| 12:55 | <boblet> | anyone want to give me some feedback on an HTML5 ‘choose the right element’ flowchart? |
| 12:57 | <Dashiva> | If you just post the link and hang around, someone might :) |
| 13:15 | <boblet> | Dashiva: up: http://oli-studio.com/temp/sectioning-flowchart-horiz.png :) |
| 13:16 | <Dashiva> | Shouldn't "any semantics" be the first block? |
| 13:16 | <Dashiva> | Or rather, non-sectioning semantics |
| 13:17 | <boblet> | Dashiva: do you mean the text in the first diamond? |
| 13:18 | <Dashiva> | I mean reordering the choices so the last one is first |
| 13:19 | <boblet> | I’m trying to start from the most specific stuff and get gradually more general, as the other direction won’t work |
| 13:20 | <boblet> | although I see what you mean in that I’m taking it as a given there are some semantics |
| 13:20 | <boblet> | (although that’s a reasonable assumption I think) |
| 13:20 | <Dashiva> | To take a somewhat contrived example, I'm writing a recipe as a list of steps |
| 13:20 | <Dashiva> | It works as a separate entry in a feed, so I should use <article> instead of <ol>? |
| 13:21 | <boblet> | aha, but it wouldn’t make sense in a feed without a title (recipe name) |
| 13:21 | <boblet> | or at least knowledge that the feed contained recipes |
| 13:22 | <boblet> | (might be worth adding “with a title” to <article> tho huh) |
| 13:22 | <Philip`> | boblet: The text rendering on that diagram looks ugly :-( |
| 13:22 | <Dashiva> | I think it's worthwhile to specify that the steps are to sort out sectioning elements |
| 13:22 | <Dashiva> | And that non-sectioning elements should be handled separately |
| 13:22 | <Philip`> | Also: s/it's/its/ |
| 13:23 | <boblet> | Philip`: yeah I know — I used a for-print serif font (making as a PDF), need to change it… |
| 13:24 | <boblet> | Dashiva: good feedback |
| 13:24 | <boblet> | Philip`: ouch, embarassing! |
| 13:24 | <boblet> | thanks |
| 13:24 | <Dashiva> | (Also, you have 'yes' choices, but no 'no' for the alternative paths) |
| 13:26 | <boblet> | Dashiva: that was done for space reasons, and in the hope that it’s obvious enough to not require labeling. not obvious enough huh? |
| 13:26 | <Dashiva> | It is, just a style nitpick |
| 13:27 | <boblet> | I was hoping to put a no between the final diamond and div |
| 13:38 | <AryehGregor> | How can the media fragments WG have existed for so long without even being sure how the media fragments are supposed to be displayed? Shouldn't the use-cases make that obvious? Or did they start with no use-cases, only "create a media fragment syntax"? |
| 13:46 | <Philip`> | AryehGregor: http://www.w3.org/2008/WebVideo/Fragments/WD-media-fragments-reqs/ looks like their use cases |
| 13:48 | <Philip`> | Seems to mix problem and solution without much analysis of what would actually be a suitable solution, though |
| 13:49 | <Philip`> | e.g. "Sebo is Deaf and enjoys watching videos on the Web. Her friend sent her a link to a new music video but she doesn't want to waste time and bandwidth receiving any sounds. So when she enters the URI in her browser's address bar, she also adds an extra parameter to select the video track only." |
| 13:49 | <Philip`> | Why would anyone want to use a URI for that, rather than configuring their browser to just never download audio? |
| 13:54 | <doublec> | it would be the same solution though wouldn't it? |
| 13:54 | <doublec> | If they flipped that switch in the browser the browser would use a media fragmment request on all media requests |
| 13:55 | <Philip`> | Requests don't contain fragments |
| 13:56 | <doublec> | I mean add the fragment request to the URI - or however the media fragment spec says it's done |
| 13:58 | <Philip`> | http://www.w3.org/2008/WebVideo/Fragments/WD-media-fragments-spec/MF-SD-ServerSideSetup.png seems to be what it suggests - the UA converts the URI fragment into some kind of HTTP headers depending on what data it needs from the server |
| 13:59 | <Philip`> | If the UA does it automatically then you don't need any URI fragment at all, you just need whatever HTTP features exist for requesting partial resources |
| 13:59 | <Philip`> | and it looks like the Media Fragments URI spec is only defining the URIs, not the HTTP features |
| 14:01 | <doublec> | you need the URI fragment so people can link to it in blogposts, etc |
| 14:01 | <doublec> | haven't you ever wanted to link to specific time ranges in a video cropped to a particular rectangle? |
| 14:03 | <Philip`> | That's a sensible use case for media fragments, but I was referring to the deaf user who supposedly modifies a URI manually in order to save time and bandwidth |
| 14:04 | <Philip`> | which seems a much less compelling use case |
| 14:04 | <doublec> | true |
| 14:04 | <doublec> | I fear that the grammar for fragments will evolve into some turing complete language to rip apart, demux, remux, crop, etc media |
| 14:04 | <doublec> | and be near impossible to implement and ue |
| 14:04 | <doublec> | s/ue/use |
| 14:06 | <Philip`> | Just let people encode entire Avisynth scripts in the fragment |
| 14:06 | <Philip`> | That should be sufficient power |
| 14:06 | <doublec> | hehe |
| 14:11 | <gsnedders> | Anyone taken a look at http://www.google.com/logos/js/pacman10-hp.3.js? |
| 14:11 | <doublec> | I saw the logo and played the game |
