| 00:00 | <Philip`> | caden: Is there some browser in which it does clear the bitmap? |
| 00:00 | <caden> | it turned out i didn't put style position relative in the element |
| 00:00 | <caden> | i was doing some code from this sample drawing program http://dev.opera.com/articles/view/html5-canvas-painting/ and it was always drawing the line about 200 pixesl too low |
| 00:01 | <caden> | i'm using iceweasel-- it's firefox 3.0 compliant, not 3.5 yet |
| 00:01 | <caden> | (the debian firefox) |
| 00:02 | <caden> | i tried using setTransform and the identity matrix to see if there was a transpose setting from all the previous draws i did |
| 00:02 | <caden> | but that wasn't it |
| 00:02 | <caden> | but it did clear the graph i had drawn on the canvas |
| 03:19 | <Darxus> | Ack, the link to explain the "this is a 'should' and not a 'must' because it has proven to be impossible" doesn't explain it! |
| 03:26 | <Darxus> | inimino: I agree that Philip`'s answer was a valid answer to my question, and that I did not ask the right question. However, that does not change the fact that Philip`'s answer was not useful. |
| 03:31 | <inimino> | Darxus: it might have been useful to you, had it caused you to reconsider the question |
| 03:31 | <inimino> | (Usefulness is subjective.) |
| 03:59 | <MikeSmith> | are there any major browsers that actually put processing instructions from text/html content into the DOM? |
| 03:59 | <MikeSmith> | are processing instructiong in the HTML DOM even defined? |
| 05:06 | <MikeSmith> | ȯ̭̥͈̭ͪa̲͡f̣̝̜̦̳̼̬͗͊̅ͤ |
| 05:16 | <MikeSmith> | COMBINING SEAGULL BELOW |
| 05:16 | <MikeSmith> | one of the best Japanese noise bands |
| 09:51 | <hallvors> | why doesn't the spec allow multi-letter accesskey values? |
| 09:54 | <Philip`> | hallvors: "For future extensibility, I have also required tokens with more than one character to be ignored for now. We can later add identifiers for various purposes, e.g. accesskey="help" to mean that the user agent's default "help" key should be used, whatever that is." |
| 09:54 | <Philip`> | (http://www.mail-archive.com/whatwg⊙lwo/msg14673.html) |
| 10:01 | <hallvors> | Thanks, Philip`. Rules out using @accesskey for GMail-type "gi" shortcuts, unfortunately. Is that a stronger or weaker use case? Or what about <a accesskey="F1"> - should we go there? |
| 10:02 | <Philip`> | Might want to send feedback to the list - everyone's probably asleep in here :-) |
| 10:07 | <hallvors> | sure. tried to ask the same question on IRC two days ago but I guess that was at the wrong time too :-p |
| 10:07 | <Lachy> | people are usually here at this time of day |
| 10:07 | <Lachy> | maybe they just see you and run |
| 10:07 | <Lachy> | :-) |
| 10:08 | <hallvors> | :-) |
| 10:08 | <hallvors> | perhaps I'll try if public-html is usable for less political issues then.. (though it wouldn't surprise me if someone turned it into a political one) |
| 10:08 | hallvors | is a scary chatter, apparently |
| 10:08 | <Lachy> | access keys have an impact on accesibility, and so it is very likely to turn into one |
| 10:09 | <hallvors> | does that mean I ought to ask on the whatwg list instead? ;) |
| 10:10 | <Lachy> | no, Chaals prefers that we use public-html |
| 10:10 | hallvors | thinks politics is a waste of time |
| 10:12 | <Philip`> | You can always post to public-html and then ignore all replies |
| 10:18 | <Lachy> | why doesn't anyone trim the CC list on that One Vendor, One Veto thread? |
| 10:19 | <gsnedders|work> | Lachy: Because pressing Reply All is simpler |
| 10:21 | <hallvors> | Philip`: I can't make sense of the following quote from the E-mail you linked to: "I have allowed the <command> element to be given an access key, which |
| 10:21 | <hallvors> | allows any scriptable behaviour to be given a keyboard shortcut." |
| 10:21 | <hallvors> | I don't find that in the spec |
| 10:22 | <hallvors> | it looks like the "command" concept is strictly tied to elements (only exposed through element.command properties etc) |
| 10:23 | <hallvors> | ..and I also don't see how the spec makes it possible to do this: "Once a command is defined, other parts of the interface can refer to the same command" |
| 10:23 | <hallvors> | http://www.whatwg.org/specs/web-apps/current-work/multipage/interactive-elements.html#concept-command |
| 10:23 | <hallvors> | I'm probably missing something really basic about the command concept.. |
