| 04:10 | <Hixie> | Lachy: yt? |
| 04:10 | <Lachy> | yeah |
| 04:11 | <Hixie> | someone submitted a blog entry |
| 04:11 | <Hixie> | (just fyi, i'm sure you'll see the e-mail in due course!) |
| 04:13 | <Lachy> | I haven't yet. Who sent you the mail? |
| 04:13 | <Hixie> | it was sent to the list |
| 04:15 | <Lachy> | oh, I thanks for pointing it out. It would have been a while before I read it |
| 04:16 | <Lachy> | there's actually 2 waiting for publishing. One is a short description and a link to http://blog.to-be.pl/2007/08/11/nowe-elementy-w-html-5/ |
| 04:16 | <Lachy> | the other is about ruby |
| 04:16 | <Hixie> | cool |
| 04:16 | <Lachy> | not sure what language that is, though |
| 04:17 | <Lachy> | I like to include the language in the title, if the post isn't in english |
| 04:17 | <Hixie> | .pl is poland |
| 04:17 | <Hixie> | i think |
| 04:17 | <othermaciej> | that's Polish |
| 04:17 | <Lachy> | oh, I didn't notice the domain |
| 04:17 | <Hixie> | (btw, i recommend not reviewing the posts -- so long as they're not spam, i'd just post them) |
| 04:18 | <Hixie> | (after all, we only added the review step because of spammers) |
| 04:18 | <othermaciej> | it's about "new elements in html5" |
| 04:18 | <Lachy> | ah, that's a translation of that IBM article I read recently |
| 04:18 | <Lachy> | ah, there's a link to it at the bottom http://www-128.ibm.com/developerworks/library/x-html5/ |
| 04:19 | <othermaciej> | I can read Polish |
| 04:19 | <othermaciej> | although technical terms sometimes confuse me |
| 04:26 | <Lachy> | oops! I should have set the post slug before publishing - it's a bit long. http://blog.whatwg.org/xhtml-5-will-have-the-only-usable-implementation-of-ruby-markup |
| 04:26 | <Lachy> | oh well, too late |
| 04:26 | <Hixie> | heh |
| 06:40 | <othermaciej> | which is up to #3 on my todo queue |
| 09:46 | <Hixie> | sigh |
| 09:46 | <Whiskey_M> | what's up? |
| 09:46 | <Hixie> | i scan several billion documents, and robert burns isn't happy with the quality of my research |
| 09:46 | <zcorpan_> | Hixie: you're not trying hard enough ;) |
| 09:47 | zcorpan_ | hides |
| 09:47 | <Whiskey_M> | I hope you did them by hand ;-) |
| 09:47 | <krijnh> | Yeah, we should get ourselves a new editor ;) |
| 09:47 | <Hixie> | and the other reply to my last public-html mail tells me that i shouldn't be making decisions |
| 09:47 | <krijnh> | Robert Burns likes writing.. |
| 09:47 | <Hixie> | oh well |
| 09:59 | hendry | sighs at the weather |
| 10:07 | Hixie | updates the <img alt> section |
| 10:11 | <karlUshi> | I think what robert burns is saying |
| 10:11 | <karlUshi> | is that stats are not enough |
| 10:12 | <karlUshi> | stats give us a rough evaluation of quantity, but not the quality. |
| 10:12 | <Hixie> | hah, the great alt="" revamp got the auspicious revision number r1000. |
| 10:12 | <karlUshi> | quality being what are the processes which have led to this stats. How authoring tools operate etc |
| 10:12 | <Hixie> | karlUshi: except in the case he's talking about, the stats were specifically about determining the quality of the content. |
| 10:13 | <gsnedders> | Hixie: scan the whole google cache. kthxbai. |
| 10:13 | <Hixie> | that's not far from basically what i'm doing actually (i work for google) |
| 10:13 | <karlUshi> | quality (in my sentence above) = using authoring tool by hand and sees how the markup is produced in an user interaction. |
| 10:14 | <karlUshi> | s/an user/a user/ |
| 10:14 | <gsnedders> | Hixie: I'm aware of what you're doing. I'm just saying _whole_ :) |
| 10:14 | <Hixie> | (though i scan a subset of the whole thing usually for reasons of not wanting to wait three months to get the results) |
| 10:14 | <Hixie> | i've done the whole thing sometimes :-) |
| 10:14 | <gsnedders> | Hixie: I'm just wondering how you can satisfy Robert Burns better with the data you have access to |
| 10:14 | <karlUshi> | gsnedders: that would not solve the problems ;) |
| 10:14 | <gsnedders> | karlUshi: I know |
| 10:15 | <Hixie> | gsnedders: i don't think bigger numbers would make the slightest difference |
| 10:15 | <Hixie> | gsnedders: i'm really not sure what would |
| 10:15 | <Hixie> | ok, any immediate comments on the alt text revamp? |
| 10:15 | <gsnedders> | Hixie: can the entire web. |
| 10:15 | <karlUshi> | exactly because bigger numbers don't do any differences |
| 10:15 | <gsnedders> | *scan |
| 10:15 | <gsnedders> | even if that is technically impossible! |
| 10:15 | <karlUshi> | quantity versus quality. |
| 10:15 | <Hixie> | statistically, what i've scanned is close enough to the entire web as makes no difference, to be honest |
| 10:16 | <gsnedders> | there is realistically no point in doing anything bigger, though |
| 10:16 | <Hixie> | right |
| 10:16 | <gsnedders> | you'd just get so similar results |
