| 00:05 | <hober> | well, think of it in terms of postel's law. a UA is at the 'liberal in what you accept' end, matching your "don't mess" behavior. |
| 00:05 | <hober> | an editor / authoring tool is at the 'conservative in what you send' end |
| 00:06 | <hober> | hence authoring requirements being stricter than UA requirements |
| 00:16 | <sayrer> | hober, I am saying that those requirements are incompatibl |
| 00:16 | <sayrer> | e |
| 00:17 | <hober> | how are they incompatible? |
| 00:20 | <roc> | they're incompatible if you require round-tripping |
| 00:20 | <roc> | which is what most people want from an editor |
| 00:23 | <hober> | hmm. conformant documents round-trip with conformant tools. non-conformant documents round-trip if your tools are also non-conformant. |
| 00:24 | <takkaria> | gsnedders: yup, Linkoping |
| 00:46 | <ojan> | Hixie: ping |
| 01:11 | <Hixie> | ojan: here |
| 01:21 | <ojan> | Hixie: i'm looking at http://www.whatwg.org/specs/web-apps/current-work/#attr-fe-disabled |
| 01:21 | <ojan> | it's not clear to me from that description whether text selection in a disabled input should work |
| 01:22 | <Hixie> | looking... |
| 01:23 | <Hixie> | it's up to the UA |
| 01:23 | <ojan> | FWIW, IE allows text selection and Gecko/WebKit don't from my quick test |
| 01:23 | <ojan> | ok |
| 01:23 | <ojan> | thx |
| 01:24 | <Hixie> | do you want a note in there to that effect? |
| 01:25 | <ojan> | meh. i guess text selection stuff is mostly up to the UA, right? |
| 01:25 | <ojan> | so a note just in that one instance doesn't seem necessary |
| 01:25 | <Hixie> | ok |
| 02:46 | <adu> | hi |
| 07:39 | <hsivonen> | Hixie: yes, I pointed out the absolute URL thing to DanC |
| 07:58 | <annevk42> | seems that Web ECMAScript needs to define (new Date(NaN).toString()) as well |
| 08:02 | Mrmil | received a xhtml list of features coded in <br />'s from a top-notch application developer. *sigh* |
| 08:05 | <Hixie> | hsivonen: yt? |
| 08:06 | <hsivonen> | Hixie: yes |
| 08:07 | <Hixie> | hsivonen: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-May/019843.html and http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-May/019852.html |
| 08:07 | <Hixie> | hsivonen: how would you like me to proceed? |
| 08:07 | <Hixie> | hsivonen: do you think i should define a mapping to infoset for microdata as you describe? |
| 08:08 | <hsivonen> | If it's proven by implementation, yes. However, at this point it's not proven by implementation. |
| 08:08 | <hsivonen> | I guess you could put it in and cut it at CR if it sucks. |
| 08:10 | <Hixie> | i'm leaning towards waiting and seeing if there's demand |
| 08:10 | <Hixie> | would that be ok? |
| 08:11 | <Hixie> | and re your latest e-mail, i now understand what you meant by "literal non-characters", my bad |
| 08:14 | <hsivonen> | Hixie: yeah, it would be OK to wait with the XML mapping |
| 08:14 | <annevk42> | oh yay, Gecko puts elements in the XHTML namespace too now |
| 08:19 | <Hixie> | ok, thanks |
| 08:20 | <hsivonen> | annevk42: I really hope the change proves to be a non-event as far as Web compat goes |
| 08:20 | <annevk42> | from prior ruby feedback it seems better to just obsolete the fallback mechanism |
| 08:21 | <annevk42> | (and require everyone to support ruby properly) |
| 08:21 | <hsivonen> | annevk42: what about degrading gracefully? |
| 08:24 | <hsivonen> | https://twitter.com/plhw3org/statuses/2090392921 |
| 08:26 | <annevk42> | hsivonen, e.g. myakura said the fallback was not good |
| 08:26 | <annevk42> | hsivonen, heh, I see you replied as well, me too: https://twitter.com/annevk/status/2090405819 |
| 08:30 | <annevk42> | cheap shot I suppose |
| 08:31 | <remysharp> | hi chaps (and chapesses?), I wanted to sign up to the mailing lists, but which one is best to give feedback on implementing the markup? |
| 08:31 | <remysharp> | is it the web designers and html authors - or |
| 08:31 | <remysharp> | implementations (though I think this might be the browser implementors) |
| 08:31 | <remysharp> | *implementers |
| 08:34 | <annevk42> | remysharp, questions go to help, feedback goes to whatwg, and questions on implementing features in user agents (e.g. html5lib, browsers, etc.) go to implementors |
| 08:35 | <remysharp> | okay, so feedback on how elements work or are used from an authors point of view, I want whatwg⊙wo then. |
| 08:35 | <remysharp> | cheers. |
| 08:35 | <Hixie> | macs really need to get better at handling hard disk failures |
| 08:36 | <Hixie> | hanging the entire OS when the disk fails to respond is not acceptable |
