| 00:00 | <gsnedders> | This endless requirement of CSS implementers is for it to be applying while iterating in tree order? |
| 00:22 | <MikeSmith> | Hixie: I see you're using tircd |
| 00:28 | <takkaria> | Hixie: what are you actually doing atm? |
| 00:30 | <Hixie> | MikeSmith: it roxors |
| 00:31 | <Hixie> | takkaria: going through and providing non-normative descriptions for all the apis aimed at authors rather than implementors |
| 00:32 | <Hixie> | takkaria: look for the green <dl>s e.g. at http://www.whatwg.org/specs/web-apps/current-work/#images which i just did |
| 00:33 | <Hixie> | if you're using a decent browser you can try one of the experimental alternative style sheets, too |
| 00:37 | <MikeSmith> | Hixie: I been using tircd for a while too.. definitely rocks the bells |
| 00:38 | heycam | is just trying it, seems good |
| 00:39 | <heycam> | i was just using the web interface before. polling is for suckers... :/ |
| 00:41 | <Hixie> | tircd makes twitter usable for me |
| 00:46 | <gavin> | oh hey |
| 00:47 | <gavin> | I hacked a mozbot to broadcast tweets onto IRC using the RSS feed |
| 00:47 | <gavin> | that looks much better |
| 00:48 | <Hixie> | tircd is way better than a bot because it makes each user have its own user in your ircclient |
| 00:48 | <Hixie> | it's like bitlbee |
| 00:49 | <gavin> | and it's two-way, right? |
| 00:49 | <Hixie> | with tircd you can follow new people with /invite, find out who people are with /whois, unfollow with /kick, block with /ban, etc |
| 00:49 | <Hixie> | yeah |
| 00:49 | <Hixie> | if you talk it tweets your message |
| 00:49 | <Hixie> | it's the hotness |
| 00:51 | <tantek_> | Hixie, I presume this is the tircd you speak of? http://code.google.com/p/tircd/ |
| 00:55 | <Hixie> | yeah |
| 00:58 | <annevk> | timbl joins the RDFa debate |
| 00:58 | <annevk> | funs |
| 01:07 | <Hixie> | ok who wants to be the one to tell tim about the bazillions of xmlns:* attributes in legacy content |
| 01:16 | <Hixie> | 40% done |
| 01:22 | <annevk> | oh lol, I had not read timbl's message yet and just see the part mjs quoted |
| 01:22 | <annevk> | "I can't imagine that there are a lot of people who have accidentally used the string xmlns: inside attribute names in the past. :-)" |
| 01:23 | <annevk> | -- timbl |
| 02:06 | <Hixie> | oh good, mjs answered that one |
| 02:06 | <Hixie> | i was worried i might have to |
| 02:20 | <roc> | We should get tbl to be a Mozilla intern for the summer |
| 02:21 | <Hixie> | good luck with that |
| 02:24 | <MikeSmith> | I don't blame anybody for not wanting to work on a browser project |
| 02:24 | <MikeSmith> | most sane people don't want to work in the sausage factory |
| 02:27 | <Hixie> | the problem is the people who want to design the sausage |
| 02:27 | <Hixie> | but haven't worked in the factory |
| 02:27 | <Hixie> | (or not worked in it this century) |
| 02:30 | <MikeSmith> | new guy in the sausage factory: "hey! spec says you aren't supposed to take that disgusting part of the big/cow/rat-that-got-caught-in-the-chopper and make it into food!" .. guy who's been working in the sausage factory: "Heh. That's what you been eating, friend. For a long time." |
| 02:31 | <Hixie> | "In fact if we stopped using that part... your digestive system would break down and you'd only be able to eat peanuts for the rest of your life." |
| 02:33 | annevk | raises an eyebrow |
| 02:34 | <Hixie> | read the AB minutes and raise more! :-) |
| 02:34 | <roc> | how many have you got |
| 02:34 | <Hixie> | ...several. |
| 02:40 | <karlcow> | hmm sausages… yum! Just had a bun bo hue, but would not say no to a sausage |
| 02:41 | <Hixie> | gah i hate the table dom model |
| 02:41 | <Hixie> | does anyone ever even use this nonsense? |
| 02:42 | <heycam> | karlcow, is vietnamese food popular/widely available in france? and is that because of it being a former colony? |
| 02:43 | karlcow | is waiting for the day of people who wants a spec which is forkable and when the next EvilEmpire Inc comes around with a super marketing force creates a spec which is imposed to the mass of people. Think about Intel Inside, Think MS Word, etc. |
| 02:44 | <karlcow> | heycam: Vietnamese food is popular in France indeed, because of the 100 years of colony of this asian region. (but I'm not in France) |
