| 00:00 | <Philip`> | http://code.google.com/p/nativeclient-sdk/source/browse/trunk/src/examples/hello_world/hello_world.html#84 - lovely |
| 00:01 | <jcranmer> | Philip`: well, I suppose the people who want to use it are mostly the people who want to screw compatability |
| 00:01 | <bga_> | jcranmer https://bugzilla.mozilla.org/show_bug.cgi?id=255942 79 votes |
| 00:02 | <TabAtkins> | Philip`: Ah, it looks like maybe we support arm there? In addition to x86-32bit and x86-64bit. |
| 00:02 | <gsnedders> | What about all these MIPS-based internet-enabled TVs? |
| 00:03 | <jcranmer> | bga_: no, 35 votes |
| 00:03 | <Philip`> | What about everybody who's going to not bother compiling for x86-64 because there's not enough users to be worth the bother? |
| 00:04 | <TabAtkins> | I dunno how all this works. |
| 00:04 | <Philip`> | Seems it'll be exactly like plugins, just minus the security issues |
| 00:32 | <TabAtkins> | Philip`: Yeah, that's basically what it is - a native plugin system. |
| 00:33 | <jacobolus> | beowulf: okay, I added a   glyph to Delicious (and even made it slightly thinner than a usual  ), and now it renders fine in Opera. Still, I wonder why it picked such a honking huge replacement before |
| 00:50 | <nessy> | question for those here that understand WebSRT: how can I address a single cue through the CSS selectors? |
| 00:51 | <nessy> | IIUC you can only address all cues or all cue-parts of a certain node type |
| 00:54 | <TabAtkins> | nessy: Right now, the only way to do so is to give it a unique voice and target that. |
| 00:54 | <nessy> | ok, thanks |
| 01:08 | <_bga> | hm |
| 01:08 | <_bga> | MonoTouch - Mono for iOS |
| 02:12 | <tau> | hi |
| 02:12 | <tau> | when i make d = html5lib.parse(f) it simply returns none. |
| 02:13 | <TabAtkins> | I believe the people you'd want to talk to are asleep, since they're european. |
| 02:13 | <tau> | TabAtkins, hm, could you help me then ? |
| 02:14 | <TabAtkins> | Not really. ^_^ html5lib was just written by a couple of the dudes who hang out in this room. |
| 02:14 | <tau> | TabAtkins, where are you from ? |
| 02:14 | <TabAtkins> | west coast of america |
| 02:14 | <tau> | cool |
| 02:18 | <tau> | lolwut |
| 02:19 | <TabAtkins> | Um? |
| 02:31 | <nessy> | TabAtkins: just to confirm about WebSRT - if I want to position a cue 5 lines higher than the bottom rendering area of the video, it would need to get a L: 5 cue setting, right? |
| 02:31 | <nessy> | or is it L: −5 ? |
| 04:33 | <boogyman> | Hixie, AryehGregor is there a tentative date, or general timeframe, when the HTML5 spec is supposed to mature to the next phase? |
| 04:33 | <MikeSmith> | boogyman: what do you mean by next phase? |
| 04:33 | <MikeSmith> | do you mean as far as W3C process, or something else? |
| 04:34 | <boogyman> | currently in working draft, I assume the next is candidate recommendation |
| 04:35 | <MikeSmith> | yeah, that's the next big step |
| 04:35 | <MikeSmith> | but prior to that there's another step |
| 04:35 | <MikeSmith> | which is so-called Last Call |
| 04:35 | <MikeSmith> | which is actually more like First Call |
| 04:35 | <m0> | How long is the Last Call? |
| 04:36 | <boogyman> | Is there a tentative eta for that mike? |
| 04:36 | <MikeSmith> | we have not decided how long the LC for it will be |
| 04:37 | <MikeSmith> | boogyman: schedule from the chairs is here: |
| 04:37 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-html/2010Sep/0074.html |
| 04:37 | <boogyman> | Thanks |
| 04:37 | <MikeSmith> | the date we are aiming to start LC is May 22 |