| 08:36 | <Hixie> | (external disk) |
| 09:08 | <annevk42> | maybe we should introduce the concept "legacy fallback" |
| 09:08 | <annevk42> | to make it clear how various elements differ in fallback model |
| 09:10 | <Hixie> | hsivonen: doesn't seem to be much point pushing <video> for freeness if one then uses java as the fallback |
| 09:10 | <hsivonen> | making Hixie allow <applet> and/or <object classid> in <video> content would be a start... |
| 09:10 | <hsivonen> | Hixie: Java is royalty-free, no? |
| 09:11 | <hsivonen> | Hixie: also, it's about making video play in IE and Opera which are both already proprietary |
| 09:11 | <Hixie> | if all you care about is money, it might be ok, i dunno |
| 09:12 | <hsivonen> | Hixie: I care about Freedom |
| 09:12 | <Hixie> | (money isn't the part of freedom that i particularly care about) |
| 09:12 | <Hixie> | java is basically a single-vendor technology for all intents and purposes |
| 09:12 | <hsivonen> | Hixie: to have Free as in Freedom implementations of the platform, it's useful to enable the same videos to play in IE |
| 09:13 | <hsivonen> | and that particular mechanism may be non-Free but royalty-free |
| 09:13 | <Hixie> | didn't we establish that java could use <object> or <embed> |
| 09:14 | <hsivonen> | Hixie: have we established that Java could use <object> in IE without classid? |
| 09:14 | <Hixie> | no idea |
| 09:14 | <Hixie> | i'm finding it difficult to care about java fallback in IE |
| 09:15 | <hsivonen> | Hixie: is your story that Cortado fallback should focus on IE only? (If you don't allow <applet>, it's pretty obvious that authors who put in a fallback won't bother tweaking it for Opera) |
| 09:15 | <annevk42> | are you sure <object classid> does not work in Opera? |
| 09:15 | <hsivonen> | Hixie: I think I have an extraordinarily legitimate use case for Java applets here |
| 09:15 | <hsivonen> | annevk42: I am not sure |
| 09:15 | <Hixie> | frankly personally i don't really see why anyone is using video at all ye, since we don't have a codec yet |
| 09:18 | <hsivonen> | hmm. if it turns out that Opera pays attention to classid, classid would no longer be single-vendor technology |
| 09:19 | <Hixie> | classid is already defined in html5 |
| 09:19 | <hsivonen> | whoa. I just looked |
| 09:19 | <Hixie> | it's just not conforming, because it's values aren't necessary to obtain the effect it does |
| 09:19 | hsivonen | looks again |
| 09:19 | <hsivonen> | well, then |
| 09:19 | <Hixie> | (type="" does everything classid does) |
| 09:20 | <Hixie> | s/it's/its/ |
| 09:20 | <hsivonen> | Hixie: do you have a demo for bootstrapping Java in IE using object without classid? |
| 09:20 | <Hixie> | no, i don't really care enough about java to have a demo |
| 09:21 | <hsivonen> | Hixie: how do you know that "(type="" does everything classid does)" |
| 09:21 | <hsivonen> | in the present tense |
| 09:21 | <hsivonen> | if you haven't demos |
| 09:21 | <Hixie> | classid="" is a dispatch mechanism. type="" is a dispatch mechanism. we only need one dispatch mechanism. |
| 09:21 | <hsivonen> | but the present tense has all the relevance to *fallback* |
| 09:22 | <Hixie> | there is no present tense with <video> |
| 09:22 | <Hixie> | <video> has no defined codec yet. |
| 09:22 | <Hixie> | so nobody should be using it. |
| 09:22 | <Hixie> | and if they do, they are on their own |
| 09:22 | <hsivonen> | Hixie: there is a present tense to having a fallback in IE8 |
| 09:25 | <Hixie> | there is no need for fallback when there is nothing to fallback _from_ |
| 09:26 | <hsivonen> | Hixie: if we get a codec in the future, falling back in IE8 will be relevant |
| 09:26 | <hsivonen> | Hixie: by your logic, we could make <video> itself non-conforming while pondering the codec |
| 09:27 | <Hixie> | who knows what the fallback needs will be then |
| 09:27 | <Hixie> | <video> _is_ non-conforming today |
| 09:27 | <Hixie> | html5 is a draft |
| 09:27 | <annevk42> | you're talking past each other |
| 09:27 | <annevk42> | right |
| 09:27 | <Hixie> | hasn't even reached last call |
| 09:27 | <hsivonen> | Hixie: yet, in practice, the browsers that have shipped <video> or are about to have exactly one common codec set |
| 09:27 | <annevk42> | Hixie is talking about the future and hsivonen is talking about now |
| 09:27 | <Hixie> | hsivonen: they are? |
| 09:28 | <hsivonen> | Hixie: with XiphQT for Safari, yeah |
| 09:28 | <Hixie> | hsivonen: one browser doesn't even have <video> |