| 10:17 | <Hixie> | it's actually amasing how similar results others have been getting with just samples of a few thousand |
| 10:17 | <karlUshi> | Hixie: If you want different results, imho, you have to categorize |
| 10:18 | <Hixie> | and that's like 0.001% of what i've scanned |
| 10:18 | <karlUshi> | small random sample = big random sample with less confidence |
| 10:19 | <karlUshi> | what is interesting is what is hidden in the data when we just do samples |
| 10:19 | <karlUshi> | http://roslingsblogger.blogspot.com/ |
| 10:22 | <virtuelv> | Hixie: re usemap: I'd like to see the opposite stats for what you did here, http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-August/012334.html |
| 10:22 | <virtuelv> | tell me which percentage breaks in which way |
| 10:23 | <virtuelv> | for some weird reason, that seems to work better as a justification for people |
| 10:24 | <Hixie> | virtuelv: well, go through the list of URLs there and categorise them into the various reasons they fail (or pass), and you'll have that data :-) |
| 10:24 | <karlUshi> | "the problem is not ignorance but preconceived ideas" |
| 10:25 | <Lachy> | Hixie: which UAs support <input useemap>? |
| 10:25 | <Hixie> | firefox and opera, for some definition of "support" |
| 10:26 | <Hixie> | (i mean, the html4 spec was no more detailed about what it should do than it is for anything else) |
| 10:26 | <Lachy> | ok, so the fact that IE doesn't support it should be a good reason for not bothering to include it |
| 10:26 | <virtuelv> | Hixie: neh, just mentioning for future reference. I only follow certain parts of html5 work |
| 10:27 | <Lachy> | though I'm really confused by Robert's statement "Dropping client-side image maps for forms in a recommendation that is focussed on web applications seems completely counter-intuitive to me." |
| 10:27 | <Hixie> | yeah i didn't follow his e-mail at all |
| 10:27 | <Lachy> | I have no idea what use cases he's imaginign for it |
| 10:27 | <Whiskey_M> | is there a URL where his e-mail is stored? |
| 10:28 | <Lachy> | http://lists.w3.org/Archives/Public/public-html/2007Aug/0462.html |
| 10:28 | <Hixie> | http://lists.w3.org/Archives/Public/public-html/2007Aug/0462.html |
| 10:28 | <Hixie> | d'oh |
| 10:28 | <Whiskey_M> | 'ta :) |
| 10:32 | <zcorpan_> | Hixie: i love the examples on <img alt> :D |
| 10:32 | <Hixie> | :-D |
| 10:35 | <Hixie> | there are about 10 in-jokes in those examples |
| 10:35 | <Hixie> | some obvious |
| 10:35 | <Hixie> | some very obscure |
| 10:36 | <gsnedders> | Hixie: I'm now going to get no work done today, just looking for them :P |
| 10:36 | <Hixie> | hah |
| 10:37 | <gsnedders> | at first glance I thought "XYZ.gif" read "YYZ.gif" |
| 10:37 | <Hixie> | XYZ.gif is one of the less obvious references |
| 10:37 | gsnedders | ponders |
| 10:37 | <gsnedders> | (and starts playing YYZ while doing so) |
| 10:39 | <Lachy> | heh, I can't wait till I get an iPhone so I can compare the reflectiveness of my apple logo :-) |
| 10:39 | <gsnedders> | Hixie: please don't say 1100670787_6a7c664aef is a joke. |
| 10:39 | <Hixie> | hah |
| 10:39 | <Hixie> | gsnedders: yup :-) |
| 10:39 | gsnedders | stabs Hixie |
| 10:39 | <gsnedders> | sorry, but that's just evil. |
| 10:39 | <Hixie> | actually that one is one of the more funny ones |
| 10:39 | <Hixie> | and one of the easiest to track down |
| 10:39 | <gsnedders> | I can make several guesses about what it is |
| 10:40 | gsnedders | wonders if he can search such things for such an ID easily |
| 10:41 | <Hixie> | actually i don't know how you would track it down, come to think of it |
| 10:41 | <Hixie> | oh, i know |
| 10:42 | <Hixie> | aha, yes |
| 10:42 | <Lachy> | oh, don't reveal the answer yet, I want to try and figure them all out ! |
| 10:42 | <Hixie> | i can do one search for a very specific, obvious string, on one site, to find the reference that's to |
| 10:42 | <gsnedders> | OBVIOUS!? |
| 10:43 | <gsnedders> | I really should work :\ |
| 10:43 | zcorpan_ | too |
| 10:43 | krijnh | too |
| 10:43 | <Hixie> | hah |
| 10:43 | <Lachy> | I'm trying to work out what "sequence of rounded squares with varying shades of green and bold white outlines" refers to. That doesn't sound like the old WHATWG logo |
| 10:43 | <zcorpan_> | Hixie: stop wasting our time! ;) |
| 10:43 | <gsnedders> | we're so hard working here |
| 10:43 | <Hixie> | tee hee |
| 10:43 | <gsnedders> | just on the wrong things |
| 10:43 | <Hixie> | Lachy: that one is one of the other pretty funny ones |
| 10:44 | <Lachy> | oh, got it! |
| 10:45 | <Hixie> | hah |
| 10:45 | gsnedders | still ponders |
| 10:45 | <Lachy> | I missed the bit about the blue book, and then that reminded me of John's book |
| 10:46 | <gsnedders> | Lachy: shhhh |
| 10:46 | <Hixie> | assuming you all got the reference in the first example (and the other later one with the same text), then there are only three that are really possible to get |