| 04:38 | <MikeSmith> | "LC resolution carries" = resolution from the group to agree to start LC |
| 04:38 | <MikeSmith> | and what LC at W3C means is that the group must record every comment it receives about the spec during the LC period |
| 04:39 | <MikeSmith> | and at the end of the LC period, publish a report listing the "disposition" for every comment |
| 04:40 | <MikeSmith> | including whether whether any change was made to the spec based on the comment |
| 04:40 | <boogyman> | ok |
| 04:42 | <MikeSmith> | what the HTML WG is doing right now is, the chairs are reviewing all escalated issues that have come in so far |
| 04:42 | <MikeSmith> | where "escalated issues" means, cases where somebody in the group asked for a change to the spec but the change was rejected |
| 04:43 | <MikeSmith> | those are cases that have to be adjudicated by the chairs |
| 04:43 | <MikeSmith> | based on their review of all the available information about the issue |
| 04:44 | <MikeSmith> | and the plan is to try to get all those open issues decided by the chairs, on behalf of the group, by April 22 |
| 04:44 | <MikeSmith> | m0: is some groups, the LC period is just 1 month or so |
| 04:44 | <m0> | Ah cool |
| 04:44 | <MikeSmith> | in this case, it will be significantly longer, though |
| 04:45 | m0 | reading the mailing list |
| 04:45 | <karlcow> | plus theoritically we never exit a phase, but always enter a new phase. Aka we enter in a phase when all the requirements for entering are met |
| 04:45 | <karlcow> | but that is just semantics |
| 04:45 | <karlcow> | :) |
| 04:46 | <MikeSmith> | yeah, the phases all just accumulate |
| 04:46 | <m0> | I am surprised I never joined w3 working group :x Doing that now. |
| 04:46 | <MikeSmith> | like karmic dust |
| 04:49 | <MikeSmith> | anyway, Hixie originally estimated we would go to CR in 2012 |
| 04:49 | <MikeSmith> | and a lot of people scoffed at that as being too far in the future |
| 04:49 | <MikeSmith> | when he said it back in 2006 or early 2007 or so |
| 04:50 | <boogyman> | From prior experience and looking at the development thus far, I'd say that Feb/Mar 2010 sounds about right |
| 04:50 | <karlcow> | (and given the type of specification that people were used to produce aka not a microscopically detailed specification ;) ) |
| 04:50 | <boogyman> | 2012* |
| 04:50 | <karlcow> | different expectations |
| 04:51 | <MikeSmith> | yeah |
| 04:51 | <MikeSmith> | anyway, all the above said, none of these dates really have any effect on end users of the actual technologies |
| 04:51 | <MikeSmith> | or on most adopters of the technologies |
| 04:52 | <MikeSmith> | I mean on content providers or developers |
| 04:52 | <MikeSmith> | what really matters is what has actually been implemented |
| 04:52 | <MikeSmith> | and how interoperably it's been implemented |
| 04:52 | <MikeSmith> | and how stable it is |
| 04:52 | <MikeSmith> | an example is WebSocket |
| 04:53 | <boogyman> | I'm just happy MS is developing IE9 to be compliant with the new spec so developers wont have to waste another decade bickering over the inequities |
| 04:54 | <MikeSmith> | yeah, they have done some great stuff so far, as far as getting important features implemented in IE9 |
| 04:54 | <MikeSmith> | but they have a lot left to do |
| 04:54 | <MikeSmith> | everybody does |
| 04:54 | <MikeSmith> | an good example is HTML5 forms support |
| 04:54 | <MikeSmith> | we still have a long way to go on that |
| 04:55 | <MikeSmith> | as far as getting it implemented |
| 04:55 | <MikeSmith> | and the example of WebSocket is one where we do already have some solid implementations |