| 10:26 | <Philip`> | hallvors: http://www.whatwg.org/specs/web-apps/current-work/multipage/interactive-elements.html#the-command ? |
| 10:27 | <hallvors> | that helps! but I still don't see any text about @accesskey being allowed for <command> |
| 10:27 | <Philip`> | Hixie: That section says 'the checkbox" maps to the Checkbox' and is missing a quote and probably a word, though I'm too lazy to send mail about this |
| 10:27 | <Philip`> | hallvors: Maybe it's in the global attributes? (I would look myself but Opera crashes as soon as I clicked the 'global attributes' link :-p ) |
| 10:27 | <hallvors> | ah, "Global attributes" |
| 10:28 | <hallvors> | found it now, yes. |
| 10:28 | <Philip`> | s/crashes/crashed/ |
| 10:34 | <hallvors> | I'm assuming that the way to link <command> to an actual "scripted behaviour" is through normal event listeners (<command onclick=""> or command.addEventListener('click', ..) ) but I still don't see how other elements can "refer to" a defined command. Say, if I do <menu><command id="c" onclick="foo()"></menu> and want a <button> elsewhere to represent that command, how do I link the <button> element to the <command>? |
| 10:37 | <Philip`> | hallvors: Why do you want to link a <button> to the <command>? |
| 10:37 | <Philip`> | It seems like the idea is that <command>s just get put into context menus or toolbar menus, and you activate them from there |
| 10:38 | <Philip`> | though it also seems you can do <button onclick=document.getElementById('c').click()> if you want to activate it remotely |
| 10:40 | <hallvors> | I thought that was sort of the point of the spec saying "Once a command is defined, other parts of the interface can refer to the same command". |
| 10:41 | <hallvors> | and it would be a nice convenience if doing document.getElementById('c').setAttribute('disabled', 'disabled') would disable all "associated" <button>/<input> elements. |
| 10:41 | <hallvors> | for example |
| 10:42 | <hallvors> | guess I'm just misunderstanding the "other parts of the interface" sentence |
| 10:42 | <Philip`> | Oh, I didn't know the spec said that |
| 10:43 | <Philip`> | I haven't actually read it |
| 10:43 | <hallvors> | it's in the first blurb on http://www.whatwg.org/specs/web-apps/current-work/multipage/interactive-elements.html#concept-command |
| 10:43 | <Philip`> | In fact I didn't even know this whole command thing existed, until twenty minutes ago |
| 10:43 | <hallvors> | it's a new thingy ;) |
| 10:43 | <Philip`> | so I should defer to someone who knows what they're talking about :-) |
| 10:44 | <hallvors> | you've been very helpful though :) |
| 11:09 | <Darxus> | inimino: It did cause me to recognize the flaw in my question. However, the intent of my original question was sufficiently obvious that there was no use in recognizing the flaw or correcting it. |
| 11:12 | Philip` | doesn't understand the intent of the original question |
| 11:14 | <Philip`> | since I don't know how much flexibility you would want the schema language to have, or how much it could be tailored specifically for HTML5 |
| 11:38 | <gsnedders|work> | eighty4: yt? |
| 11:40 | annevk4 | wonders if it's worth reviewing the RDFa draft given that it contains stuff like "While deprecation of the xmlns: attribute is attempted in HTML5" |
| 11:55 | <Lachy> | annevk4, it's worth reading to see what they're proposing, but I haven't yet decided if it's worth responding with any feedback since it so clearly ignores all prior feedback about not relying on namespaces |
| 11:58 | <MikeSmith> | Philip`: I notice spec-splitter.py doesn't report filenames of all sections it's generating files for |
| 11:59 | <MikeSmith> | would be nice if it did |
| 11:59 | MikeSmith | goes to read the code |
| 12:00 | <MikeSmith> | d'oh |
| 12:00 | <MikeSmith> | my bad |
| 12:00 | MikeSmith | hits himself with the idiot stick |
| 12:00 | <MikeSmith> | bad me |
| 12:09 | annevk4 | gets to the formal objection thread |
| 12:15 | <annevk4> | a Chris Wilson summarizes in that thread how I feel about it "The fun conversations always happen when I'm on vacation." |
| 12:15 | <annevk4> | s/a/ah/ |
| 12:24 | <eighty4> | gsnedders|work: yes |
| 12:27 | <Philip`> | MikeSmith: Um, okay |
| 12:28 | <Philip`> | MikeSmith: Does that mean I don't have to understand or respond to your comment? :-) |
| 12:34 | <MikeSmith> | Philip`: pretend I nver said it |
| 12:41 | <takkaria> | annevk4: what thread? |
| 12:42 | <annevk4> | on public-html? |
| 12:42 | <annevk4> | and www-archive |