| 10:46 | <gsnedders> | Hixie: so kind |
| 10:46 | <Hixie> | that's the XYZ one, the "consider for a moment" one, and the 1100670787_6a7c664aef one |
| 10:47 | hsivonen | found 1100670787_6a7c664aef |
| 10:47 | <Hixie> | teehee |
| 10:47 | <Hixie> | you're good |
| 10:47 | <gsnedders> | hmmm… |
| 10:47 | <gsnedders> | it'll be some ID specific to some site |
| 10:47 | <gsnedders> | (probably) |
| 10:48 | <Hixie> | (the other examples are either references that aren't funny, or just made up things based on the other in-jokes) |
| 10:48 | <gsnedders> | Hixie: but no, I didn't get that reference in the first example |
| 10:48 | <Hixie> | the first example should be immediately obvious to anyone here :-P |
| 10:49 | <gsnedders> | where it comes from, or what? |
| 10:49 | <Hixie> | yeah |
| 10:49 | <Hixie> | it's not funny |
| 10:49 | <gsnedders> | yeah, I got that straight away |
| 10:49 | <Hixie> | right |
| 10:49 | <gsnedders> | I simply glanced over that, looking too deeply |
| 10:49 | <Hixie> | hehe |
| 10:49 | gsnedders | needs to implement that, actually |
| 10:49 | <Lachy> | how is the XYZ company a joke? |
| 10:50 | <hsivonen> | I like the /images/logo example |
| 10:50 | <Hixie> | lachy: it's a reference to something quite old, and the third example from that series is based on it |
| 10:50 | <Hixie> | not saying it's especially funny mind you |
| 10:50 | gsnedders | laughs at what iTunes randomly plays |
| 10:50 | gsnedders | is listening to I Still Haven't Found What I'm Looking For by U2 from Rattle And Hum |
| 10:50 | <Hixie> | hah |
| 10:51 | <gsnedders> | myself and a friend have a running joke about iTunes being psychic. it _always_ does such things, but _only_ on OS X. |
| 10:51 | <Lachy> | Hixie: the 3rd example in the whole spec from the same series? |
| 10:52 | <Lachy> | or, actually, I'm not sure what you meant by "the third example from that series is based on it" |
| 10:53 | <gsnedders> | I can't get the ΑΒΓ joke at all |
| 10:53 | <Hixie> | Lachy: the third example in the xyz/alpha-beta-gamma series is a reference to something specific |
| 10:54 | <Hixie> | gsnedders: which one? |
| 10:54 | <gsnedders> | Hixie: both |
| 10:54 | <gsnedders> | Hixie: can't get the XYZ one either |
| 10:54 | <Hixie> | ah :-) |
| 10:54 | <gsnedders> | am I just too young? :) |
| 10:55 | <Hixie> | could be, it's reference to something from the late 90s |
| 10:55 | <Whiskey_M> | am I right in thinking (sorry doing other stuff so slow), that the <input type="image" usemap="..." >, functionality would trigger a get user agent request? (sorry for being dim, obviously need more tea) |
| 10:55 | <Hixie> | though i don't expect anyone here to get it without searching for it |
| 10:55 | gsnedders | feels young again |
| 10:55 | <Lachy> | oh, I see. I thought "series" was referring to some TV series or something |
| 10:56 | <Hixie> | Whiskey_M: how do you mean? when clicked? |
| 10:56 | <Whiskey_M> | yup |
| 10:57 | <gsnedders> | "sequence of rounded squares with varying shades of green and bold white outlines" |
| 10:57 | <gsnedders> | hmmm… |
| 10:57 | <Whiskey_M> | sorry, into a meeting now |
| 10:57 | <Hixie> | Whiskey_M: yeah, it would work much the same way as <img usemap> |
| 10:57 | <gsnedders> | I'd expect it to have been published on a blue book |
| 10:58 | <gsnedders> | 11:06+01 and I haven't started work yet. :\ |
| 11:02 | <Lachy> | Hixie: "... but if the image makes it a lot easier for users of visual browsers to understand the concept." - either remove "if" or finish the statement. |
| 11:03 | <Hixie> | oops, missing "including" |
| 11:04 | <Lachy> | I think you should s/if/including/ |
| 11:05 | <Hixie> | i was going to do s/if/if including/ |
| 11:05 | <Hixie> | but yeah, i'll remove the second if |
| 11:05 | <Lachy> | nah, that doesn't work |
| 11:05 | <Hixie> | (and replace it with including) |
| 11:06 | hsivonen | doesn't get the specific reference of the alpha-beta-gamma pie chart |
| 11:06 | <Lachy> | it sounds like it could be some Dilbert reference or something |
| 11:06 | <Hixie> | much more obscure than that |
| 11:06 | <Hixie> | (much closer to home than that, too) |
| 11:07 | <hsivonen> | I thought it was a reference to the latest Stevenote but the colors are wrong |
| 11:07 | <Hixie> | oooh, i didn't think of having examples to keynotes |
| 11:07 | <Lachy> | Hixie: where's the obligatory Stargate reference in this section? |
| 11:08 | <Hixie> | heh |
| 11:08 | <Hixie> | haven't seen stargate in months, it's not on the mind |
| 11:09 | <Hixie> | talking about not understanding references, can anyone work out wtf this e-mail is talking about? http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-August/012351.html |