| 04:56 | <MikeSmith> | but the spec is not completely stable |
| 04:56 | <MikeSmith> | for the protocol part |
| 04:56 | <boogyman> | I'm intrigued to see what roll web-sockets and <canvas> once decently implemented |
| 04:56 | <MikeSmith> | and implementations are all going to need to change still |
| 04:57 | <MikeSmith> | canvas is probably the most stable new feature from HTML5 that's actually been implemented |
| 04:57 | <boogyman> | I'm not really sold on regular expressions defined in dev source though |
| 04:57 | <MikeSmith> | huh? |
| 04:57 | <MikeSmith> | in what context? |
| 04:58 | <boogyman> | for instance input type=email pattern="RegEx" |
| 05:00 | <MikeSmith> | ah |
| 05:01 | <boogyman> | although, i guess that's more of an implementation bit, as it does have valid use for non-security type instances, such as a name or date |
| 05:02 | <MikeSmith> | yeah, but I supposed naive developers are certainly going to find ways to do stupid things with it |
| 05:02 | <MikeSmith> | but that's true of exposing any kind of feature like that to the masses |
| 05:02 | <jacobolus> | svg is more exciting than canvas IMO. once IE supports it, adoption is going to take off |
| 05:03 | <MikeSmith> | jacobolus: it's not an either-or at all |
| 05:03 | <jacobolus> | sure |
| 05:03 | <jacobolus> | SVG has I think more interesting applications than canvas does |
| 05:03 | <MikeSmith> | the market seems to thing they are both pretty exciting |
| 05:03 | <jacobolus> | definitely |
| 05:04 | <boogyman> | I've not really broken the surface with svg, I was going to do that on my holiday break while traveling |
| 05:04 | <jacobolus> | I've been playing with protovis for a couple weeks. After a somewhat steep learning curve, it's pretty neat |
| 05:04 | <MikeSmith> | what's protovis? |
| 05:04 | <MikeSmith> | authoring tool? |
| 05:04 | <jacobolus> | http://vis.stanford.edu/protovis/ |
| 05:05 | <jacobolus> | infographic/visualization/chart tool by a prof. and a grad student at stanford |
| 05:05 | MikeSmith | looks |
| 05:05 | <jacobolus> | better link: http://vis.stanford.edu/protovis/ex/ |
| 05:05 | <MikeSmith> | nice |
| 05:05 | <karlcow> | svg and html are TSA approved, you can view the source. With canvas it is a bit more opaque ;) |
| 05:06 | <m0> | lol |
| 05:06 | <jacobolus> | karlcow: if you're generating SVG from javascript, it's not always super helpful |
| 05:06 | <m0> | jacobolus: well, you can inspect the page of generated JS and see the SVG |
| 05:07 | <jacobolus> | you can inspect the javascript drawing on the canvas too.... :) |
| 05:07 | <MikeSmith> | I'm not sure most JS programmers would agree that code for a canvas-based image/app is more opaque than that for a SVG one |
| 05:08 | <jacobolus> | but yeah, the era of multiplayer web games is on us, in a way it never was with flash |
| 05:08 | <jacobolus> | (and also multiplayer web other stuff) |
| 05:08 | <jacobolus> | ("multi-user"?) |
| 05:08 | <MikeSmith> | yeah, god help us all |
| 05:09 | <boogyman> | jacobolus: have you seen http://10k.aneventapart.com/Entry/83 |
| 05:09 | <MikeSmith> | great productivity booster that multiplayer games are |
| 05:10 | <karlcow> | "god" in French is a… (no, I'll pass on this one) |
| 05:10 | karlcow | is going to bed |
| 07:06 | <JonathanNeal> | Does anyone here know how I can correct for webkit's underflow? http://sandbox.thewikies.com/float-bugs/ |
| 07:09 | <MikeSmith> | JonathanNeal: is that a problem only in webkit? |
| 07:10 | <MikeSmith> | looks same to me in Minefield |
| 07:10 | <MikeSmith> | ah wait |
| 07:10 | <JonathanNeal> | I don't see it in Firefox 3.6.12 or Firefox 4.0b6 |