| 09:28 | <Hixie> | hsivonen: and if we can ask people to download a new codec, why not ask the other people to download a custom plugin? |
| 09:29 | <hsivonen> | Hixie: you still get out-of-the-box goodness in products from two other vendors and you bypass the plug-in prison on Snow Leopard (AFAICT) |
| 09:30 | <Hixie> | (for some definition of "goodness") |
| 09:43 | <Hixie> | ok i can't get java to work even with <applet>, so the odds of me getting it to work with anything else are minimal |
| 09:44 | <Hixie> | it crashed firefox twice, IE refuses to render anything but a red cross, and safari claims java can't be enabled. |
| 09:44 | <Hixie> | which from what i hear is a good thing since i don't think apple has fixed the java exploit yet |
| 09:50 | <Hixie> | ah, i don't seem to even have java installed. |
| 09:51 | <Hixie> | oh jesus java tried to install the MSN toolbar |
| 09:51 | <Hixie> | hsivonen: i'm less and less convinced that this is a useful route to follow for video fallback |
| 09:52 | <hsivonen> | Hixie: note that this is for casual windows users who have already installed the JRE for other reasons |
| 09:53 | <hsivonen> | Hixie: not for non-windows users who take a closer look at what they install |
| 09:54 | <Hixie> | the installer seems to have hung |
| 09:54 | <Hixie> | so the odds of casual windows users having it installed seem low to me :-) |
| 09:55 | <hsivonen> | Hixie: I'd agree with you based on my own installation experience, but random XP boxen seem to have the Java thingy in tray surprisingly often |
| 09:55 | <hsivonen> | Hixie: I don't know how it ends up installed, but it seems it does |
| 10:02 | <Hixie> | hsivonen: http://damowmow.com/playground/demos/java/001.html |
| 10:03 | <hsivonen> | Hixie: the second </applet> should be </object>, right? |
| 10:03 | <Hixie> | oops |
| 10:03 | <Hixie> | fixed |
| 10:04 | <Hixie> | man windows 7 sounds more like the wii than a serious OS |
| 10:04 | <annevk42> | of course, "application/x-java-applet" is not really conforming either, but I guess the validator does not go that far |
| 10:05 | <Hixie> | i'll let sun worry about _that_ |
| 10:06 | <Hixie> | notwithstanding the fact that the video doesn't actually play for me, i got it working for <applet>, <object>, and <embed> in IE8 without violating HTML5 for <object> or <embed>. |
| 10:06 | <hsivonen> | Hixie: OK. cool.. |
| 10:06 | <hsivonen> | I guess I shouldn't trust Sun's documentation |
| 10:06 | <hsivonen> | at all |
| 10:07 | <annevk42> | none of it works for me btw; guess I don't have Java |
| 10:07 | <Hixie> | i get as far as cortado loading and saying "buffering... 25%" (for some arbitary %) |
| 10:07 | <Hixie> | sometimes cortado crashes with a NullPointerException |
| 10:07 | <Hixie> | and it never plays the video back |
| 10:08 | <Hixie> | but i figure those problems are besides the point here |
| 10:11 | <hsivonen> | http://hsivonen.iki.fi/test/moz/video-fallback-validation/object-type.html WFM in IE8 |
| 10:11 | <hsivonen> | and validates |
| 10:13 | <nessy> | for me, safari on the mac buffers 1% of each of the three videos; firefox 3.5 doesn't display any of them |
| 10:13 | <hsivonen> | http://hsivonen.iki.fi/test/moz/video-fallback-validation/object-type.html WFM in Opera on Windows, too. |
| 10:14 | <hsivonen> | I'll close the bug. sorry about the false alarm |
| 10:14 | <annevk42> | is already closed |
| 10:14 | <hsivonen> | oh. Hixie closed it already. |
| 10:14 | <Hixie> | :-) |
| 10:15 | <Hixie> | no idea why it works for you but not me |
| 10:15 | <Hixie> | that's weird |
| 10:15 | <Hixie> | oh wel |
| 10:15 | <Hixie> | l |
| 10:15 | <Hixie> | (for your page but not mine, i should say) |
| 10:16 | <hsivonen> | Hixie: on surface, I notice that I used .ogg instead of .ogv and I have carefully configured Apache's types in advance |
| 10:16 | <hsivonen> | Hixie: Also, I obtained cortado fresh from svn |
| 10:16 | <hsivonen> | I'm positively surprised that it compiled with no classpath hell whatsoever |
| 10:19 | <Hixie> | i used the cortado build on their ftp site from just now, and use the same mime type as you |
| 10:19 | <Hixie> | do you think they examine the filename? |
| 10:19 | <Hixie> | i could change that i guess... |
| 10:20 | <Hixie> | same result |
| 10:20 | <Hixie> | <applet> crashed, <object> says "Buffering...", and <embed> is blank. |
| 10:20 | Hixie | continues to experience hell whenever dealing with java |
| 10:30 | <jgraham> | "[@summary] may just be a one trick pony. It does that trick well". Well I hope it is now obvious why I think lots of accessibility people are uninterested in enganing in a feedback loop about its usefulness |