| 11:10 | <Hixie> | "My sense is that a consensus is building in the group to use a generalized XML-like syntax for the examples and illustrations in the semantics chapter." |
| 11:10 | <Hixie> | it is? |
| 11:11 | <hsivonen> | Hixie: no |
| 11:11 | <Hixie> | no to the e-mail, or to the examples' syntax? :-) |
| 11:11 | <Hixie> | or both? :-) |
| 11:11 | <hsivonen> | Hixie: the latter |
| 11:12 | <hsivonen> | Hixie: syntax that is |
| 11:12 | <Hixie> | ah ok |
| 11:12 | <Hixie> | well good |
| 11:12 | <Hixie> | though someone should probably let him know |
| 11:12 | <Lachy> | I found 1100670787_6a7c664aef, but I don't know who the person in the photo is |
| 11:12 | <Hixie> | not it! |
| 11:12 | <Hixie> | Lachy: me |
| 11:12 | <Lachy> | oh |
| 11:14 | <gsnedders> | hah |
| 11:15 | <Hixie> | right |
| 11:15 | <Hixie> | bed time |
| 11:15 | <Hixie> | hsivonen: i added the red boxes like you suggested |
| 11:16 | <gsnedders> | Lachy: mind PMing me the link? |
| 11:16 | <gsnedders> | Lachy: to 1100670787_6a7c664aef? |
| 11:16 | <Lachy> | no |
| 11:16 | <gsnedders> | why was I expecting that? :) |
| 11:16 | <hsivonen> | Hixie: thanks. the div one is rather opinionated ;-) |
| 11:16 | <Lachy> | I thought it would be obvious which site it came from based on the filename. After that, it's not too hard to figure out what to search for |
| 11:16 | <Hixie> | hsivonen: :-D |
| 11:17 | <Lachy> | the div one? Which one is that? |
| 11:17 | <gsnedders> | Lachy: but surely such a filename could be any photo gallery? |
| 11:17 | <Hixie> | hsivonen: i have but one thing to say about <div>: http://gallery.mac.com/emily_parker |
| 11:18 | <hsivonen> | Lachy: see rev 1001 |
| 11:19 | <Hixie> | that url also serves well as an argument for irrelevant="" |
| 11:19 | colione | found the 1100670787_6a7c664aef too |
| 11:19 | zcorpan_ | also, btw |
| 11:19 | gsnedders | feels dumb |
| 11:20 | colione | wonders how many really famous photo sharing sites there is.. :) |
| 11:20 | <Hixie> | colione: well, i just posted a link to one above :-D |
| 11:21 | <Hixie> | so at least 2 ;-) |
| 11:21 | gsnedders | can't get what to search for |
| 11:21 | <colione> | :D |
| 11:22 | <gsnedders> | I'm probably on the right site, at least |
| 11:23 | <gsnedders> | ah. searching for a substring of what is needed doesn't help :D |
| 11:23 | <gsnedders> | hmm… it only has two views :\ |
| 11:26 | <hsivonen> | gsnedders: try the API |
| 11:27 | <gsnedders> | hsivonen: multiple have confirmed I got it right, though |
| 11:27 | zcorpan_ | got the logo joke |
| 11:27 | <gsnedders> | *multiple people |
| 11:29 | <Hixie> | zcorpan_: woot |
| 11:31 | <Hixie> | right, *gone* |
| 11:35 | <zcorpan_> | xml doesn't define error handling?? |
| 11:37 | <gsnedders> | WHAT!? |
| 11:37 | <zcorpan_> | http://www.w3.org/mid/54DFD15E-C8EA-445C-A674-8611405F77CD⊙rc |
| 11:37 | <gsnedders> | actually, I think XML says it's an error, but not fatal |
| 11:38 | <gsnedders> | no, that must be something else |
| 11:39 | <gsnedders> | here we go. |
| 11:39 | <gsnedders> | "Note that non-validating processors are not obligated to to read and process entity declarations occurring in parameter entities or in the external subset; for such documents, the rule that an entity must be declared is a well-formedness constraint only if standalone='yes'." |
| 11:39 | <gsnedders> | http://w3.org/TR/xml/#wf-entdeclared |
| 11:40 | <zcorpan_> | right, so if there's an external dtd and you're non-validating, entity references you don't know about is an error but not fatal |
| 11:40 | <gsnedders> | it doesn't specify what to do, though |
| 11:40 | <Lachy> | zcorpan_: yes |
| 11:41 | <Lachy> | but I have no idea where Rob is getting his ideas about stray & and < characters |
| 11:41 | zcorpan_ | neither |
| 11:41 | <Lachy> | they're defined as fatal well formedness errors |
| 11:42 | <gsnedders> | an entity to be treated as above still has to match the requirements for an entity ('&' Name ';') |
| 11:43 | <zcorpan_> | indeed |
| 11:43 | <gsnedders> | which does mean there will be some where that is the case |
| 11:44 | <gsnedders> | surely we have to assume that we have a standard XML parser that creates a DOM, so we can't do anything to it? |
| 11:44 | <hsivonen> | fwiw, the reason why unrecognized entities are fatal in Gecko is that as far as expat is concerned, it hasn't skipped the DTD but parsed a zero-length DTD |
| 11:45 | <hsivonen> | the Gecko entity resolver resolves unknown external entities to a zero-length stream |
| 11:45 | <hsivonen> | unknown: those that don't have a mapping in the built-in catalog |
| 11:49 | <hsivonen> | Rob is right that by configuring the XML processor to skip extenal entities and by using a reference to an external entity in the document, one can open up a loop hole that allows entity references to be expanded on the application level |
| 11:50 | <zcorpan_> | hsivonen: but he says that stray & and < are not well-formedness errors |