| 02:44 | <deltab> | Hixie: I've done so, though I've been told the implementations are very slow |
| 02:44 | <heycam> | karlcow, oh right you're in canada, forgot |
| 02:44 | Hixie | writes half-hearted authoring descriptions for the six thousand apis the table elements have |
| 02:44 | <Hixie> | deltab: curse you! :-P |
| 02:45 | <deltab> | it does seem odd that tables get special attention in the DOM |
| 02:45 | <karlcow> | :) no issue. I moved recently from Japan to Canada. :) |
| 02:45 | heycam | loves him some cà phê sữa đá |
| 02:46 | <MikeSmith> | claims of threat of spec forking is total FUD |
| 02:46 | <heycam> | (with funky diacritical characters copied from somewhere, since i have no idea how to type them) |
| 02:46 | <MikeSmith> | hmm, guess I shouldn't say that so forcefully |
| 02:46 | <MikeSmith> | um, some people might say that claims of threat of spec forking is total FUD |
| 02:47 | <Hixie> | given html5's history, claims that copyright has any ability to stop spec forking are ridiculous |
| 02:47 | MikeSmith | caught up in excitement of CSS WG f2f and got carried away |
| 02:47 | <Hixie> | and given html5's present (and its two forks), claims that spec forking are a problem are also ridiculous imho |
| 02:48 | <Hixie> | but i understand that a lot of people are scared |
| 02:48 | <karlcow> | hehe |
| 02:48 | <Hixie> | on another note, why are html5's forks only written by robs |
| 02:49 | <Hixie> | and is roc going to make one too? :-P |
| 02:49 | <heycam> | ha |
| 02:49 | <MikeSmith> | we need some rocspec |
| 02:49 | <karlcow> | I love maciej's message :) 1) and 2) did nothing so let's change the license to achieve the same nothing. huh… hihi That is logical. poooof |
| 02:50 | <karlcow> | a big circus of clowns |
| 02:50 | <Hixie> | oh sheesh, <td> has a whole swathe of apis too |
| 02:50 | Hixie | cries |
| 02:51 | <Hixie> | the table elements are actually proving more of a pain than <canvas> and <video>! |
| 02:51 | <Hixie> | er, <tr>, not <td> |
| 02:52 | <roc> | I think spec forking could be a problem but copyright threats are not remotely the right solution |
| 02:52 | <Hixie> | food time |
| 02:53 | <annevk> | karlcow? |
| 02:53 | <roc> | *actual* forks that is. The *ability* to fork is essential, *actual* forks are harmful |
| 02:53 | <annevk> | we already have "actual" forks in a way |
| 02:55 | <karlcow> | annevk: my sincere and honest opinion about that is the license is just fine as it has always been. And all the comments around it are FUD. I do not believe in W3C but I trust W3C and its social process. To create a license for prose which protects it and puts it in the hand of a group is a good thing. |
| 02:56 | <karlcow> | I still have no trust in whatwg social structure because of the way in history they have acted. |
| 02:56 | <karlcow> | I understand perfectly that people can disagree with me and that's fine. |
| 02:58 | <karlcow> | And my trust in the social structure of whatwg can't be improved as long that the whatwg is used as a gun on my head saying if you don't behave as we wish, we (opera, mozilla, apple) will do our own way. That is not acceptable in a social process. |
| 02:58 | <annevk> | you believe the use cases put forward are FUD? |
| 02:59 | <karlcow> | annevk: here… you just put words in my mouth I have not said. Classical! |
| 02:59 | <annevk> | you said that all the comments around the license argument are FUD |
| 02:59 | <karlcow> | yes. |
| 02:59 | <annevk> | use cases are comments around the license argument |
| 02:59 | <annevk> | we want it changed because of the use cases |
| 03:00 | <karlcow> | the use cases are hypothetical cases which never happened. |
| 03:00 | <annevk> | it may be that I don't know what you're saying though; admittedly I often find that difficult |
| 03:00 | <annevk> | karlcow, that's not true |
| 03:00 | <roc> | Henri already cited a case where he was warned not to put spec text in his code |
| 03:00 | <annevk> | karlcow, e.g. hsivonen uses text from the WHATWG draft in his code |
| 03:00 | <karlcow> | for a group of people who say all the time that we live by "real world" dogma |
| 03:00 | <karlcow> | I find it quite amusing |
| 03:01 | <annevk> | karlcow, software copies IDL fragments from drafts |
| 03:01 | <karlcow> | roc: henri cited a case which never happened really and to completion |