| 07:10 | <MikeSmith> | one pixel |
| 07:10 | <MikeSmith> | yeah, I wasn't looking close enough |
| 07:14 | <JonathanNeal> | Is there a css property to adjust for this? |
| 07:48 | <MikeSmith> | https://github.com/ry/node/raw/19478ed4b14263c489e872156ca55ff16a07ebe0/README is great |
| 07:49 | <MikeSmith> | WHEREAS, The usage of threads has complicated computer programming; and WHEREAS, V8 javascript comes free of I/O and threads; and WHEREAS, Most operating systems do not provide asynchonous file system access. Now, therefore: This set server and client libraries were made to build simple but fast servers. They are provided free of charge under a permissive simple license. |
| 07:50 | <MikeSmith> | JonathanNeal: no clue from me about whether there's a CSS property for adjusting that |
| 07:50 | <MikeSmith> | but it seems like a bug |
| 07:50 | <MikeSmith> | that you should report at least |
| 07:50 | <MikeSmith> | (which is realize doesn't help you for now with trying to get it to work the same) |
| 07:51 | <JonathanNeal> | Should I report @ https://bugs.webkit.org/ ? |
| 07:51 | <MikeSmith> | yep |
| 07:51 | <JonathanNeal> | or is it better to do it on the Chrome side? Okay, you got it. |
| 07:52 | MikeSmith | steps away for a bit |
| 07:52 | <MikeSmith> | http://wordsquared.com/ |
| 07:53 | <MikeSmith> | node-based multiplayer word game |
| 07:53 | MikeSmith | now really steps away for a bit |
| 08:02 | <JonathanNeal> | MikeSmith, there we go @ https://bugs.webkit.org/show_bug.cgi?id=50006 hope that helps. |
| 08:03 | <JonathanNeal> | In the meantime I'm trying to think of ways of correcting it with JS, without breaking FF or IE. |
| 08:51 | <phrearch> | hi |
| 08:51 | <phrearch> | does a contenteditable div have the same events as a regular textarea? |
| 08:52 | <phrearch> | im trying to build a wysiwyg editor on top of jinfinote |
| 10:13 | <Hixie> | http://blogs.gnome.org/alexl/2010/11/23/gtk3-vs-html5/ oh dear. |
| 10:13 | <Hixie> | someone's gonna have to explain html to them. |
| 10:24 | <hsivonen> | Hixie: the same person who is going to explain HTML and JavaScript to Google? |
| 10:24 | <hsivonen> | (GWT team in particular) |
| 10:24 | <Hixie> | yeah, probably |
| 10:24 | <Hixie> | or indeed to the bespin team :-) |
| 10:26 | <hsivonen> | WebSocket would be more appropriate for the use case than XHR |
| 10:26 | <Hixie> | i'm far more concerned about the canvas side of it |
| 10:26 | <Hixie> | XHR vs WebSocket doesn't really harm anyone except the programmer |
| 10:27 | <hsivonen> | I think there's a pending "yo, dawg" joke in there when they get a browser running |
| 10:28 | <hsivonen> | Hixie: next, they need to implement an atk to ARIA bridge |
| 10:28 | <Hixie> | we don't provide enough hooks for them to do it |
| 10:28 | <zcorpan> | hsivonen: then they can do gtk in the browser in the gtk in the browser |
| 10:28 | <Hixie> | nor should we, the whole idea is hare-brained :-) |
| 10:29 | <Hixie> | (as implemented) |
| 10:29 | <Hixie> | would make much more sense to have gtk map to html |
| 10:29 | <Hixie> | that would actually be useful |
| 10:29 | <hsivonen> | Hixie: do we provide enough hooks for *that*? |
| 10:30 | <Hixie> | i dunno if we provide enough hooks to get you 100% of the way there |
| 10:30 | hsivonen | notes that VNC sends over bitmaps, not the semantics of the toolkit of the remote system |
| 10:30 | <espadrine> | zcorpan: didn't you see that the firefox copy he was running was actually inside a chrome tab? |