| 10:30 | <jgraham> | *engaging |
| 10:32 | <Hixie> | they seem very interested in engaging in a feedback loop if by "feedback loop" you mean a loop where they give the same feedback repeatedly ignoring all responses |
| 10:32 | <jgraham> | Heh. That wasn't quite what I had in mind... |
| 10:33 | <Hixie> | (and by "they" i mean specifically a dozen or so people in public-html, not the whole accessibility community, many of whom have privately told me that they agree that summary="" has failed) |
| 10:34 | <annevk42> | Hixie, unfortunately private communication cannot be verified by anyone so is hardly useful for common understanding of things (imo) |
| 10:34 | <Hixie> | indeed |
| 10:35 | <Hixie> | hence why i have been focusing on the research and not making claims to authority in my arguments |
| 10:36 | <annevk42> | you did make a few with respect to the US gov, to be fair |
| 10:36 | <Hixie> | hm? |
| 10:36 | <annevk42> | but they were not that relevant to the overall point |
| 10:36 | <Hixie> | not sure which case you're talking about |
| 10:39 | <annevk42> | I recall something like "even the US gov discourages summary=" |
| 10:40 | <annevk42> | but alas, I cannot find it |
| 10:42 | <Hixie> | there was a recent post where i said that, but that was citing the documentation that the fpwg gave |
| 10:50 | <Hixie> | ok bed time |
| 11:26 | <Dashiva> | Is there a term for giving examples that are nothing at all like the relevant subject at hand? |
| 11:28 | <jgraham> | Dashiva: For example? |
| 11:32 | <Dashiva> | Like talking about handicap parking in the @summary discussion |
| 11:33 | <jgraham> | Oh I'm not sure what that's called |
| 11:33 | <jgraham> | Although I think Shelly was just wrong |
| 11:34 | <jgraham> | If you measured the impact of disabled spaces then I'm pretty sure that they would come out well |
| 11:34 | <jgraham> | compared to the alternatives |
| 11:35 | <Dashiva> | What is the parking equivalent of summary on a layout table? |
| 11:36 | <jgraham> | I guess it is a disabled person using the space. But to be fair I think she was trying to make a general (and as I said, IMHO, wrong) point so there need not be a 1:1 correlation |
| 11:38 | <Dashiva> | Well, yeah. Any example to something regulated by business and law is somewhat irrelevant by default when it comes to a volunteer and amateur web |
| 11:39 | <jgraham> | My assumption is that she believes that an "empiracist" would regard disabled parking spots as a failure if it could be shown that they were more often used by non-disabled people than by disabled people |
| 11:39 | <jgraham> | It might be true that they are used in that way. But that doesn't make them a failure |
| 11:40 | <jgraham> | Because they might nevertheless increase the accessibility of shops to disabled people by increasing the chance that they will get a spot near the shops |
| 11:41 | <jgraham> | Thus allowing them to go shoppiing when they would otherwise not be able to |
| 11:42 | <Dashiva> | I've been avoid calling it an analogy because it's a very bad one :) |
| 11:42 | <jgraham> | On the other hand they would be a failure if they were less cost effective at achieving this goal than some other solution that could be used instead. |
| 11:42 | <jgraham> | And were preventing that alternative solution being deployed |
| 11:43 | <Dashiva> | You also have to look at the overall result. Non-disabled use of disabled parking still produces utility. |
| 11:44 | <Dashiva> | Whereas an ignored attribute is pure author waste |
| 11:44 | <jgraham> | True |
| 11:46 | <jgraham> | (as an example of an alternative solution, we can imagine requiring businesses to ferry disabled users directly to the door from any parking space. This would have greater benefits (door to door service) but also much greater costs) |
| 12:19 | <annevk42> | I have the feeling some of the ECMAScript folks have been working on "obsoleting" the origin security infrastructure whereas we are working on strengthening it |
| 12:20 | <annevk42> | E.g. they are working on safe subsets of ECMAScript that can run within the same origin somehow for ads etc. whereas we have the sandboxed <iframe> feature to give those ads a distinct origin |
| 13:00 | <annevk42> | (My main worry though is that I am bringing this up and nobody else.) |
| 13:02 | <jgraham> | annevk42: I agree ith you but I don't kno if it is a problem |
| 13:04 | <annevk42> | jgraham, maybe it's just a problem for me then |