| 03:01 | <gavin> | that these things are tolerated but not allowed by the license isn't a very strong argument |
| 03:01 | <karlcow> | someone at w3c asked him to do something and someone reverted his/her decision |
| 03:01 | <annevk> | you still haven't explained how I'm putting words in your mouth |
| 03:01 | <roc> | what's 'completion'? Someone is successfully sued for copyright infringement? |
| 03:02 | karlcow | took a knife this morning, I could have kill the squirel outside, but I didn't do it. Damn! |
| 03:02 | <karlcow> | roc: yes |
| 03:03 | <roc> | so if the W3C sued someone and lost, you'd say there's still no problem? |
| 03:03 | <karlcow> | if W3C lost the case that would prove that W3C was not in its rights |
| 03:03 | <roc> | to be more precise |
| 03:03 | <karlcow> | Justice system |
| 03:05 | <karlcow> | annevk: FUD the comments about the license. The use cases were exactly to try to have a practical discussion around the issue. Test cases if you prefer. |
| 03:05 | <roc> | So you're saying that what Henri was doing was not within the letter of the license, but it should be tolerated, and the W3C as a whole will tolerate it even though we know not all W3C people agree |
| 03:05 | <karlcow> | roc: no. I didn't say that |
| 03:06 | <roc> | oh |
| 03:06 | <karlcow> | IMHO what henri was doing was perfectly fine. |
| 03:06 | <annevk> | karlcow, the comments evolve around the use cases |
| 03:06 | <annevk> | karlcow, I'm confused |
| 03:07 | <karlcow> | annevk: it's ok. :) I'm confused with all this discussion, because for me I se No. zill. nada. issues with the W3C License as it is. |
| 03:08 | <karlcow> | and I love opensource and creative commons licenses, but they all have their particular context. |
| 03:08 | <annevk> | some people apparently have an issue with the license |
| 03:08 | <karlcow> | annevk: I see that. |
| 03:09 | <annevk> | at some point people didn't have an issue with Member-only groups either |
| 03:09 | <annevk> | things change |
| 03:09 | <karlcow> | some people have an issue with the existence of the whatwg ;) |
| 03:09 | <roc> | karlcow: you don't think our desire to be able to incorporate spec text in our open source projects is reasonable? |
| 03:09 | <annevk> | karlcow, prolly the same people that are fine with the license :p |
| 03:09 | <karlcow> | roc: it is perfectly reasonnable AND possible today |
| 03:09 | <gavin> | it's always been technically possible |
| 03:10 | <gavin> | but the license does not allow it |
| 03:10 | <roc> | you mean via fair use and other copyright defenses? |
| 03:10 | <karlcow> | roc: yes |
| 03:10 | <karlcow> | never the W3C will sue open source projects |
| 03:10 | <roc> | New Zealand does not have a fair use doctrine, where does that leave me? |
| 03:10 | <karlcow> | that is dumb |
| 03:10 | <roc> | maybe so, but where does that leave me? |
| 03:10 | <karlcow> | but if NextEvelEmpire Inc |
| 03:10 | <tokamak> | SLEEP |
| 03:11 | <karlcow> | does bad things, then there will be room for actions |
| 03:11 | <gavin> | copyright law is not the right tool for that |
| 03:12 | <karlcow> | roc: for me the fear of the new requested license is that it prepares the bed for the NextLevelEmpire Inc. |
| 03:12 | <roc> | copyrighted specs did not slow down Microsoft one iota |
| 03:12 | <roc> | and they will not slow down the next attacker, if it comes |
| 03:13 | <karlcow> | slow down on what? |
| 03:13 | <karlcow> | I'm talking about a technical spec |
| 03:13 | <roc> | developing their own APIs, publishing them, implementing them, evangelizing them |
| 03:13 | <karlcow> | not implementation |
| 03:13 | <gavin> | what are you afraid that NextEvilEmpire will do? |
| 03:13 | <karlcow> | heeeey? |
| 03:14 | <gavin> | if they gain enough credibility at W3C's expense, maybe they deserve it! |
| 03:14 | <karlcow> | [21:48] * karlcow is waiting for the day of people who wants a spec which is forkable and when the next EvilEmpire Inc comes around with a super marketing force creates a spec which is imposed to the mass of people. Think about Intel Inside, Think MS Word, etc. |
| 03:14 | <roc> | that's not really a concrete fear |
| 03:15 | <annevk> | unless you believe in the vast browser wing conspiracy o_O |
| 03:15 | <gavin> | you're talking about "imposing" a spec - they can do that either way, whether they're allowed to copy the original spec text or not |