| 12:44 | <takkaria> | ah |
| 12:52 | <Mrmil> | What's more appropriate: <section> as main content area with news as <article>'s, or vice versa? |
| 12:54 | <MikeSmith> | Mrmil: I would think vice versa |
| 12:55 | <annevk4> | you probably want just <article> in that scenario |
| 12:55 | <Mrmil> | But what if the main content area contains: some text, shopping assistant, product catalog and list of categories? Can still that be in an article? |
| 12:56 | <adactio> | annevk4: do you mean an <article> with multiple <article>s nested in it? |
| 12:56 | <annevk4> | there's no real wrapper element for the main content area |
| 12:56 | <Mrmil> | ok |
| 12:56 | <annevk4> | some have suggested there should be one though |
| 12:56 | <annevk4> | i.e. <main> or <content> or some such |
| 12:56 | <adactio> | Well, we have <div>. |
| 12:57 | <Mrmil> | Could you look here? http://www.whatwg.org/demos/company-home/ I wonder if it's correct :) |
| 13:05 | <annevk4> | "<q>Development</q>" is not an actual quote |
| 13:05 | <annevk4> | I'm not really sure about all the details of <section> / <article> etc. myself so I won't comment on that :) |
| 13:07 | <MikeSmith> | complete the following list: |
| 13:07 | <Mrmil> | annevk4: Hehe, ok. I saw a tutorial today showing it the other way. |
| 13:07 | <MikeSmith> | text/html parsing algorithm, Window object, innerHTML, document.write |
| 13:07 | <MikeSmith> | no answer is wrong |
| 13:07 | <MikeSmith> | unless the answer is stupid |
| 13:07 | <adactio> | Mrmil: Like many of these semantic questions, it's all a matter of opinion really, isn't it? Personally I might choose to use a <div> where you've used <section id="content"> or I might choose <footer> where you've used <aside> but I don't think there are any black and white answers. |
| 13:07 | <annevk4> | batman |
| 13:08 | <annevk4> | MikeSmith, aah |
| 13:08 | <MikeSmith> | heh |
| 13:08 | <takkaria> | <base href=""> changes |
| 13:08 | <Mrmil> | adactio: I kindof agree, it's hard to decide what markup to use |
| 13:09 | eighty4 | pokes at gsnedders|work with a djungelvrål |
| 13:09 | <adactio> | Mrmil: we demand rigidly defined areas of doubt and uncertainty. ;-) |
| 13:09 | <annevk4> | adactio, arguably that makes the new elements less useful |
| 13:09 | <MikeSmith> | takkaria: cheers |
| 13:10 | <adactio> | annevk4: not at all, it follows in the tradition of HTML: flexibility with semantic choices empowers the author. |
| 13:10 | <Mrmil> | adactio: hehehe :P |
| 13:10 | <annevk4> | MikeSmith, you're looking for things it helped clarify? |
| 13:10 | MikeSmith | smiles at "rigidly defined areas of doubt and uncertainty" |
| 13:10 | <adactio> | MikeSmith: that was Douglas Adams, I believe. :-) |
| 13:11 | <MikeSmith> | adactio: aha |
| 13:11 | <MikeSmith> | I should have guessed |
| 13:11 | <MikeSmith> | either that or Mark Pilgrim |
| 13:12 | <MikeSmith> | annevk4: yeah, biggish stuff that's been around in browsers since 1975 or whenever but that was never spec'ed at all til now |
| 13:16 | <annevk4> | window.history / window.location / XMLHttpRequest / <base target>, <a href="x x">, <form> rendering in quirks mode, etc. |
| 13:16 | <annevk4> | activation of nested <a> elements |
| 13:16 | <annevk4> | contenteditable / execCommand() and friends |
| 13:18 | <takkaria> | quirks mode / standards mode generally, really |
| 13:18 | <MikeSmith> | annevk4, takkaria - thanks, adding to my list |
| 13:19 | <annevk4> | takkaria, yeah, I suppose <form> rendering in quirks mode is pretty obscure, but it was one of the things that caused us compat issues :) |
| 13:24 | <zcorpan> | MikeSmith: Error: Heading cannot be a child of another heading. |
| 13:24 | <zcorpan> | From line 189, column 3; to line 189, column 19 |
| 13:24 | <zcorpan> | >↩<h1>↩↩ <h2 id="editing"><span |
| 15:31 | <annevk4> | quite fascinating how blogs make this XHTML 2.0 / HTML5 thing a soap opera |
| 15:32 | <karlcow> | annevk4: drama queen :) |
| 15:32 | <gsnedders|work> | Well, if it's a soap opera, I guess I can ignore it. |
| 15:32 | karlcow | still thinks that Lachy as a drag queen on a bus singing would be cool |
| 15:32 | <Philip`> | Better than a soap firefox |
| 15:32 | <Philip`> | It wouldn't melt so fast |
| 15:33 | karlcow | is referring to http://www.dvdverdict.com/images/reviewpics/priscilla1.jpg |
| 15:33 | <annevk4> | karlcow, I feel less and less moved by the ill-informed comments :) |
| 15:34 | <karlcow> | hehe |