| 13:04 | <jgraham> | annevk42: Why is it a problem to have two possible approaches to the same problem, in this case? |
| 13:04 | <annevk42> | jgraham, some of the ECMAScript folks are asking for specific XMLHttpRequest changes that do not seem to take into account the feature set that sandboxed <iframe>s offer |
| 13:05 | <jgraham> | Oh |
| 13:05 | <jgraham> | Well I guess it depends on the merits of those changes |
| 13:05 | <annevk42> | jgraham, one security researcher makes claims about CORS based on the assumption that the Web will go in the direction of Caja/etc. |
| 13:06 | <annevk42> | jgraham, the merits heavily depend on where we end up, security wise |
| 13:06 | <jgraham> | I think it is bad to assume that either technique will dominate |
| 13:07 | <annevk42> | you also think it is bad to have several ways to do the same thing |
| 13:08 | <annevk42> | and given the API bloat that XMLHttpRequest already has I'm not inclined to sprinkle around some more |
| 13:09 | <jgraham> | In general I agree but with something like security there is more of an argument for more than one way to do it |
| 13:09 | <jgraham> | Assuming the ways are complementary and don't crowd each other out of the market |
| 13:10 | <jgraham> | (with different browsers implementing different solutions) |
| 13:11 | <jgraham> | Or cause undue confusion about the right approach to use |
| 13:12 | <annevk42> | I think you need one approach as it affects how ad platforms have to work, how you have to do communication securely, whether origins can be trusted, etc. |
| 13:13 | <jgraham> | I don't think we will (in general) want to assume different origins can be trusted any time soon |
| 13:13 | <annevk42> | it's about whether you can trust your own |
| 13:13 | <jgraham> | ad platforms should just pick whichever solution meets their needs best |
| 13:15 | <jgraham> | annevk42: Well I don't really understand the details... I assume it has something to so with embedding external script in your own origin but using caja to limit what it can do? |
| 13:16 | gsnedders | grumbles at Adobe |
| 13:16 | <annevk42> | yes |
| 13:16 | <annevk42> | versus embedding it inside a sandboxed iframe |
| 13:16 | gsnedders | has sent a form of verification they say is valid and they're saying it isn't |
| 13:16 | <jgraham> | (it is possible that my opinions on this topic are nonsensical since I don't really understand it all that well and it is rather complex) |
| 13:17 | <gsnedders> | "Official, current report card indicating name of school and student" — uh, yes, what I sent is that. |
| 13:17 | <jgraham> | annevk42: So what, in detail, is the problem with allowing ad networks / authors to choose whichever approach they feel is best |
| 13:18 | <annevk42> | it complicates XHR and potentially other APIs |
| 13:18 | <jgraham> | How much? |
| 13:18 | <gsnedders> | Oh, duh. |
| 13:18 | <jgraham> | Does it break back compatibility for example? |
| 13:18 | <annevk42> | needs a new constructor that changes behavior |
| 13:19 | <jgraham> | That seems like a pretty big change. What is the new behaviour |
| 13:19 | <jgraham> | ? |
| 13:19 | <annevk42> | acting as if the sending origin was null and not include any credentials in the request |
| 13:20 | <annevk42> | a sandboxed iframe gives you exactly that |
| 13:20 | <jgraham> | Yeah |
| 13:21 | <jgraham> | Well maybe it is worth trying to reduce the overlap between the two solutions if thay are going to cause significant duplication of functionality but in different places |
| 13:24 | <annevk42> | my current tactic is "look, sandboxed iframes" |
| 15:39 | <tantek> | Hixie, classid and type do have different semantics though (what piece of code is preferred to run this thing, what is the data type of this thing), that have been shown to have utility in various systems, even before the web. |
| 15:39 | <tantek> | E.g. MacOS files have the notion of Type and Creator. If the Creator is found for opening a file then it is used, otherwise, the system looks for what other applications support the Type of document, and give the user the option of opening the document in one of those apps. |
| 15:40 | <tantek> | And you could certainly base support for classid on existing web content that uses it. |
| 15:41 | tantek | is not a huge supporter of classid, but there are the facts behind it that should be considered from a scientific perspective before rejecting/removing it. |
| 15:47 | <gsnedders> | In title case, should à be capitalized (en français)? |
| 17:52 | <annevk42> | someone joked about grids for the Web and it taking ten years |