| 11:50 | <hsivonen> | however, the question that one must ask what good is using XML if you do your utmost to exploit loop holes to escape the nature of XML |
| 11:50 | <hsivonen> | zcorpan_: yeah, that part is bogus |
| 12:02 | <takkaria> | Hixie: typo in new <img> text, grep for "Standadrs" in one of the examples |
| 12:18 | <hendry> | anyknow what what WICD is about? |
| 12:18 | <hendry> | anyone know what WICD is about? http://www.w3.org/TR/WICD/ |
| 12:22 | <zcorpan_> | hendry: it's wicked |
| 12:22 | <hendry> | zcorpan_: not crazy? |
| 12:23 | <zcorpan_> | </pun> |
| 13:23 | <Lachy> | I wrote a script to mark the issues in the spec with the text "Big Issue:" http://lachy.id.au/dev/utilities/#mark-issues |
| 13:23 | <Lachy> | I made a bookmarklet to do it manually, though we could probably get the script added to the spec |
| 13:28 | <hsivonen> | Lachy: wouldn't CSS generated content work fine? |
| 13:28 | <Lachy> | how well does that work with assistive technology? e.g. JAWS running on IE |
| 13:29 | <Lachy> | I wrote it based on Steve Faulkner's post to public-html |
| 13:37 | <hsivonen> | Lachy: oh. IE... |
| 13:37 | <Lachy> | yeah, I just fixed that bug, reload the script |
| 13:38 | <Lachy> | now I can't figure out why it's not working in Opera |
| 13:39 | <Philip`> | http://www.twaynet.com/trade/t.asp?ID=1100670787 - aha, that alt text example is a reference to Peek Pi Pps Engineering Plastics Bearings! |
| 13:41 | <hsivonen> | Philip`: huh? |
| 13:42 | <Philip`> | It's the same number as in 1100670787_6a7c664aef.jpg and there's only a one-in-a-billion chance of it being a coincidence :-) |
| 13:42 | <hsivonen> | http://diveintomark.org/archives/2007/08/13/cc-elephants accessibility guy uses alt for title & authorship data |
| 13:42 | <Lachy> | ah, I had uploaded the old version of the bookmarklet. Now it works! :-) |
| 13:42 | <hsivonen> | Philip`: oh |
| 13:43 | <virtuelv> | Lachy: got a URL I can test that script on? |
| 13:43 | <Lachy> | yeah, the spec |
| 13:44 | <Philip`> | Hixie: Other typos in the <img> thing: "greek" -> "Greek"; "aquisition" -> "acquisition"; "shallot.jpeg" -> "shalott.jpeg" |
| 13:47 | <Lachy> | hmm. It doesn't work in Safari :-( |
| 13:47 | <Lachy> | is there some trick to getting bookmarklets to work in it? |
| 13:52 | <Lachy> | hmm. mark-issues.js works fine in Safari, but the bookmarklet is broken. Oh well. |
| 14:02 | <virtuelv> | Lachy: it's the bookmarklet itself that breaks |
| 14:02 | <Lachy> | yeah, I know. Do you know why? |
| 14:03 | <zcorpan_> | http://simon.html5.org/temp/zon/ working prototype parser for a custom tests format i came up with to be able to write lots of tests for getting innerHTML |
| 14:04 | <virtuelv> | Lachy: looking into it, because it should not break |
| 14:08 | <virtuelv> | Lachy: dunno exactly what happened, but this: javascript:(function(){var d=document.documentElement.appendChild(document.createElement('script'));d.src='http://lachy.id.au/dev/utilities/mark-issues.js'})() works in my version of Opera |
| 14:09 | <Lachy> | yeah, it works in Opera fine now. I fixed the bug that was causing Opera to break already |
| 14:09 | <Lachy> | oh, did you mean Safari? |
| 14:10 | <virtuelv> | don't have Safari here |
| 14:11 | <Lachy> | oh, then why was my version breaking in Opera? Were you using the older version? |
| 14:11 | <virtuelv> | no, I tested in an internal version, in which your bookmarklet broke |
| 14:12 | <colione> | Nice work zcorpan_ |
| 14:15 | <zcorpan_> | thanks colione |
| 14:26 | <Lachy> | virtuelv: ok, you'd better file a bug report for it then :-) |
| 14:26 | <Lachy> | virtuelv: I didn't realise you worked at Opera |
| 14:27 | <virtuelv> | Lachy: but you changed the bookmarklet somewhat? I retested it, and now it seems to work |
| 14:28 | <Lachy> | the last change I made to it was to add the outer (function() { ... })() |
| 14:28 | <Lachy> | the original one I had uploaded omitted it by mistake and, although it worked in Firefox, it broke in Opera |
| 14:29 | <virtuelv> | I tested the one that didn't wrap it in a function |
| 14:29 | <Lachy> | ok |
| 14:39 | <zcorpan_> | "User agents may adjust prefixes and namespace declarations in the serialisation (and indeed might be forced to do so in some cases to obtain namespace-well-formed XML)." -- should this perhaps be fleshed out and be specced as a set of requirements so that it can be tested? |
| 14:39 | <zcorpan_> | getting innerHTML in XML that is |
| 15:01 | <gsnedders> | or even speccing it out as an algorithm, so everything must implement it consistently? |
| 15:03 | <zcorpan_> | yeah |
| 15:03 | <zcorpan_> | that would be good |
| 15:26 | <Lachy> | charlvn added the script to the annotate-web-apps.js script, so now the issues get annotated automatically :-) |