| 15:34 | <karlcow> | annevk4: I guess it means that html5 reached a community a bit farther than w3c and whatwg. Somehow… finally the real world. |
| 16:01 | <Darxus> | Somebody named something "grand unified schema" and it's only useful for biological stuff :( |
| 16:02 | <ezyang> | LDAP! |
| 16:04 | <MikeSmith> | Philip`: thanks. working now |
| 16:06 | <Philip`> | MikeSmith: Neato |
| 16:10 | gsnedders|work | waves |
| 16:10 | <gsnedders|work> | Anyone have a clue about running reftests in Fx? |
| 16:10 | <gsnedders|work> | I'm getting stuff that looks like failures and no output. Something seems wrong. |
| 16:11 | <Philip`> | MikeSmith: Now all that's missing is the equivalent of http://www.whatwg.org/specs/web-apps/current-work/multipage/404.html#the-draggable-attribute :-) |
| 16:11 | <Philip`> | (though preferably an equivalent that actually works, instead of having a temporarily broken fragment-links.js) |
| 16:12 | <Philip`> | MikeSmith: (The theory is that that's useful when pages get renamed and a link to an old page gets 404ed) |
| 16:18 | <MikeSmith> | Philip`: ok, i can add that too |
| 16:25 | <mookid> | Hello internet friends |
| 16:26 | <mookid> | Philip`: do you have 5 mins to spare me? |
| 16:26 | <mookid> | :> |
| 16:30 | <Philip`> | mookid: Probably not, unless it's about something interesting |
| 16:31 | <mookid> | potentially |
| 16:31 | <mookid> | it's about how browsers treat links and state transitions |
| 16:32 | <mookid> | Are state transitions in HTTP simply the URI/Verb combination, or the entire of the message? |
| 16:33 | <mookid> | by message I mean HTTP message headers basically |
| 16:35 | <Philip`> | mookid: I have no idea |
| 16:36 | <mookid> | Does the browser store page history just with the URI and HTTP method? |
| 16:37 | <gsnedders|work> | It depends on the browser; most store cache headers as well, and some title for the page |
| 16:38 | <mookid> | is it feasible they could store the entire request? (minus sensitive headers like Authorization) |
| 16:38 | <gsnedders|work> | No idea |
| 16:38 | <gsnedders|work> | I don't know what they store myself, I'm just guessing from my experience of the black box |
| 16:39 | <mookid> | fair enough - where might I get a bit of feedback on that? |
| 16:39 | <mookid> | is there a browser dev channel of some sort? |
| 16:40 | <Darxus> | XHTML 1.0 does not allow text directly in a body element, but HTML5 does?? |
| 16:40 | <gsnedders|work> | XHTML 1.0 Transitional does, Strict does not. |
| 16:40 | <gsnedders|work> | But yes, HTML 5 does. |
| 16:41 | <mookid> | \o/ |
| 16:41 | <gsnedders|work> | mookid: Being around here at the right time is probably your best bet |
| 16:41 | <gsnedders|work> | (Don't ask me when right time is though) |
| 16:41 | <Philip`> | mookid: You could look at e.g. Firefox's places.sqlite, which seems to be where it stores the page history information |
| 16:41 | <Philip`> | particularly in the table that's like: CREATE TABLE moz_places (id INTEGER PRIMARY KEY, url LONGVARCHAR, title LONGVARCHAR, rev_host LONGVARCHAR, visit_count INTEGER DEFAULT 0, hidden INTEGER DEFAULT 0 NOT NULL, typed INTEGER DEFAULT 0 NOT NULL, favicon_id INTEGER, frecency INTEGER DEFAULT -1 NOT NULL); |
| 16:41 | <Philip`> | (frecency? Hmm...) |
| 16:42 | <mookid> | :P |
| 16:42 | <mookid> | bloody hippies |
| 16:42 | <Philip`> | (That's a totally made up word, but apparently someone put it in Wiktionary) |
| 16:42 | <mookid> | ok gsnedders|work thanks for the heads up |
| 16:42 | <mookid> | I read back over a little bit of the stuff in November |
| 16:43 | <mookid> | and I think I got a little hot under the collar because I got accused of being an architecture astronaught or something |
| 16:43 | <gsnedders|work> | Philip`: frecency, like netwerk |
| 16:43 | <Darxus> | gsnedders|work: I am absolutely shocked that HTML5 allows that when XHTML 1.0 strict does not :P However, I never got around to trying to figure out why XHTML 1.0 Strict didn't. |
| 16:44 | <gsnedders|work> | Philip`: And referer |
| 16:44 | <Darxus> | Awesome, HTML5 validator catches overlapping table cells, and XHTML 1.0 Strict doesn't: http://www.chaosreigns.com/code/dxschema/tests/invalidtable.html |
| 16:44 | <Darxus> | It's also pretty cool how chrome renders it. |
| 16:44 | <Darxus> | And firefox. |
| 16:45 | <mookid> | anyway I'm going to try and formulate a more complete case for adding conneg attributes |