| 14:11 | <gsnedders> | doublec: I don't care about that : |
| 14:12 | <gsnedders> | * :) |
| 14:13 | <gsnedders> | Just seems to be using a div and Flash for the sound |
| 14:14 | <gsnedders> | Just loads of elements |
| 14:14 | <gsnedders> | Unusually readable for Google JS, quite a lot of identifiers still there |
| 14:15 | <doublec> | someone linked to a prettified version on reddit recently |
| 14:18 | <gsnedders> | Gah, the normal version with whitespace added in is really readable by Google terms. More people should try looking at Gmail. :) |
| 14:47 | <nessy> | note about the media fragments: this far the discussion have concentrated on time ranges and we're fairly confident that side is well defined now; but pictures and spatial stuff hasn't been discussed sufficiently |
| 14:48 | <nessy> | AryehGregor: keeping just a focus on a part of the picture rather than actually zooming in makes more sense for video than for single images |
| 14:49 | <nessy> | but the spec doesn't really prescribe the presentation, since that's to be discussed with browser vendors |
| 14:50 | <nessy> | just like the URI spec doesn't prescribe how to present html pages ;) |
| 14:57 | <boblet> | try two: http://oli-studio.com/temp/sectioning-flowchart2.png — Dashiva’s reordering (ruins the nice line I had going tho), screen capt from PDF not Graffle for less chunky fonts, addition of nav since it’s sectioning elements… |
| 15:02 | <Dashiva> | Hm, I was thinking div should be the last step, and appropriate element first (just to exclude non-sectioning) |
| 15:04 | <boblet> | Dashiva: wouldn’t that basically involve a question like “is this a sectioning element?”? :P |
| 15:05 | <Dashiva> | Yeah |
| 15:06 | <boblet> | I’m still a little unsure if I prefer the any semantics q at the start or end |
| 15:07 | <boblet> | I’ll run it past a few more people. need to get some feedback from neubs :) |
| 15:19 | <boblet> | pulling those flowcharts until they’re done (should be live at HTML5Doctors on Tuesday)… |
| 15:23 | <boblet> | Dashiva, Philip` — thanks for your feedback yo |
| 18:48 | <peta> | hello everybody |
| 18:51 | <peta> | I don't know if i'm right here ... I'm writing a little tool that extracts ttf/otf font tables from WOFF files and saves them into a generic ttf/otf file. However I've got problems with decompressing the font tables. Does someone know more about it? |
| 19:23 | <Dashiva> | peta: You're not supposed to do that :) |
| 19:24 | <peta> | Dashiva: ouh, you mean in regard to license/copyright violations? |
| 19:24 | <Dashiva> | Nah, that applies to copying the original woff files too |
| 19:25 | <peta> | nono ... that's not my intention. I just was curious about the format spec ... and wanted to try it for the sake of it |
| 22:18 | <zcorpan_> | http://krijnhoetmer.nl/irc-logs/whatwg/20100521#l-465 - brendan maybe? |
| 22:26 | <annevk> | TabAtkins, so that email about "Staying at my place" just arrived in my gmail folder... |
| 22:26 | <annevk> | TabAtkins, delay much? o_O |
| 22:27 | <zcorpan_> | html4 errata? |
| 22:28 | <jgraham> | What, where? |
| 22:28 | <Dashiva> | Isn't that html5? |
| 22:28 | <zcorpan_> | http://dev.w3.org/html5/profiles/drafts/ED-html5-profiles-20100522/ |
| 22:28 | <gsnedders> | I blatantly need to catch up on the mailing list |
| 22:29 | <jgraham> | Non normative errata? That couldn't posibly cause confusion |
| 22:29 | <gsnedders> | Is it inappropriate to complain that that document doesn't comply with ISO 2145? |
| 22:29 | <annevk> | lol |
| 22:29 | <annevk> | gsnedders, it's inappropriate to ask that more than once |
| 22:30 | <jgraham> | Or, to be brief, yes |
| 22:30 | gsnedders | hangs head in shame |
| 22:30 | <annevk> | gsnedders, about specifications complying with ISO 2145 that is and since you already did... |
| 22:30 | <Dashiva> | What exactly does the profile add in the rdfa example? |
| 22:31 | <jgraham> | Dashiva: I'm not sure I understand |
| 22:31 | <jgraham> | They seem to be suggesting it should change the predicate |
| 22:31 | <jgraham> | (s) |
| 22:32 | <Dashiva> | To me the tuples look like what I'd expect without the profile too |
| 22:32 | <zcorpan_> | adding more attributes seems like a good way to make rdfa easier to understand |
| 22:32 | <jgraham> | Dashiva: I hink you are supposed to assume no vocabularly-specific knowledge |
| 22:32 | <jgraham> | *think |
| 22:35 | <jgraham> | It is hard to tell though because the document only specifies that the profile attribute is allowed and that URIs are processed from left to right (dunno what that means really) |
| 22:36 | <jgraham> | So the triple generation stuff seems to be entirely unrelated to the rest of the document |
| 22:36 | zcorpan_ | wonders if the people in the Built-in image sprite support in HTML5 thread know about #SVGView() |
| 22:38 | <jgraham> | (I guess "These instructions and vocabulary terms apply to the current element and all descendant elements." is intended to be a normative requirement but it is extremely odd since this spec doesn't actually provide any processing for @profile. Therefore the relevant conformance class for that reuirement is presumably other specs) |
| 22:38 | <Dashiva> | So I guess it's just a talisman then |
| 23:51 | <boblet> | hey all, where are HTML5 unit tests living? |