| 03:16 | <karlcow> | roc: I like open source projects and community projets, I like community social structures giving protection to a body of work with a kind of equilibrium of forces. I do NOT like at all Marketing Department of huge companies. The issue is that many people *believe*. |
| 03:17 | <roc> | I don't understand how marketing departments are relevant here. If you're talking about marketing to the masses, the copyright status of HTML is completely irrelevant to the masses |
| 03:17 | <karlcow> | roc: marketing is what is relevant specifically. |
| 03:17 | <karlcow> | and where the evil comes from most of the time. |
| 03:18 | <roc> | and for marketing, trademarks are what you want |
| 03:18 | <karlcow> | roc: would you accept that W3C trademark html5 ? |
| 03:18 | <roc> | by all means trademark "HTML" or whatever you can get your hands on |
| 03:18 | <roc> | "W3C HTML: The Real Web" |
| 03:18 | <roc> | karlcow: yes |
| 03:18 | <karlcow> | … |
| 03:18 | karlcow | is lost |
| 03:19 | <roc> | we trademark Firefox |
| 03:19 | <roc> | that's how we protect our brand from evil people who try to sell malware as "Firefox" etc |
| 03:20 | <roc> | no-one is impeded from taking our code and doing what they like with it, they just can't call it "firefox" necessarily |
| 03:20 | <roc> | everybody's happy, except for some misguided Debian users |
| 03:20 | <karlcow> | And some people said that Mozilla was evil because of the trademark. |
| 03:21 | <roc> | they're in the minority, I believe |
| 03:21 | <karlcow> | w3c license case: every's happy, except for some misguided html wg participants |
| 03:21 | <roc> | they probably haven't realized that there actually are people out there offering "Firefox downloads" containing malware |
| 03:21 | <karlcow> | :) i can do sentences too ;) |
| 03:22 | <roc> | karlcow: everybody's happy except for three out of four browser vendors and other people developing open-source software related to these specs |
| 03:23 | <roc> | that's a pretty big part of the spec audience |
| 03:23 | <roc> | even the Debian users are pretty happy, they've got their Iceweasel build without the Firefox trademark |
| 03:23 | <karlcow> | roc: for a few people into the browser communities |
| 03:24 | <roc> | I daresay Maciej can speak for Webkit |
| 03:24 | <karlcow> | I think I hear recently that for example for the Mozilla's case, people didn't like when someone was saying "I talk for Mozilla" |
| 03:24 | <roc> | and I doubt anyone at Mozilla would disagree with Jonas or dbaron's position |
| 03:24 | <roc> | we don't all agree about everything but we do tend to agree pretty strongly about our open source licensing |
| 03:25 | <karlcow> | roc: I have nothing against open source code. |
| 03:25 | <karlcow> | quite the opposite |
| 03:26 | <roc> | nevertheless you prefer to put us at legal risk for incorporating spec text in our code |
| 03:27 | <karlcow> | heh http://sam.zoy.org/wtfpl/ |
| 03:27 | <karlcow> | roc: there is no risk. |
| 03:29 | <roc> | We've already been over that. Henri was threatened. There is risk. |
| 03:30 | <karlcow> | and it's where I come with the argument fud. So I guess we will not find a solution today :) |
| 03:32 | <roc> | if you think the people who have spoken out against the license are just "a few of the browser communities", please try to find one person from the Mozilla or Webkit community to speak out in support of the license |
| 03:32 | <roc> | just one |
| 03:33 | <karlcow> | Let's say that you rely on a very rigid and binary interpretation of law. |
| 03:33 | <karlcow> | and I rely on social structures and their organic natures, I rely on humans and need trust for it. |
| 03:33 | <karlcow> | Having been closed of quite a number of participants of this story, my trust is still on the side of the social structure of W3C. |
| 03:33 | <roc> | no, make it two |
| 03:33 | <karlcow> | roc: what does it take to be part of Moz community? |
| 03:34 | <karlcow> | employed by Moz Corp? or anyone participating? |
| 03:34 | <roc> | for the sake of argument, I'll say "commit access to the repository" |
| 03:34 | <roc> | which includes hundreds of non-MoCo people |
| 03:34 | <karlcow> | yep |