| 16:46 | <mookid> | not that I expect it to be taken seriously but just so I don't look like I was causing a fuss for no reason :) |
| 16:48 | <Darxus> | Yeah I've been futilly agravating the whatwg list for a couple days trying to convince them it could be useful to have a standard way to mention the target HTML specification for the document :/ |
| 16:48 | <mookid> | to be fair I think there's quite a lot of us asking for stuff |
| 16:49 | <mookid> | so hate me |
| 16:49 | <mookid> | because I hate you |
| 16:49 | <Darxus> | Haha. |
| 16:49 | <mookid> | ? |
| 16:49 | <mookid> | seriously. |
| 16:49 | <mookid> | :D |
| 16:49 | <Darxus> | And I'm not allowed to be amused? |
| 16:50 | <mookid> | was joking |
| 16:50 | <mookid> | although we can have an argument about it if you want |
| 16:50 | <mookid> | since we're both into that kind of thing |
| 16:50 | <Darxus> | Haha. |
| 16:50 | <mookid> | ? |
| 16:50 | <Darxus> | Have you seen my posts? |
| 16:50 | <mookid> | no |
| 16:50 | <Darxus> | Or are you just basing that belief on what I've said here? |
| 16:51 | <Darxus> | Okay. |
| 16:51 | <Darxus> | Working on it. |
| 16:51 | <gsnedders|work> | Darxus: There have been several hundred emails on that subject already, FWIW |
| 16:53 | <Darxus> | I want a way, specified in HTML5, to state which HTML version I expect a document to conform to. |
| 16:54 | <Philip`> | Darxus: It seems your posts started with a solution (adding a language version identifier to documents) and then tried to describe a problem for it, whereas you might have more success with defining a practical problem and then suggesting and comparing possible solutions (of which a language version identifier might be one, with various pros and cons) |
| 16:54 | <Darxus> | I would be fine with something like either <meta name="html version" content="5"> or <body data-html-version=5>. |
| 16:54 | <mookid> | mmm |
| 16:54 | <mookid> | use an HTTP header |
| 16:54 | <mookid> | -_- |
| 16:55 | <mookid> | or extend the MIME type |
| 16:55 | <mookid> | what is the MIME type for html5? |
| 16:55 | <mookid> | is it just text/html >? |
| 16:56 | <Darxus> | Philip`: I totally agree. My last case was where someone has thousands of HTML5 conformant files and is gradually converting them to HTML5.1. And wants to be able to spider validate his entire site, while ensuring that the old files are still valid HTML5, and the updated files are still valid HTML 5.1. |
| 16:56 | <Philip`> | Yes, text/html |
| 16:56 | <Darxus> | text/html *or* application/xhtml+xml. |
| 16:56 | <Darxus> | Or I think application/xml. |
| 16:56 | <mookid> | don't use anything other than text/html |
| 16:56 | <mookid> | it breaks stuff. |
| 16:56 | <Darxus> | You don't get the full awesomeness of the browser using an xml parser unless you serve it as an xml content type. |
| 16:56 | <annevk4> | Darxus, the idea is that we do not make incompatible changes unless we really need to |
| 16:57 | <annevk4> | Darxus, and if we really need to authors better update their pages too because the feature they're using is harmful in some way |
| 16:57 | <Darxus> | mookid: Grr. Serve text/html to anyone with "MSIE" in their User-Agent, and application/xhtml+xml to everyone else. With the same document. |
| 16:57 | <gsnedders|work> | Darxus: HTML 5 cannot be served as application/xhtml+xml; XHTML 5 can. |
| 16:57 | <Philip`> | Darxus: That just punishes non-IE users, by increasing the chance of them seeing a fatal error instead of your page |
| 16:57 | <gsnedders|work> | Darxus: Have fun if you accept any user input with XML formats. |
| 16:57 | <mookid> | Darxus: yeah I know that - but be normal and just use text/html -_- |
| 16:58 | <annevk4> | Darxus, besides what Philip` said, XML parsers are also less optimized than HTML parsers |
| 16:58 | <Darxus> | gsnedders|work: I have not yet fully grasped the differences between the xml and html types. |
| 16:58 | <Darxus> | annevk4: Less optimized? They seem really freaking fast to me. |
| 16:58 | <mookid> | current html isn't valid XML |
| 16:58 | <annevk4> | Darxus, i.e. not all XML parsers have fast paths for known elements to create the appropriate DOM constructs right away |
| 16:58 | <Darxus> | They have the advantage of knowing they are getting a perfectly formed document. |
| 16:58 | <annevk4> | Darxus, in browsers |
| 16:58 | <annevk4> | Darxus, that's not really an advantage |
| 16:59 | <annevk4> | Darxus, have you written parsers? |
| 16:59 | annevk4 | has |