| 15:44 | <Lachy> | ah... how should I respond to this? http://www.w3.org/mid/46C1BF25.9060808⊙RAU |
| 15:45 | <gsnedders> | Lachy: that phrasing confuses me slightly… |
| 15:46 | <Lachy> | in what way? |
| 15:46 | <gsnedders> | Lachy: "Many sites that used usemap="" on <input type="image"> cease to work in browsers that implement usemap, therefore it has been dropped from HTML 5." |
| 15:46 | <Philip`> | Try giving an example where it breaks |
| 15:46 | <gsnedders> | Lachy: I dunno. It just is confusing. |
| 15:46 | <Philip`> | (preferably of a real site) |
| 15:47 | <Lachy> | yeah, I was just making a test case |
| 15:47 | <Lachy> | to see how it works |
| 15:47 | <Philip`> | and then say that UAs which don't implement <img usemap> work better than those which do |
| 15:47 | <Philip`> | and hopefully that'd be clear enough :-) |
| 15:48 | <Lachy> | ok. |
| 15:48 | <gsnedders> | and list UAs |
| 15:48 | <Whiskey_M> | <img usermap, or <input type="image" usermap ? |
| 15:48 | <gsnedders> | Whiskey_M: the latter |
| 15:48 | Philip` | wonders if this is the kind of place where the HTML 5 test suite should explicitly test that UAs don't support usemap |
| 15:48 | <takkaria> | Lachy: I was very tempted to ask if he could just read it again and save us all bandwidth ;) |
| 15:49 | <Philip`> | (and how you can do that kind of testing nicely when the spec doesn't talk about usemap at all) |
| 15:49 | <Lachy> | ha! |
| 15:49 | <gsnedders> | Philip`: does anything prohibit UAs that implement HTML5 from supporting further elements? |
| 15:50 | <Philip`> | Oops, yes, I didn't mean <img usemap> |
| 15:50 | <Lachy> | perhaps, in cases like this, the spec should give explicit requirements for certain attributes to be ignored |
| 15:52 | <Philip`> | gsnedders: Don't know - it just seems good for UA quality assurance if the HTML test suite can test that UAs work as well as possible, and ignoring usemap appears to be some that is needed for a UA to work as well as possible |
| 15:53 | <Philip`> | and I don't know how that situation would be best handled |
| 15:58 | <takkaria> | I think the spec requring a certain attribute not be handled that was defined in a previous version of the spec would be a good solution; or at least saying that for compatiblity, UAs should not implement the attribute |
| 16:02 | <Philip`> | I posted http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-June/012063.html a while ago about supporting non-standard features - keeping non-standardised support for compatibility reasons seems like a bad idea in terms of having a relevant-in-the-real-world spec |
| 16:02 | <Philip`> | (because if it's needed for compatibility, it should be in the spec, and if it's not needed, it shouldn't be in the implementations) |
| 16:03 | <Philip`> | It feels kind of cruel to look in people's source code and create test cases specifically for them to fail, though |
| 16:06 | <gsnedders> | Philip`: keeping? what about adding? |
| 17:11 | <hsivonen> | Lachy: I don't understand what difference between Safari and Firefox http://alforag.com/donate/ illustrates |
| 17:13 | <Lachy> | it's the Sign up for updates form that breaks, were you testing the other one below (which I didn't notice originally) |
| 17:14 | <Lachy> | In Firefox, when you click the go button, it's treated as a link to a file:// URI and goes nowhere. In Safari, it submits the form. |
| 17:15 | <hsivonen> | Lachy: oh, right. I clicked the image outside the red rectangle, so I didn't notice |
| 17:16 | <Lachy> | oh, I didn't notice one could click outside the red |
| 17:29 | <gsnedders> | Lachy: being very observant today? :) |
| 17:30 | <Lachy> | gsnedders: me being observerant? I don't think so! |
| 17:31 | <gsnedders> | maybe this is why I try and do backend development. less observation needed. |
| 18:11 | <takkaria> | there does seem to be a strong expert-bias in public-html |
| 18:15 | <Lachy> | takkaria: what do you mean by expert-bias? |
| 18:17 | <takkaria> | the comment that brought it about was Philip Taylor's "allowing the craziness of a small number of individuals to outweigh the combined wisdom of all those who formulated the specification for HTML 4.01" |
| 18:19 | <takkaria> | the idea being that an expert's opinion is worth far more than anyone else's, in short |
| 18:19 | <hsivonen> | takkaria: curiously, those who formulated HTML 4.01 are automatically assumed to have had wisdom whereas those who formulate HTML 5 aren't |
| 18:19 | <hsivonen> | I guess that happens when a spec is old enough |
| 18:19 | <takkaria> | interesting how people acquire mythical status over time. :) |
| 18:20 | <Lachy> | yeah, despite the fact that we have a whole 10 years worth of real world evidence that the HTML4 WG didn't have |
| 18:23 | <gsnedders> | why don't we just continue using SGML? fuck the real world, it keeps us consistent with the wisdom of HTML 4.01 |