| 03:34 | <roc> | that's the funny thing. We in the open source communities advocate using social structures, not copyright law, to prevent forking |
| 03:35 | <roc> | you think that the threat of lawsuit hanging over everyone's head is the way to get people to behave |
| 03:35 | <karlcow> | you still didn't get my opinion, did you? |
| 03:35 | <karlcow> | there is no threat |
| 03:36 | <roc> | if we can have that in writing, then the problem is solved |
| 03:37 | <roc> | although you'd need to extend your amnesty to all downstream recipients of our code, or else we won't be able to distribute our own code under GPL/LGPL |
| 03:37 | <karlcow> | the funny thing is that if the license changes I don't care. If I look carefully in my W3C staff mail archives, I'm even pretty sure I have been advocating at least one or twice for creative commons licenses when I was working at W3C. |
| 03:38 | <roc> | great, then why resist it? |
| 03:38 | <karlcow> | roc: it is not *my* amnesty |
| 03:38 | <roc> | ok, if it's the W3C's policy they can write it down in a binding manner |
| 03:39 | <karlcow> | roc: isn't it what was in the proposal just made by W3C? |
| 03:45 | <roc> | the latest proposal imposes conditions that make it impossible to distribute derived works of parts of the spec with our code |
| 03:45 | <roc> | where's the amnesty? |
| 03:46 | <karlcow> | roc's rippling |
| 03:49 | karlcow | is throwing the towel. I tried to explain… but I'm not rigid enough be a wall. License might change, and the world will continue. The only thing I know is that it will still orbit around the sun. |
| 03:51 | <karlcow> | A breed whose testimony is behaviour, |
| 03:51 | <karlcow> | What we are, we are—nativity is answer enough |
| 03:51 | <karlcow> | to objections; |
| 04:19 | <annevk> | bit annoying that workers have race conditions through cookies |
| 04:19 | <annevk> | hopefully on idiot site starts relying on specifics of execution |
| 04:20 | <annevk> | s/on i/no i/ |
| 05:25 | <annevk> | hmm, I guess I should put http://wiki.whatwg.org/wiki/RelExtensions in order :/ |
| 05:27 | <annevk> | maybe it's very easy to write a script for that |
| 05:30 | <annevk> | I wonder why some values are listed there that are already defined in the specification itself |
| 05:40 | <annevk> | was indeed very easy |
| 05:40 | <annevk> | list = rawdata.split("\n|-\n") |
| 05:40 | <annevk> | list.sort() |
| 05:40 | <annevk> | print "\n|-\n".join(list) |
| 05:40 | <annevk> | well, I also lowercased EditURI first |
| 06:07 | <annevk> | I removed the values already defined in the spec |
| 06:07 | <annevk> | is there some way to grey out a line in wiki tables to mark it obsolete? |
| 08:21 | <hsivonen> | does anyone have a reference to the CSS hack the tests the namespace of xmlns="..." on root to determine whether a document is in the HTML or XML mode? |
| 09:55 | <hsivonen> | streaming XPath impl: http://code.google.com/p/jlibs/wiki/XMLDog |
| 10:27 | <Lachy> | I updated the list of obsolete elements. http://dev.w3.org/html5/html-author/#obsolete-elements |
| 10:27 | <Lachy> | let me know if I missed any. That's all the elements I found mentioned in the spec |
| 10:28 | <Lachy> | the list does seem a little short though |
| 10:30 | Philip` | wonders why datagrid isn't a link, in the table just above there |
| 10:31 | <Philip`> | Lachy: Are you aware that the DOM interfaces are (except for their first line) indented by two tabs too many? |
| 10:32 | <Lachy> | Philip`, because the spec stuck a note saying "The API here needs rewriting. Don't implement this right now." between the heading and the <dl class="element">, and so my script to extract it from the spec didn't find it |
| 10:32 | <Philip`> | Ah |
| 10:32 | <Lachy> | whereas, I have a separately maintined datafile from which I generate those tables, which includes datagrid |
| 10:33 | <Lachy> | http://dev.w3.org/html5/html-author/utils/elementdesc.txt |
| 10:34 | <Lachy> | the DOM interfaces are extracted as-is from the spec. They will be replaced with something a lot more author friendly later on |
| 10:34 | <Lachy> | so I'm not too concerned about the indentation now |
| 10:35 | <Philip`> | Also, tabs are aesthetically displeasing and you should indent with spaces instead ;-) |
| 10:35 | <Lachy> | I suppose I could set my editor to use soft tabs instead |
| 10:36 | <Lachy> | which means 4 spaces |