| 16:59 | <mookid> | sex god |
| 16:59 | <Darxus> | mookid: I looked into specifying the HTML version in the Content-Type, it looks entirely possible, as: Content-Type: text/html; v=5 |
| 16:59 | <mookid> | right |
| 16:59 | <Darxus> | annevk4: I have not yet written parsers. |
| 16:59 | <mookid> | what's wrong with that? |
| 16:59 | <gsnedders|work> | Darxus: Knowing something is perfectly formed makes no difference. |
| 16:59 | <Darxus> | What's wrong with what? |
| 17:00 | <gsnedders|work> | mookid, Darxus: What do you gain by it? |
| 17:00 | <mookid> | nothing I'm just giving him a solution |
| 17:00 | <annevk4> | XML also requires more complicated checks on element names and such that require some extra time |
| 17:00 | <mookid> | dont' use html anyway use JSON it works better |
| 17:00 | <Darxus> | mookid: The thing that's wrong with specifying the html version number in the Content-Type is that I want to specify it in the document, so I don't have to keep seperate records are which files are to which spec. |
| 17:01 | <mookid> | you're storing your html in flat file? |
| 17:01 | <mookid> | with no metadata? |
| 17:01 | gsnedders|work | finds a test case that says, "This page tests something" |
| 17:01 | <Darxus> | mookid: Yes. |
| 17:01 | <mookid> | ewww |
| 17:01 | <Darxus> | Heh. |
| 17:01 | <Philip`> | Storing HTML in files? How antiquated! |
| 17:02 | <mookid> | how's 1994 these days? |
| 17:02 | <Philip`> | You need to be more enterprisey |
| 17:02 | <Darxus> | Maybe I'm wrong about XML vs. HTML, but I thought XML was better. I'll try to remember to look into it. |
| 17:02 | <gsnedders|work> | You need to store everything in a database! |
| 17:02 | <Darxus> | This is my personal web site, it's almost all static html :) |
| 17:02 | <Darxus> | Heh. |
| 17:02 | <gsnedders|work> | You know, just store everything in a single column table with the datatype of BLOB. |
| 17:02 | <mookid> | I said no metadata :/ |
| 17:02 | <gsnedders|work> | Because databases are cool. |
| 17:03 | <gsnedders|work> | mookid: I know, I'm just being silly about the whole file thing. |
| 17:03 | <mookid> | :P |
| 17:03 | <Darxus> | So what reason is there to not include a method of specifying the HTML version in the HTML5 spec? |
| 17:03 | <mookid> | RDBMS's are totally unnecessary in the majority of cases |
| 17:03 | <mookid> | it's just lower barrier to entry |
| 17:03 | <gsnedders|work> | mookid: Seriously though, most FS can store little to no metadata |
| 17:03 | <mookid> | apache can do clever thins with metadata though right? |
| 17:04 | <gsnedders|work> | It stores metadata in seperate config files though |
| 17:04 | <gsnedders|work> | Or serializes it into the filename |
| 17:04 | <mookid> | yeah and rules |
| 17:04 | <mookid> | well that's flat files for ya |
| 17:04 | <mookid> | �-_- |
| 17:05 | <mookid> | just store the metadata, work out how your rules are going to work, and tack on the version to the content-type header |
| 17:06 | <mookid> | isn't it really inefficient to parse XML instead of just looking at a header? |
| 17:06 | gsnedders|work | is quite tempted to go and get some food |
| 17:07 | <gsnedders|work> | mookid: Parsing MIME types is expensive, though. |
| 17:07 | <mookid> | is it? |
| 17:07 | <mookid> | o.O |
| 17:07 | <gsnedders|work> | What metadata do we want? |
| 17:07 | <gsnedders|work> | If it's something in the prolog, you can just stop parsing XML after that. |
| 17:08 | <gsnedders|work> | mookid: Yes, the format of MIME types is quite complex. |
| 17:08 | <mookid> | meh, you standards people suck. |
| 17:08 | <gsnedders|work> | mookid: Why? |
| 17:08 | <mookid> | that's a joke |
| 17:08 | <mookid> | your brain is much bigger than mine |
| 17:08 | <mookid> | relax |
| 17:08 | <mookid> | :P |
| 17:08 | gsnedders|work | takes a deep breath |
| 17:08 | <gsnedders|work> | :P |
| 17:08 | <mookid> | ;D |
| 17:09 | <mookid> | still waiting for the right time to ask that question again |
| 17:09 | <gsnedders|work> | Hmm, if I go home now, then I need to work an extra hour and half over for the next two days |
| 17:09 | <Darxus> | Even better: http://www.chaosreigns.com/code/dxschema/tests/invalidtable.html |
| 17:10 | <gsnedders|work> | mookid: The conneg one? Ask it on the mailing list if you want a more diverse set of opinions, and want the proposal to go anywhere |
| 17:10 | <mookid> | can't you just write a script to auto-respond to emails with 'no its not getting included there's no use case' ? |