| 10:31 | <Hixie> | but the 80% of the way there that you get will be 100% "native" compared to the vnc-like solution which can get you 100% of the way there in terms of pixels and precisely 0% of the way there in terms of native |
| 10:31 | <espadrine> | (I hope it wasn't) |
| 10:31 | <hsivonen> | the word "accessibility" doesn't appear in the comment section of the blog post |
| 10:31 | <Hixie> | (where native means hooking into system things like IME, accessibility, etc) |
| 10:39 | <hsivonen> | speaking of native, I find it amusing (in a sad way) that the W3C offers an HTML/CSS/SVG reference as an Android app |
| 10:41 | <annevk> | there's also an online version I think |
| 10:41 | <espadrine> | I hope they don't go over the network in the app version, at least... |
| 10:42 | <hsivonen> | annevk: I'm aware of the dogfood version |
| 10:42 | <zcorpan> | where's this reference? |
| 10:43 | <hsivonen> | zcorpan: http://dev.w3.org/2009/cheatsheet/doc/android |
| 10:49 | <zcorpan> | hsivonen: thanks. i don't understand how i get to the "web" version from http://dev.w3.org/2009/cheatsheet/doc/ |
| 10:50 | <espadrine> | http://www.w3.org/2009/cheatsheet/ |
| 10:50 | <zcorpan> | ah |
| 10:53 | <zcorpan> | no <datalist> |
| 10:53 | <Rik`> | http://www.w3.org/2009/cheatsheet/#search,datalist |
| 10:53 | <zcorpan> | i mean, the app doesn't use <datalist> |
| 10:54 | <Rik`> | well, datalist is quite unusable considering the webkit implementation |
| 10:54 | <annevk> | does use a non-polyglot DOCTYPE and a namespace declaration |
| 10:54 | <annevk> | hurray |
| 10:57 | <jgraham> | Hmm, I think the GTK-in-HTML thing is rather neat. I mean, doing it with HTML widgets would be better but would probably have the disadvantage of not actually working |
| 10:58 | <jgraham> | Obviously a11y is an issue |
| 12:03 | <annevk> | "After briefly trying out Debian and Fedora, I suspect Ubuntu might be the worst distro choice, except for all the others." :/ |
| 12:12 | <jgraham> | I thought that was common wisdom |
| 12:31 | <annevk> | this MS extensions to window.screen thread is completely useless |
| 12:32 | <annevk> | aah, found the start of the thread |
| 12:33 | <annevk> | aah okay |
| 12:34 | <annevk> | it's about APIs nobody wants to add |
| 12:34 | annevk | moves along |
| 12:34 | annevk | deletes thread |
| 20:07 | <AryehGregor> | Could someone tell me if these look the same or different in IE9 beta? |
| 20:07 | <AryehGregor> | data:text/html,<!doctype html><img src='data:image/svg+xml,<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN" "http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd"><svg xmlns="http://www.w3.org/2000/svg"; width="200" height="200"><circle r="100" fill="red" /></svg>' style="height: 5em"> |
| 20:07 | <AryehGregor> | data:text/html,<!doctype html><img src='data:image/svg+xml,<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN" "http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd"><svg xmlns="http://www.w3.org/2000/svg"; width="200" height="200" viewBox="0 0 200 200"><circle r="100" fill="red" /></svg>' style="height: 5em"> |
| 20:08 | <AryehGregor> | WebKit and Opera render them the same, Firefox 4.0b7 renders them differently (clipping the former instead of scaling it). |
| 20:08 | <AryehGregor> | I can't figure out from the SVG spec which is correct. It doesn't say what you do if viewBox isn't defined. |
| 20:10 | <AryehGregor> | I have no Vista or 7 machine that I can install IE9 on, since my father needs IE8 for work (sigh). |
| 20:12 | <AryehGregor> | CSS height/width scales everything else, so I don't see why it wouldn't scale SVG. |