| 12:41 | <Lachy> | Added description of the html element http://dev.w3.org/html5/html-author/#the-html-element |
| 14:16 | <Philip`> | Lachy: Perhaps it could do with a few more pixels of left margin on the heading text in those green boxes |
| 14:17 | <Lachy> | ok |
| 14:17 | <Philip`> | By the way, I hope you're not wanting any substantial comments :-p |
| 14:17 | Philip` | tries actually reading the text |
| 14:18 | <Lachy> | Philip`, of course not. Don't even think about provinding any constructive feedback |
| 14:18 | <Lachy> | *providing |
| 14:19 | <Philip`> | s/langauge/language/ |
| 14:20 | <Philip`> | "Every document must begin with this element" - that seems to be mixing the concepts of elements and tags - the document as a DOM tree doesn't have a 'beginning', and the document as a character stream has tags rather than elements |
| 14:20 | <Philip`> | (Or if it's not mixing elements and tags, it's mixing trees and character streams) |
| 14:21 | <Lachy> | The character stream needs to begin with the DOCTYPE (possibly preceded by whitespace or comments) |
| 14:22 | <Lachy> | how do you suggest I rephrase it? |
| 14:23 | <Philip`> | Maybe something like "The html tag must be the first tag in the document" would be more accurate? (with a subsequent paragraph clarifying that that really means "The html tag must, if not omitted entirely, be the first tag in the document") |
| 14:24 | <Philip`> | I don't know if that's just more confusing, though |
| 14:25 | <Philip`> | The discussion of languages should link to something describing what languages are, e.g. that they're two-character codes from some list somewhere so you write lang="en" and not lang="english" |
| 14:27 | <Philip`> | (Maybe that'd be part of the definition of the global lang attribute, once it's defined?) |
| 14:29 | <Philip`> | (If so, it might still be good to mention it explicitly in the html definition, since that's where most people are going to use lang, and most aren't going to bother following every single cross-reference in the document) |
| 14:33 | <Philip`> | It seems like it'd be a good idea to include the doctype in XHTML examples, because people will skim through the document and see a block of code that uses the element they want to use, and copy and paste it and use it as their own document, and never know that they should add a doctype |
| 14:34 | <Lachy> | the lang and xml:lang attributes will be described in the global attributes section, and that will elaborate on langauge tags |
| 14:34 | <Philip`> | If they're advanced authors, they might notice that it's XHTML, and think "ooh, XHTML is the cool new thing so I'll copy that one rather than the HTML example" and have the same problem |
| 14:35 | <Lachy> | the way I have it set up is that "HTML Example" means an HTML only examle, "XHTML Example" means XHTML only, and "Example" means it's compatible with both. |
| 14:36 | <Philip`> | Indeed, but nobody's going to care when they're copying-and-pasting chunks of code |
| 14:36 | <Philip`> | so it's beneficial to improve the chances of them copying-and-pasting a chunk of code that has a doctype and won't be quirks mode |
| 14:37 | <Philip`> | (Hmm, isn't the "XHTML Example" for <html> already compatible with both, so it should be "Example"?) |
| 14:41 | <Philip`> | http://www.codinghorror.com/blog/archives/001234.html - "HTML Validation: Does It Matter?" |
| 14:47 | <Dashiva> | ho ho: "Google actually ranks it's indexed pages. The more valid the (X)HTML of your pages, the higher it'll appear in a search." |
| 14:50 | <Philip`> | Dashiva: Shhhh, don't tell people that that's untrue |
| 14:51 | <Philip`> | People's selfishness combined with their blind following of SEO 'experts' makes it a very effective way to encourage them to write valid pages |
| 14:53 | <jgraham_> | Apparently Google ranks your pages higher if you send me cake |
| 14:55 | <Philip`> | jgraham_: Nobody's going to believe you in IRC - you have to set up an SEO blog first, and then link to it from thousands of other SEO blogs' comment threads |
| 14:56 | <Dashiva> | Philip`: Or you can RT it |
| 14:57 | Philip` | is insufficiently down with the lingo |
| 14:57 | <Philip`> | Oh, do you mean retweet? |
| 14:58 | <Dashiva> | Yes |
| 14:58 | <Philip`> | That introduces the problem of bootstrapping yourself with followers, I assume |