| 17:52 | <annevk42> | first time it was discussed on a public list was early '95 apparently: http://lists.w3.org/Archives/Public/www-style/1995Jun/0000.html |
| 17:53 | <annevk42> | where "early" means "halfway" |
| 18:03 | <beowulf> | don't we already have grids in tables... |
| 18:04 | beowulf | gets his coat |
| 18:44 | <jgraham> | annevk42: The lack of grid layouts is far and away the biggest failing of CSS |
| 18:44 | <jgraham> | Probably one of the biggest failings in the web standards world |
| 18:45 | <annevk42> | The group responsible for CSS mostly appears interested in print and documents |
| 18:46 | <annevk42> | Though to be fair grid layout is more seriously being worked on now |
| 18:47 | <jgraham> | annevk42: It has been "being worked on" for a long time. There have been drafts since at least 2000 right? |
| 18:47 | <annevk42> | I don't think so |
| 18:48 | <annevk42> | I think flex is really cool, but there's nobody to really drive it :/ |
| 18:48 | <krijnh> | Wasn't Hixie planning to pull a WHATWG on the CSSWG? |
| 18:49 | <jgraham> | It is pretty sad that Mozilla havn't managed to drive it to completion |
| 18:49 | <annevk42> | Though we do have a draft of it now: http://dev.w3.org/csswg/css3-flexbox/ |
| 18:49 | <krijnh> | Somewhere around 2023 |
| 18:49 | <annevk42> | But that is the old draft |
| 18:49 | <annevk42> | krijnh, we need an editor first :) |
| 18:50 | <krijnh> | By that time we can clone people |
| 18:50 | <jgraham> | annevk42: Is dbaron not working on it anymore? |
| 18:51 | <annevk42> | jgraham, I believe he is looking into it now, yes |
| 18:51 | <jgraham> | awesome |
| 18:51 | <jgraham> | I hope he is able to finish it soon |
| 18:53 | <annevk2> | problem with flex is that it's 1D; it's not grids |
| 18:54 | <annevk2> | there was some 2D flex proposal floating around based on tables, but only in email and very vague |
| 18:54 | <jgraham> | Does XUL have more than that? |
| 18:54 | <jgraham> | If webapps can have good enough layout primitives to do everything that XUL can do it is a huge win |
| 18:54 | <jgraham> | even if better ones would be nice |
| 18:55 | <annevk2> | XUL has some kind of weird grid model besides flex |
| 18:55 | <annevk2> | from what I heard we don't wanna copy that |
| 19:00 | <krijnh> | What's wrong with position: inline-absolute? http://www.shauninman.com/archive/2006/05/22/clearance_position_inline_absolute |
| 19:09 | <_trace> | question re:html5... |
| 19:09 | <_trace> | any good examples on 'article' and 'section' element usage? |
| 19:10 | <_trace> | should I be using a single 'article' element with multiple 'section's for each article on the same page, or multiple 'article's? |
| 19:11 | <jgraham> | _trace: multiple <article> elements |
| 19:11 | <jgraham> | One per article |
| 19:11 | <hober> | _trace: "for each article" suggests multiple <article>s |
| 19:11 | <jgraham> | <section> for subsecions of the article, if any |
| 19:11 | <_trace> | jgraham: thanks, that's what I thought, but I was seeing conflicting info in various articles |
| 19:12 | <_trace> | hober: thanks |
| 19:12 | <jgraham> | _trace: Got any examples of articles you found confusing? |
| 19:15 | <_trace> | well, http://www.alistapart.com/articles/previewofhtml5/ didn't specifically mention multiple articles, http://www.w3schools.com/tags/html5_article.asp mentioned 'article' was for content from an 'external' provider... |
| 19:15 | <jgraham> | Oh wow. w3schools ftw |
| 19:16 | <jgraham> | (it is totally misleading) |
| 19:16 | <_trace> | but the current draft of the HTML5 spec was vague, but I read it as meaning a single 'article' per article... |
| 19:16 | <_trace> | yeah, I ignored that since I didn't read the HTML5 draft that way at all |
| 19:16 | <jgraham> | If you have suggestions for clarifying the spec, you should talk to Hixie |
| 19:17 | <_trace> | jgraham: excellent |
| 19:23 | <annevk2> | krijnh, looks complex |
| 19:27 | <hober> | I only use Google's SearchWiki feature to remove w3schools results... |
| 19:28 | <annevk2> | the fact is though that nobody has come up with a good competitor to w3schools |
| 19:28 | <annevk2> | you'd think that a wiki/Web 2.0 like approach would be an instant hit |
| 19:30 | <annevk2> | maybe the competition gets all up in terminology all the time though and competes using "element" rather than "tag" |
| 19:31 | <krijnh> | annevk2: complex? |
| 19:31 | <krijnh> | :S |
| 19:34 | <annevk2> | krijnh, besides it doesn't do a whole lot of things that grids/flex would allow for |