| 20:13 | <AryehGregor> | Does CSS say clearly whether replaced content should be scaled vs. clipped? |
| 20:13 | <AryehGregor> | (when given width/height) |
| 20:14 | <AryehGregor> | I now hate all specs other than HTML5. They're much too vague. :( |
| 20:14 | <AryehGregor> | (well, and some other post-HTML5 specs) |
| 20:14 | <Ms2ger> | AryehGregor++ |
| 20:14 | <AryehGregor> | (and some non-web specs, and possibly some web specs before HTML5 that I haven't heard of) |
| 20:22 | <david_carlisle> | AryehGregor: IE9 doesn't seem to acept data: in that context and if save the contentto local .html files and then load them, they each just render as a broken image symbol |
| 20:22 | <AryehGregor> | Well, it could be rewritten as a proper document. |
| 20:22 | <AryehGregor> | One sec. |
| 20:23 | <david_carlisle> | yes thats what I meant i saved from <!doctype to the end (and checked ift worked in FF) |
| 20:24 | <AryehGregor> | http://aryeh.name/tmp/test1.html http://aryeh.name/tmp/test2.html |
| 20:24 | <AryehGregor> | How do those look? |
| 20:24 | <AryehGregor> | Whee, Chrome messes up the second one in a weird way. |
| 20:24 | <TabAtkins> | Whether to scale or not is up to SVG. CSS asks it for its intrinsic dimensions, then figures out a box for it to render in and just tells it to go crazy. |
| 20:24 | <AryehGregor> | (Opera doesn't handle inline SVG at all in the version I'm using, which is why I used data URLs) |
| 20:25 | <TabAtkins> | (http://dev.w3.org/csswg/css3-images/#sizing) |
| 20:26 | <david_carlisle> | 1 quater circle 2 half |
| 20:26 | <david_carlisle> | sorry need to go... |
| 20:27 | <AryehGregor> | SVG 1.1 is what I should be looking at, right? |
| 20:27 | <TabAtkins> | ...probably. I think that's the one everyone implements, yeah. |
| 20:28 | <AryehGregor> | TabAtkins, typo for you to fix there: "conain". |
| 20:28 | <TabAtkins> | Yeah, there's a few typos in that section that I'm just now seeing. |
| 20:34 | <TabAtkins> | Argh, I can't get a cubic-bezier timing function to work in animation. Hrm. |
| 20:37 | <TabAtkins> | Ah, there it is. It's very annoying trying to figure out exactly what combination of things need to be prefixed in a complex experimental syntax. |
| 20:40 | <AryehGregor> | Is there any chance we can get browsers to start ignoring doctype and xmlns for SVG? |
| 20:40 | <AryehGregor> | Because man, those are obnoxious. |
| 20:40 | <TabAtkins> | That's the primary reason I don't use SVG in practice, definitely. |
| 20:42 | <AryehGregor> | So can we persuade the SVGWG about this, or work around them, or what? Has it come up there? |
| 20:43 | <TabAtkins> | I doubt SVG will stop being an XML language, so probably a no-go in SVG itself. |
| 20:43 | <AryehGregor> | You don't need doctype or xmlns to be an XML language. |
| 20:43 | <AryehGregor> | XML doesn't require either of those. |
| 20:44 | <TabAtkins> | Oh, yeah, you're right. Damn things just sprout them as a matter of course, so I forget. |
| 20:44 | <TabAtkins> | Um, then maybe! |
| 20:44 | <roc> | just write your SVG in HTML |
| 20:44 | <Ms2ger> | It works in Firefox! :) |
| 20:45 | <roc> | and Chrome, I think |
| 20:45 | <AryehGregor> | So just do <img src="foo.html"> where foo.html is like <!doctype html><svg>...</svg>? Does that work? |
| 20:45 | <roc> | mmmm no |
| 20:45 | <roc> | <object> would though |
| 20:45 | <roc> | so would <iframe> |
| 20:45 | <roc> | I suppose we could make HTML images work, although someone might blow a gasket |
| 20:46 | <TabAtkins> | Well, just doing the svg-in-html certainly works in Chrome just fine. Yay!@ |
| 20:46 | <AryehGregor> | Do height and width work as expected on object and iframe? |