| 15:01 | <Dashiva> | I'm sure there's a group of SEO twits (tweeters? twitterers?) who dedicate themselves to being a complete graph |
| 15:02 | <jgraham_> | Philip`: Frankly you are the wrong sort of people to get cake from anyway. I only really care about people trying to SEO their luxury good sites so that I get decent cake |
| 15:03 | <jgraham_> | s/good/goods/ |
| 15:04 | <jgraham_> | For example, my insider knowledge tells me that the search ranking of internationally famous London department stores is heavily contingent on a steady supply of cake |
| 15:10 | <beowulf> | finally, something i can do |
| 15:11 | beowulf | loads a pallet of cake onto a boat for jgraham |
| 15:11 | beowulf | awaits fame and fortune |
| 15:13 | <jcranmer> | I've seen SEO spam by companies who weren't in the top 10 results for "SEO" on Google |
| 15:13 | <jcranmer> | there are two explanations for this: 1. they're so bad, they need to spam for advertisement purposes, or 2. the spam is how they do SEO |
| 15:14 | <Dashiva> | Or 3. The 10 companies are their customers |
| 15:14 | <Philip`> | "We're Ranked No 1 For SEO" says the sponsored link at the top of Google's search results, linking to a site that is ranked No 11 |
| 15:15 | <Dashiva> | E.g. http://xkcd.com/125/ |
| 16:44 | <Almad> | Hi, I'd like to ask about html5lib: Is there an easy way to have only a specific portion of html sanitized? For example, I have img src attribute and I'd like to use html5lib to return an escaped/sanitized/stripped string. Is there an API for that? |
| 16:46 | <jgraham> | Almad: I think the short anser is "no" and the longer anser is "what are you trying to achieve, exactly"? |
| 16:46 | <Lachy> | jgraham, I think I might have found a bug in html5lib. http://validator.nu/?doc=http%3A%2F%2Fdev.w3.org%2Fhtml5%2Fhtml-author%2F&showsource=yes#l1033c16 |
| 16:46 | <Lachy> | although, I will update to trunk first and see if it's been resolved |
| 16:47 | <jgraham> | Specifically what is the situation in which you can sanitize a single attribute and be reasonably sure you are doing the right thing? |
| 16:47 | <Almad> | jgraham: I'm having a wiki language and I have construct like "((picture http://picture))";. I have it parsed and have "http://picture"; string and I'd like to sanitize it using html5lib |
| 16:48 | <jgraham> | Lachy: Interesting |
| 16:50 | gsnedders | agrees with that conclusion |
| 16:51 | <Lachy> | jgraham, from lookiing at my source document, the bug appears to occur with <section> <p>foo </section> <section><h1>...</h1></section> |
| 16:51 | <Lachy> | although, I haven't made a minimal TC to verify that |
| 16:51 | <jgraham> | Almad: You could do something like s = sanitizer.HTMLSanitizerMixin() |
| 16:52 | <jgraham> | s.sanitize_token({"type":"StartTag", "name":"a", "data":["href":href]} |
| 16:52 | <jgraham> | ) |
| 16:52 | <jgraham> | hich would return a token with the href sanitized |
| 16:52 | gsnedders | fails at tweeting |
| 16:53 | <jgraham> | but it's not like that isn't a hack |
| 16:53 | <Almad> | jgraham: thanks |
| 16:53 | <jgraham> | Almad: All the other ways I can think of off the top of my head require you to sanitize the whole document |
| 16:54 | <jgraham> | Which seems safer, but slower |
| 16:54 | <Almad> | jgraham: well another solution in my mind is to expand to full <img src="unsanitized content"/>, then parse it, get proper node and return it |
| 16:54 | <jgraham> | We should really refactor the HTMLSanitizer to make this use case easier |
| 16:54 | <Almad> | hmm |
| 16:54 | <Lachy> | jgraham, the bug seems to be fixed in trunk, so nevermind |
| 16:55 | <jgraham> | Almad: Yes, you could of course do that |
| 16:55 | <Almad> | jgraham: ...but this still seems to be hacky for me ,) |
| 16:55 | <jgraham> | Almad: Indeed |
| 16:55 | jgraham | wonders if rubys has an opinion |
| 16:56 | jgraham | wonders if rubys read the logs |
| 16:56 | <jgraham> | s/d/ds/ |
| 16:56 | <Lachy> | gsnedders, in anolis, what's the parameter to make it quote all attribute values in the output? |
| 16:56 | <gsnedders> | Lachy: Whatever html5lib's parse.py uses |
| 16:56 | <Almad> | I'm afraid about that expanding wighout sanitizing part, at least user will probably have some unexpected formatting output (at least) |
| 16:56 | <Lachy> | any idea what that is? |
