21:53 | <a-ja> | jgraham: possible, i suppose...guess i'll followup with hixie, in case it's just an oversight |
| 21:53 | <a-ja> | oh, guess i *did* miss it, then |
| 21:53 | <gsnedders> | a-ja: search is grouped with text in the spec. What makes you think it isn't supported? |
| 21:56 | <a-ja> | haven't read spec section lately....just recent change log comments / diffs re:placeholder |
| 21:57 | <a-ja> | makes sense if search is just subtype of text |
| 22:15 | <gsnedders> | \emph{That's} whom you should have married, […], instead of all that we \emph{teased} her to death! |
| 22:15 | <gsnedders> | What should the position of comma be around the […]? |