| 20:46 | <david_carlisle> | AryehGregor: don't you want a mime type for "svg serialised as html" |
| 20:46 | <Ms2ger> | image/html |
| 20:46 | <AryehGregor> | Ms2ger++ |
| 20:46 | <TabAtkins> | AryehGregor: No, they work differently than they do for <img>. |
| 20:46 | <roc> | it should work in IE9 as well |
| 20:47 | Ms2ger | is reminded of image/xhtml+xml |
| 20:47 | <AryehGregor> | I vaguely remember trying out object/iframe for SVG on my site and got annoyed at how they worked differently, so I used img. |
| 20:49 | <AryehGregor> | It looks like Firefox lets me leave out doctype, but not xmlns. |
| 20:50 | <Ms2ger> | In XML? That's what I'd expect |
| 20:50 | <AryehGregor> | Ditto Chrome and Opera. |
| 20:50 | <AryehGregor> | Oh well. |
| 20:51 | <TabAtkins> | Huh? Chrome lets me leave out xmlns in html. |
| 20:51 | <AryehGregor> | In <img>. |
| 20:51 | <AryehGregor> | In text/html you're supposed to be able to leave out xmlns, yeah. |
| 20:53 | <TabAtkins> | Ah, darn, you're right. I can't use svg-in-html-in-img. |
| 20:53 | <TabAtkins> | ;_; |
| 20:54 | <jgraham> | Do the SVG doctypes map to the magic set of entities? |
| 20:54 | <jgraham> | Like they do for MathML? |
| 20:55 | <jgraham> | Not that you need them for SVG, typically |
| 20:55 | <Ms2ger> | I don't think so |
| 20:56 | <jgraham> | So you can use them in MathML, XHTML, HTML, but not SVG? Weird |
| 20:56 | <jgraham> | consistency++ |
| 21:03 | <TabAtkins> | Ah, now I remember what the problem I had with Firefox's SVG-in-<img> support - it doesn't do SMIL while inside an <img>. I tried to change over to <object> but it failed pretty bad, so I just decided that FF doesn't get animated conveyor belts. |
| 21:08 | <david_carlisle> | so don't you want an image/svg+html mime type for svg serialised as per html5 to use with image? |
| 21:09 | <david_carlisle> | It only took 12 years to (not) register the mime type for svg as xml... |
| 21:10 | <david_carlisle> | jgraham, you can use whatever entities you line in svg as xml if you use an appropriate dtd |
| 21:10 | <david_carlisle> | 'cept that in a browser context you don't fetch the dtd... |
| 21:16 | <webr3> | anybody know if xhr.statusText should be '200 OK' or just 'OK'? (as in with code or without) |
| 21:24 | <jgraham> | david_carlisle: Yeah, you could use an internal subset but it's about as appealing as hacking your own arm off. With a spoon. |
| 21:26 | <david_carlisle> | or the browsers could all decide to parse all xml with an effective catalogue that predefined all the entities always rather than just looking for certain old legacy doctypes and adding the mathml ones, sometimes. |
| 21:36 | <TabAtkins> | david_carlisle: That would be nice, yes. Then maybe Anne could drop doctype handling from his XML5 stuff. |
| 21:37 | <TabAtkins> | (Right now, he only parses doctypes for entities. If he didn't need that, he could drop it down to a handful of states just describing how to ignore a doctype.) |
| 21:38 | <david_carlisle> | sIn an environment where you don't fetch external dtd and don't validate, |
| 21:38 | <david_carlisle> | just always using the same set of entities makes sense to me |
| 21:38 | <TabAtkins> | Me too. |
| 21:39 | <david_carlisle> | (and means i didn't waste half a lifetime maintaining them through changes in Unicode:-) |
| 21:41 | <webr3> | IIRC all the us gov data (new) all uses doctypes w/ entities &myns; etc |
| 21:41 | <webr3> | several thousand data sets |
| 21:41 | <david_carlisle> | rdf? |
| 21:42 | <webr3> | yeah everything via RPI |