| 19:34 | <sayrer> | grids/flex god |
| 19:34 | <krijnh> | True |
| 19:34 | <sayrer> | please |
| 19:35 | <sayrer> | I can't stand to read another article on lame CSS to simulate that |
| 20:12 | <rubys> | there is a required audio format, at the moment there is no required video formats, is (are?) there any required image format(s)? |
| 20:15 | <annevk2> | they're not listed |
| 20:15 | <annevk2> | but I believe the plan is to list them at some point |
| 20:17 | <annevk2> | PNG is required for <canvas> |
| 20:17 | <annevk2> | at this point |
| 20:18 | <rubys> | png as an *output* format |
| 20:18 | <annevk2> | good point |
| 20:19 | <jgraham> | We should probably list gif/jpeg/png since they are required to support the web |
| 20:19 | <annevk2> | (though everyone expects roundtripping to work of course) |
| 20:19 | <rubys> | svg's status seems unclear |
| 20:20 | <annevk2> | unclear how? |
| 20:21 | <rubys> | can a browser correctly claim to support HTML5, but support only gif/jpeg/png as image formats? |
| 20:21 | <jgraham> | An image format in <img>? |
| 20:22 | <jgraham> | I don't see why not at the moment |
| 20:22 | <annevk2> | SVG and MathML are part of HTML5 in a way |
| 20:22 | <sayrer> | I think it needs to support ico |
| 20:22 | <rubys> | at the moment, that appears to also be true for png; but I'm not asking about "at the moment", but the intent |
| 20:22 | <jgraham> | I don't know if we have defined what should happen for svg-in-<img> yet |
| 20:23 | <jgraham> | sayrer: .ico in <img> or for favicons? |
| 20:23 | <sayrer> | jgraham: hmm, not sure about img, but browsers can also view favicons in the content window |
| 20:23 | <jgraham> | (svg-in-<img> has differnt behaviour compared to <svg>-in-html with respect to e.g. scripting) |
| 20:24 | <rubys> | jgraham: if it has any behavior at all :-) |
| 20:24 | <annevk2> | heh |
| 20:24 | <jgraham> | sayrer: How? Is it needed for webcompat? You could make a browser today that didn't have favicons |
| 20:25 | <jgraham> | and the UI would probably suck but otherwise it would be OK, no? |
| 20:25 | <annevk2> | I'm not sure what the complete list is rubys, but I imagine APNG to be there |
| 20:25 | <annevk2> | and BMP |
| 20:27 | <annevk2> | heh, this <nfsw> thing has a whole share of followers suddenly |
| 20:27 | <annevk2> | e.g. http://twitter.com/corewarrior/statuses/2107076157 |
| 20:28 | <annevk2> | it's also being reported as part of HTML5: http://twitter.com/silner/statuses/2106913972 |
| 20:28 | <rubys> | s/nfsw/nsfw/ |
| 20:29 | <jgraham> | We totally need a National Freedom of Speech Week tag |
| 20:31 | <annevk2> | and also nfslw then |
| 20:32 | <annevk2> | to please our anti-fanboy |
| 20:32 | jgraham | can't work out all the letters |
| 20:33 | <annevk2> | Last Week! |
| 20:33 | <jgraham> | It was the nfs I was having trouble with... |
| 20:34 | <annevk2> | on the one hand it seems kind of bad for HTML5 to require specific formats, but on the other hand not doing it for things like e.g. <canvas> would be very impractical |
| 20:34 | <annevk2> | oh heh, nsflw |
| 20:34 | <annevk2> | jgraham, I just based it on yours |
| 20:35 | <jgraham> | Oh OK. I was hoping for "Not Safe For Mr Last Week" |
| 20:36 | <jgraham> | <nsfmlw>Hixie is my God and the W3C sucks</nsfmlw> |
| 20:36 | <jgraham> | That sort of thing |
| 20:36 | <jgraham> | Then when he read the logs it would appear as "Mr last week rules!" |
| 20:39 | <annevk2> | it's amazing there's still 73 people in this channel with all the nonsense that's going around :p |
| 20:41 | <jgraham> | I assume there are 71 people ignoring this channel |
| 20:42 | <annevk2> | prolly true, although I'm here twice |
| 20:57 | <Hixie> | did the pfwg just announe their intention to violate w3c process? |
| 20:57 | gsnedders | points out that we did that ages ago |
| 20:58 | <Hixie> | we did? |
| 20:59 | <jgraham> | Hixie: Pointer? |
| 20:59 | <gsnedders> | Is not basing decisions on their logical merits against the process as it isn't consensus? |
| 20:59 | <Hixie> | http://www.w3.org/mid/20090610153947.GE3371⊙srn |
| 21:00 | <jgraham> | gsnedders: We haven't published anything at the W3C without consensus |
| 21:00 | <Hixie> | gsnedders: but that's not the process the wg is following, that's only the process i'm following to make my chartered first proposals |
| 21:00 | <gsnedders> | That is true. |
| 21:02 | <jgraham> | is insure an en-US spelling of ensure? |
| 21:04 | <Hixie> | it appears to be an often used alternative spelling in US-english |