| 17:10 | <gsnedders|work> | Hmm, I'm not sure whether NLP is up to that :P |
| 17:10 | <mookid> | yeah, it's tricky stuff |
| 17:10 | <mookid> | -_- |
| 17:14 | <annevk4> | mookid, also, conneg is pretty much dead |
| 17:16 | <Darxus> | (invalidly) Overlapping cells with 50% alpha background colors: http://www.chaosreigns.com/code/dxschema/tests/invalidtable.html |
| 17:21 | <mookid> | annevk4: that's because browsers and HTML don't support it - not because there's anything wrong with it |
| 17:21 | <mookid> | it's actually a really good thing for stuff like cache-invalidation mechanisms |
| 17:23 | <Darxus> | The HTML5 spec is a bit verbose. |
| 17:25 | <Philip`> | Verbose, or detailed? |
| 17:26 | <Philip`> | It says a lot but I think a lot of what it says is necessary |
| 17:27 | <Darxus> | Yeah I don't object. |
| 17:27 | <gsnedders|work> | Hixie: http://wiki.whatwg.org/wiki/Developing_HTML5_derivatives is no longer true (where have header-whatwg and header-w3c gone?) |
| 17:47 | <Darxus> | "This specification should be read like all other specifications. First, it should be read cover-to-cover, multiple times. Then, it should be read backwards at least once." - HTML5 |
| 17:51 | <Darxus> | Ack, attribute values are not all required to be quoted? |
| 17:52 | <Philip`> | They aren't |
| 17:52 | <Philip`> | <input type=checkbox checked> etc |
| 17:53 | <tantek> | Philip' is that for both the HTML and XHTML serializations? |
| 17:53 | <Lachy> | tantek, no |
| 17:54 | <tantek> | Lachy, good, that's what I thought. |
| 17:54 | <Philip`> | XHTML is just XML and doesn't change any of XML's rules |
| 17:54 | <Philip`> | so there shouldn't be anything to worry about there :-) |
| 17:56 | <Darxus> | Ah, good. |
| 17:59 | <Darxus> | Why wasn't CSS originally implimented in sgml, and why isn't it now conformant to xml similarly to HTML5/XML? |
| 18:02 | <Philip`> | Because if the syntax was SGML/XML, it would be incredibly painful to write |
| 18:03 | <ttepass-> | DSSSL! |
| 18:12 | <Darxus> | <css> <definition> <target> <element>div</element> <class>container</class> </target> <attribute> <top>0</top> <left>0</left> <width>100%</width> </attribute> </definition> </css> |
| 18:12 | <Darxus> | Philip`: Point taken. |
| 18:13 | <Darxus> | Or I guess <attribute top="0" left="0" width="100%">. |
| 18:14 | <Philip`> | <style selector="div.foo" top="0" left="0" width="100%"/> |
| 18:14 | <Darxus> | Wow, thanks. I have much to learn. |
| 18:15 | <Philip`> | if you don't mind using a custom non-XML language for selectors |
| 18:15 | <Philip`> | There's nothing to learn, I just made that up :-) |
| 18:15 | <Philip`> | Don't know how to extend it to handle things like !important |
| 18:15 | <Darxus> | I disagree. It is a much shorter valid xml way of representing the same thiing I typed. |
| 18:16 | <Darxus> | And I want to write a schema language :P |
| 18:17 | <eighty4> | <target><element>div></element>.....<imporance><level>1000</level></importance>....</target>? |
| 18:20 | <ttepass-> | Just a thought: A XML-Version of CSS in Philip's style had a great use of CURIES: |
| 18:20 | <ttepass-> | <style xmlns:svg="http://www.w3.org/TR/SVG"; selector="[svg:path] .... /> |
| 18:34 | <tantek> | Darxuz - see http://bytes.com/groups/css/99746-why-css-not-xml |
| 18:37 | <tantek> | and while looking for that, I found: |
| 18:37 | <tantek> | http://www.w3.org/MarkUp/2004/xhtml-faq |
| 18:38 | <tantek> | I'm pretty sure I read this long ago in the past, but I thought folks here might find it insightful from a development of HTML/XHTML standpoint. |
| 18:39 | <tantek> | even better answer to the "Why is CSS not XML syntax" FAQ: |
| 18:39 | <tantek> | "CSS has its own syntax because readability is very important for a language that is used by nearly as many people as HTML itself. It could have been based on SGML, and originally there were indeed proposals for SGML-based syntaxes (see [Lie99]) and there even existed a US military standard ([FOSI]), but there was never any question that CSS syntax is better. (XML didn't exist at the time, but it wouldn't have helped since it is more verbose than SGML and al |
| 18:39 | <tantek> | from http://www.w3.org/People/Bos/DesignGuide/readability.html |
| 18:39 | <tantek> | by Bert Bos, one of the creators of CSS |
| 18:40 | <inimino> | well put |
| 18:46 | <Darxus> | Cool. |