| 18:25 | <Lachy> | gsnedders, XML is the future, SGML failed |
| 18:25 | <gsnedders> | Lachy: but that goes against the wisdom of the old HTML WG! |
| 18:26 | <gsnedders> | Lachy: and they must be right! |
| 18:26 | <gsnedders> | Lachy: and you can't say anything to change that! |
| 18:26 | <Lachy> | the old HTMLWG already began the migration to XML as soon as it became available |
| 18:26 | <Lachy> | they ditched HTML4 after XHTML1 |
| 18:26 | <takkaria> | I imagine who you define as an expert is a function of how much your views and theirs match up. :) |
| 18:27 | <gsnedders> | Lachy: were they still called the HTML WG? |
| 18:27 | <Lachy> | takkaria, I define you as an expert :-) |
| 18:27 | takkaria | grins |
| 18:27 | gsnedders | defines self as idiot |
| 18:27 | <Philip`> | Did they actually care about SGML when writing HTML4, or did they just keep using it because that's what the previous versions of HTML used? |
| 18:27 | <gsnedders> | Philip`: IIRC it's that they didn't want to reverse engineer everything |
| 18:27 | <Lachy> | gsnedders, yes, the status of XHTML1 still talks about the "HTML Working Group" |
| 18:28 | <gsnedders> | Lachy: when did the name change? |
| 18:28 | <gsnedders> | Lachy: the rechartering this year? |
| 18:28 | <Lachy> | it changed from HTML WG to XHTML2 WG earlier this year |
| 18:28 | <gsnedders> | so yeah |
| 18:29 | <Lachy> | I'm sure they just kept using it based on the wisdom of those who wrote HTML 3.2 using SGML |
| 18:29 | <Lachy> | which was in turn based on the wisdom of those who wrote HTML2 |
| 18:29 | <Lachy> | which was actually not based on the wisdom of TBL who didn't originally use SGML when he first created it |
| 18:30 | <gsnedders> | which was based on the criticism of the IIIR-HTML draft? |
| 18:30 | <Philip`> | So they were following the SGML cow path? :-) |
| 18:30 | <gsnedders> | was I even born when the IIIR-HTML draft was written? |
| 18:30 | <gsnedders> | 1993. ah good. I was alive! |
| 18:31 | <Whiskey_M> | gs: lol |
| 18:31 | <takkaria> | did they have cowpaths back then? :) |
| 18:31 | <Lachy> | what does IIIR stand for? |
| 18:32 | <Lachy> | ah, Integration of Internet Information Resources Working Group |
| 18:32 | <Philip`> | They should have called it the I3R |
| 18:32 | <gsnedders> | looking it up on Wikipedia gives a search page, results including "Transsexual sexuality" :\ |
| 18:33 | <takkaria> | sounds like grounds for a libel claim to me |
| 18:33 | <gsnedders> | note to self: wikipedia does not have the answers to everything. google does. |
| 20:43 | <Jero> | hey, was reading http://pearlcrescent.com/products/pagesaver/ and read their extension uses the <canvas> element to somehow make a screenshot of the page. |
| 20:43 | <Jero> | "Page Saver uses the canvas feature that was introduced in Firefox 1.5." |
| 20:43 | <Jero> | does anyone know how they pull that off? |
| 20:44 | <Philip`> | There's a (non-standard) drawWindow method, which is usable only by extensions and not normal web content |
| 20:45 | <Jero> | i see... |
| 20:46 | <Philip`> | ctx.drawWindow(window, x, y, w, h, bgcolor) |
| 20:46 | <Philip`> | (I think) |
| 20:46 | <Jero> | interesting |
| 20:47 | <Philip`> | (It's disabled for web content because of security concerns) |
| 20:47 | <Philip`> | ( http://lxr.mozilla.org/mozilla/source/content/canvas/src/nsCanvasRenderingContext2D.cpp#2305 ) |
| 20:50 | <Jero> | Thanks! |
| 20:52 | <Jero> | the thing is, i was checking out the possibilities of creating a web application that takes a URL, feeds it to Gecko, and outputs a PNG image |
| 20:53 | <Jero> | sort of like khtml2png (http://khtml2png.sourceforge.net/) |
| 20:53 | <met_> | Philip`, this drawWindow is only for Mozilla XPCOM or is it accessible from JavaScript? |
| 20:54 | <met_> | Jero, check http://www.screengrab.org/ extension |
| 20:54 | <Philip`> | http://weblogs.mozillazine.org/roc/archives/2005/05/rendering_web_p.html |
| 20:55 | <Philip`> | met_: It works in JS too |
| 20:55 | <met_> | Philip`, just trying but it gives me some security error |
| 20:56 | <met_> | ah there is example, going to try it, thx |
| 20:56 | <Jero> | met_, thanks |
| 20:56 | <Philip`> | met_: You need UniversalBrowserRead privileges, either by being chrome (like an extension) or by doing some magic to ask the user for permission |
| 20:57 | <met_> | Philip`, ah this 8-( |
| 20:57 | <met_> | so it is not primary designed for web but for XUL apps |
| 20:57 | <Philip`> | netscape.security.PrivilegeManager.enablePrivilege('UniversalBrowserRead') |
| 20:57 | <Philip`> | (Hooray for netscape) |
| 20:58 | <Philip`> | met_: Yes, since the security issues of exposing it to the web are seemingly non-trivial |
| 20:59 | <Philip`> | (I think it would be quite a useful thing to have web access to, but I don't know how it'd be implemented) |