| 21:45 | <david_carlisle> | webr3: do you have an example uri? Also for xml, ignoring any external subset and just using a dtd that defines the htmlmathml entities is (just about) conformant xml, but you'd also have to define entities defined in a local subset (in xml, Anne's xml5 can do whatever he says:-0 |
| 21:47 | <webr3> | david, http://tw2.tw.rpi.edu/data-gov/data-10048.rdf |
| 21:48 | <david_carlisle> | thanks |
| 21:50 | <david_carlisle> | webr3: that's got a lot of namespaces but no doctype (and thus) no entities to worry about. |
| 21:53 | <webr3> | hmm maybe i gave wrong link (should have checked) - I may be confused, will confirm in a mo |
| 21:53 | <webr3> | ack wrong link |
| 21:56 | <webr3> | ahh yes, http://data-gov.tw.rpi.edu/vocab.php?instance=Dataset_616 (and all data sets_ |
| 21:57 | <webr3> | aside: it caught my attention yesterday as it was a huge amount of datasets in this style, and it's not often I see doctypes and entities in RDF/XML - until this set that, there is one other big oen the same, can't remember which off hand |
| 22:02 | <david_carlisle> | webr3: thanks yes rdf gives xml a bad name:-) |
| 22:03 | <david_carlisle> | as I say it's conformant (and would work with that) to ignore any external dtd subset and substitute the html5 entities, and then to define any entities in the local subset |
| 22:05 | <david_carlisle> | it would be interesting to know what's behind that php script whether it could serve the rdf/xml without the local subset and all the namespaces uri inlined |
| 22:06 | <webr3> | probably a good time to ask, as they're migrating to different server's atm |
| 22:07 | <david_carlisle> | well it isn't my government so someone else should probably ask:-) |
| 23:31 | AryehGregor | is slightly disoriented when he reads random cryptography articles on Wikipedia and finds references regularly to things named after the guy who taught his crypto course |
| 23:31 | <TabAtkins> | Heh, that's pretty cool. |
| 23:32 | <gsnedders> | AryehGregor: Maybe he did know what he was talking about, after all, then :P |
| 23:34 | <AryehGregor> | I guess that's what you get at a place like NYU, they have top researchers. He was a really good instructor, too. |
| 23:34 | <AryehGregor> | I haven't noticed it with my math professors, but I guess math research is just way too specialized for me to know anything that touches on their research. |
| 23:38 | AryehGregor | finds http://en.wikipedia.org/wiki/Very_smooth_hash suspicious, since it claims to be collision-resistant but not preimage-resistant, which is self-evidently impossible |
| 23:40 | <TabAtkins> | My cryptography knowledge is very obviously based on me reading wikipedia and blogs, so I slip up and confused collision and preimage a lot. >_< |
| 23:44 | AryehGregor | notices that the article only cites two sources, of which one is the original paper and the other link is broken . . . |
| 23:45 | <gsnedders> | I just claim to not know much about crypto :) |
| 23:51 | <AryehGregor> | Okay, so apparently collision resistance implies that you can't do preimage attacks on *most* hashes, but you might still be able to do them on particular important subsets of hashes. |
| 23:53 | <AryehGregor> | So formally it implies preimage resistance, if you define preimage resistance in terms of the challenger selecting a hash uniformly at random, but not in practice. |
| 23:53 | <AryehGregor> | Learn something new every day. |
| 23:55 | <AryehGregor> | I don't feel so bad, because I learned this particular fact from this famous professor I was talking about. |
| 23:55 | AryehGregor | stops talking to himself |