| 21:04 | <Hixie> | (insure also means to obtain insurance, but that's separate) |
| 21:05 | <Hixie> | 37 new e-mails overnight. Man. Every day 100 steps forward, 50 steps back. |
| 21:05 | <jgraham> | It's one of those things that grates ever so slightly when I see it even though I know that is an illogical reaction |
| 21:05 | <jgraham> | (insure that is, not new email) |
| 21:05 | gsnedders | wonders whether the latest 50 emails on public-html are worth reading |
| 21:06 | <Hixie> | i don't know, i just put them on my pile and will read them next month or so |
| 21:07 | <gavin_> | "insure" as "ensure" bothers me too |
| 21:11 | <annevk2> | please tell me if I make a mistake like that |
| 22:14 | <Lachy> | people who use insure in place of ensure simply don't know how to spell. Anyone who claims it's a valid alternative spelling is wrong |
| 22:14 | <roc> | many dictionaries do |
| 22:16 | <Hixie> | i was in fact quoting the dictionary macs ship with |
| 22:17 | <gavin_> | see usage note at http://www.answers.com/topic/assure |
| 22:22 | <Lachy> | Wow, that's even crazier than people who spell "percent" as "per cent". My Australian dictionary only has the insurace related meanings for insure. |
| 22:25 | <gsnedders> | Lachy: You realize in en-gb per cent is normal? |
| 22:25 | <Lachy> | gsnedders, yes, it's the one case where en-gb is stupid |
| 22:26 | <jgraham> | Lachy: You realise that once upon a time percent was probbly spelt per cent, right? People get things "wrong", language changes, grammer nazis find something new to complain about |
| 22:26 | <gsnedders> | Lachy: No, it's just we haven't got around to putting it together yet, right? |
| 22:26 | <jgraham> | *probably |
| 22:26 | <gsnedders> | Lachy: It was once everywhere two words |
| 22:26 | <Hixie> | unless en-au has a normative spec, there's not much point arguing over what is right |
| 22:27 | <Hixie> | that's why i have a normative spec for en-gb-x-hixie :-) |
| 22:27 | <Lachy> | it makes no sense as two words because percentage is one word |
| 22:27 | gsnedders | points out there is o normative spec for en-gb either :P |
| 22:27 | <Lachy> | actually, en-au is wrong on that case too |
| 22:27 | <gsnedders> | en-gb-x-sneddy really ought to have a spec |
| 22:27 | <jgraham> | Woah, back up there. You seem to be trying to apply logic to spelling rules |
| 22:27 | <Philip`> | Lachy: "percentage" makes no sense as one word because "per cent" is two |
| 22:28 | <jgraham> | It isn't going to work |
| 22:29 | <Lachy> | it also makes no sense because the word cent on its own has its own meaning as a unit of currency, and when you write it as per cent it makes it appear as though it's referring to the currency, not % |
| 22:29 | <jgraham> | Fun fact: There is research into double negatives that shows that languages go in cycles on whether double negatives are used for negation or not |
| 22:30 | <Philip`> | Lachy: You think words should only ever have one meaning? |
| 22:30 | <Lachy> | no, I didn't say that |
| 22:30 | <jgraham> | So a language will start with single negation and people will start using double negaion for emphasis. Then people will again start to favour single negation for brevity and the cycle will repeat |
| 22:30 | <jgraham> | Or something |
| 22:31 | <gsnedders> | Lachy: cent also means one hundred :P |
| 22:31 | <gsnedders> | (At least in Latin, and Latin is a proper language) |
| 22:31 | <Lachy> | Philip`, consider the phrase "I got 5 per cent". Depending on context, that could mean "I got 5 [items] per cent (currency)" or "I got 5%" |
| 22:32 | <Lachy> | writing it without the space removes the ambiguity and makes it easier to read |
| 22:33 | <Philip`> | ("per cent" is apparently an abbreviation of "per centum", so the two-word form seems the more original) |
| 22:34 | <jgraham> | gsnedders: The phrase "proper language" used without a smilie is sure to raise ire |
| 22:34 | <Philip`> | Lachy: Trying to make all sentences unambiguous independent of context seems like a battle that was lost millennia ago |
| 22:35 | <Philip`> | Context matters, and this isn't an example that seems ambiguous when you do have context |
| 22:36 | <jgraham> | I would wager that thre has never been a language that could be understood in an entirely context free way |
| 22:36 | <jgraham> | Because the difficulty of arriving at such a thing would be huge yet its value would be rather small |
| 22:38 | <Philip`> | Not even Lojban |
| 22:43 | <jgraham> | I guess languages with no (native) speakers don't count. But even so I wonder if it is truly context free |