| 19:17 | <Darxus> | I was trying to write a schema language without having ever looked at one. I was trying to decide one / zeroorone / zeroormore / oneormore, and one / optional / (default) / required, and then looked up relax ng and discovered it uses a mix. |
| 19:29 | <Darxus> | Hah, there are constraints on W3C Schema that cannot be expressed in it? |
| 20:29 | <sebmarkbage> | How long does it typically take for a moderator to review a message to the list? My first post hasn't gone through yet. Anxious. :) |
| 20:30 | <annevk4> | which list and are you subscribed to it? |
| 20:33 | <sebmarkbage> | annevk4: whatwg⊙wo and yes |
| 20:34 | <annevk4> | there should be no moderation |
| 20:34 | <annevk4> | if you got a moderation message it means you are not subscribed |
| 20:35 | <sebmarkbage> | that's weird... I recieve the e-mails... |
| 20:35 | <annevk4> | are you sending from the same address as you're receiving? just checking... |
| 20:36 | <sebmarkbage> | yes. |
| 20:36 | <sebmarkbage> | If I send, I get a "Your message to whatwg awaits moderator approval" by whatwg-bounces⊙lwo |
| 20:40 | <sebmarkbage> | let me just check which e-mail gets the list again just to make sure... |
| 20:40 | <Philip`> | sebmarkbage: The moderators take infinitely long |
| 20:40 | <Darxus> | I found a typo! "A content attribute is said to change value only if its value new value is different" - 2.1.2, DOM trees |
| 20:41 | <sebmarkbage> | Yea, I had subscribed using a different e-mail. typo. thanx. |
| 20:42 | <Philip`> | Darxus: http://www.mail-archive.com/whatwg⊙lwo/msg16078.html - old news :-p |
| 20:44 | <Darxus> | Damn. |
| 20:48 | <Darxus> | How about "it is possible in this encodings" s/this/these/ ? |
| 20:48 | <Darxus> | 2.1.5 Character encodings |
| 20:56 | <annevk4> | sebmarkbage, seems to work now |
| 20:56 | <sebmarkbage> | yes. Thanks annevk4 |
| 20:59 | <Philip`> | Darxus: See same mail as the other typo :-) |
| 21:00 | <gsnedders> | The UK is losing 52 pubs per week! Oh noes! |
| 21:00 | Philip` | suggests looking behind the sofa |
| 21:00 | <Philip`> | Wait, that's not quite the right concept |
| 21:01 | Philip` | is mixing up lost coins and Doctor Who watching habits |
| 21:01 | gsnedders | squints |
| 23:26 | <caden> | is there a way to save the current canvas bitmap and then retrieve it, like to reset the background of an animation? |
| 23:27 | <caden> | getImageData, probably? |
| 23:27 | <Philip`> | caden: You could create a new <canvas> (with createElement) and then drawImage the current one into it, then draw it back later |
| 23:27 | <Philip`> | or use getImageData/putImageData |
| 23:28 | <Philip`> | or use toDataURL/drawImage (except that probably will throw security errors in most browsers if you do it multiple times) |
| 23:28 | <Philip`> | (Copying to a new canvas is probably going to be more efficient than getImageData, if you care about that) |
| 23:30 | <caden> | can i make that new canvas invisible then? |
| 23:31 | <caden> | right now redrawing the graph between animation frames is the application bottleneck |
| 23:31 | <caden> | i am hoping i can just refresh this snapshot instead of redrawing all the axes and lines and whatnot |
| 23:32 | <Philip`> | Yes - if you do something like "var c = document.createElement('canvas'); c.width = ...; c.height = ...; c.getContext('2d').drawImage(othercanvas, 0, 0)" then the new canvas never gets rendered since it's not added to the document |
| 23:35 | <caden> | i will try it, the get/put did not improve performance |
| 23:42 | <caden> | does this look right to ou for restoring the image? this.snapshotCanvas.getContext('2d').drawImage(this.canvasContext,0,0); |
| 23:42 | <caden> | or is this.canvasContext supposed to be a refernce to an element instead of a context |
| 23:45 | <Philip`> | caden: The argument to drawImage is the canvas object that you want to copy from |
| 23:45 | <Philip`> | rather than a context |
| 23:45 | <caden> | ok |
| 23:45 | <caden> | thank you! |
| 23:45 | <caden> | it looks like it stores the background as transparent, does that sound right? |
| 23:46 | <Philip`> | Probably - if you're not explicitly drawing over the canvas's background then it'll just be transparent |
| 23:46 | <caden> | cool |
| 23:56 | <caden> | well, i'm using a very primitive way of profiling, using just getTime() statements, and these indicate that drawImage is much faster than get/put |
| 23:57 | <caden> | although subjectively it doesn't seem faster |
| 23:57 | <caden> | maybe it is just my sucky machine |