| 16:57 | <gsnedders> | Lachy: no |
| 16:57 | <jgraham> | quote_attr_values |
| 16:57 | <jgraham> | I think |
| 16:57 | <gsnedders> | --quote-attr-values I guess through CLI? |
| 16:57 | <Almad> | jgraham: I'll probably use Your hack now and look at another solutions later, thanks for helping |
| 16:57 | <jgraham> | But I'm reading pms.net which is a bit buggy |
| 16:57 | <jgraham> | gsnedders: Yeah, probably |
| 16:58 | <jgraham> | Lachy: You are using the commandline? |
| 16:58 | <Lachy> | yes |
| 16:59 | <Lachy> | yay, that worked. Now when I open the generated file in my editor, it doesn't mess up the syntax highlighting. |
| 17:00 | jgraham | should really stop calling it pms.net since that seems to be a domain squatting site that is looking for a pre-menstrual-syndrome related buyer |
| 17:01 | <Philip`> | jgraham: You should buy that domain |
| 17:02 | <jgraham> | Philip`: That involves spending money. I don't think PimpMySpec is important enough to warrant two domain |
| 17:02 | <jgraham> | s |
| 17:02 | <Lachy> | if that domain isn't taken, I'd be surprised |
| 17:03 | <Lachy> | it is taken |
| 17:03 | <Lachy> | someone took a domain I wanted too, recently |
| 17:03 | <jgraham> | Lachy: Like I said, it's being squatted on. The first link on the page says "The domain might be for sale" |
| 17:04 | <Lachy> | I got html5reference.org and html5reference.net. I couldn't get the .com |
| 17:04 | <jgraham> | s/The/This/ |
| 17:06 | <Philip`> | http://dev.quickfire.org/perfect/?query=html |
| 17:06 | <Philip`> | Looks like the web isn't very perfect :-( |
| 17:07 | <jgraham> | Hmm mjs is arguing against standardising on a behaviour that webkit already has in ES3.1 |
| 17:07 | <jgraham> | Or at least saying that it might be problematic... |
| 17:45 | Philip` | wonders if the comments on http://www.codinghorror.com/ are usually as voluminous and full of people entirely missing the point and repeating misinformation as they are on the post about HTML validation |
| 17:46 | <sid0> | yes |
| 17:48 | <Philip`> | I guess I've noticed similar things in comments on The Daily WTF |
| 17:48 | <Philip`> | Is there something about popular technical blogs that turns their comments sections into floods of hundreds of comments that provide no value whatsoever? |
| 17:49 | <Philip`> | Actually, I suppose it's just that I'm judging popularity based on number of comments |
| 17:50 | <Philip`> | and at a certain point there are so many comments that people post without reading all the previous ones, and then nobody will bother posting anything worthwhile because they know it's going to get lost in the flood and nobody will read it |
| 18:17 | <jcranmer> | "This new approach to programming doesn't yet have a handy name like Ajax, though some refer to it as HTML 5." |
| 18:17 | <jcranmer> | uh... misinformation++ |
| 18:18 | <jcranmer> | the author also seems to equate Ajax with dynamic web pages |
| 18:19 | <jcranmer> | "Ajax also allowed Web pages to be more dynamic in other ways, letting users, say, right-click and see a menu tailored to their needs." |
| 18:19 | <jcranmer> | that's plain old DOM manipulation |
| 20:53 | svl | frowns. I don't grok the two errors this gives me: http://validator.nu/?doc=http://have-skill.com/presentation/2009/mozcamp Could someone enlighten me (or have I stumbled upon a bug in the validator)? |
| 20:54 | <gsnedders> | svl: opening pre tag implies a closing p tag, so there is no open p element there |
| 20:55 | <svl> | Ah, dagnabbit - I thought I'd checked that was 'inline'. |
| 20:56 | <svl> | So that leaves the datalist thing; which I copy/pasted from some Opera tutorial... |
| 20:57 | <gsnedders> | hsivonen: I think that's a validator bug |
| 20:57 | <gsnedders> | svl: ^^ also |
| 20:58 | <svl> | thanks. :) |
| 21:01 | svl | also spots an Opera bug with the actual <input list> thing, in that it causes a weird reflow removing all the linebreaks from the pre after having been activated. Where's anne when you need him anyway? |
| 21:02 | <gsnedders> | On a plane, I think |
| 21:02 | <Philip`> | A basic multilingual plane? |
| 21:02 | <gsnedders> | No, an aeroplane. |
| 21:03 | <Philip`> | Oh, right |
| 22:01 | <Philip`> | http://www.crummy.com/software/BeautifulSoup/3.1-problems.html |
| 22:58 | <gsnedders> | Wow. Markup by me with closing tags. |