| 21:02 | met_ | is not sure if there are still some security problems if you can draw only window from your domain |
| 21:02 | <met_> | you can screenshot filled form, but you have access to it anyway |
| 21:02 | <colione> | There will still be som XSS concerns |
| 21:03 | <met_> | colione: iframes and frames from other domains can remain blank |
| 21:03 | <Philip`> | The window from your domain can contain content from other domains (images, iframes, etc) and it's hard to filter all that stuff out correctly |
| 21:03 | <met_> | in rendered view |
| 21:03 | <met_> | not sure if images are the real problem too |
| 21:04 | <colione> | Imagine some badly coded application that takes arguments in a querystring and don't sanitize it...you can put in any arbitrary script there |
| 21:05 | <met_> | colione, but what is problem with canvas2d.drawWindow ? |
| 21:07 | <colione> | that page might lead to a registration page and you can for example hook a function to the submit button that takes a screenshot of the page and send in over the web |
| 21:07 | <colione> | prehaps not the best example, but still possible |
| 21:09 | <met_> | if you have access to this page, frame, you can do it anyway - making screenshot is only better comfort |
| 21:11 | <Jero> | Okay, so lets say if i wanted to make a web application that takes a URL to create a screenshot of it using Gecko. Do you think it would be possible to make the app call Firefox with an extension that uses the drawWindow method on the server to save the image on the server? Or would there be a better way? |
| 21:14 | <Philip`> | Jero: http://www.hackdiary.com/archives/000055.html sounds like what you're doing |
| 21:16 | <Jero> | Philip`, oh wow, awesome, thanks! |
| 21:16 | <Philip`> | (Using an embedded Mozilla is probably much easier than trying to automate the Firefox UI) |
| 21:17 | <Jero> | yeah |
| 21:27 | <Hixie> | Philip`: there's more than one-in-a-billion chance, it's a birthday attack ;-) |
| 21:31 | <Hixie> | fixed the other typos, thanks |
| 21:31 | <Philip`> | Someone should do a survey of all the numbers that are on the web, limited to ten digits or so - it could be interesting to see if any particular numbers are unrepresented... |
| 21:32 | <Philip`> | (And then publish the results somewhere, hence filling in all the gaps) |
| 21:34 | <Hixie> | it would probably show a propensity towards numbers that start with 1 |
| 21:35 | <Hixie> | there's a theory regarding the likelihood of that in any number sample |
| 21:36 | gsnedders | wonders how high up 42 would be |
| 21:36 | <Philip`> | Hixie: Something I noticed earlier today: Why add <nav> instead of reusing <map>? HTML4 says <map> can be used for navigation stuff, and http://www.noraradcliffe.co.uk/ uses it for that (though that's the only one I could find in ~8000 pages), and you can style it properly in current versions of IE and Firefox (unlike <nav>) |
| 21:41 | <zcorpan_> | Philip`: then <map> would be sectioning element or not depending on if it was used as an image map or not... and afaict using <map> meaning navigation isn't widely understood by authors |
| 21:42 | <Hixie> | Philip`: i prefer not to overload things where it isn't necessary |
| 21:48 | met_ | is lost while reading http://webkit.org/blog/115/webcore-rendering-ii-blocks-and-inlines/ Blocks inside Inline Flows section |
| 21:49 | met_ | didn't undestund why inlines around block are splitted |
| 21:49 | <zcorpan_> | jgraham__: that would be awsome (re table algorithm prototyping) |
| 21:50 | zcorpan_ | has some ideas for possible algorithms, but not really written down |
| 21:52 | <zcorpan_> | trying to figure out header cells from TD-only tables could be interesting, but might be taking it too far...? i mean, there are data tables where all cells are just data |
| 22:00 | <Philip`> | zcorpan_, Hixie: Okay, makes sense |
| 22:11 | Hixie | sighs as he contributes yet another unhelpful e-mail to public-html |
| 22:11 | <Hixie> | maybe i should just unsubscribe |
| 22:14 | <gsnedders> | Hixie: yeah, just leave the WG |
| 22:14 | <Hixie> | the funny thing is, i have no trouble keeping up with the volume of e-mail |
| 22:14 | <Hixie> | my problem is that i sometimes can't help replying to e-mails |
| 22:14 | <gsnedders> | ah. I have no problem with either. |
| 22:15 | <gsnedders> | The amount of junk makes it easier to do the first though, as there is less in-depth reading needed |
| 22:31 | <takkaria> | I've drafted several replies to things on public-html, but they go in the "drafts" folder and I never remember about things that go there |
| 22:34 | <zcorpan_> | takkaria: yep, having public-html mail in the drafts folder for a while might be a good thing... :) |
| 22:35 | <takkaria> | well, when I come back to it, I don't remember writing it, so it gets deleted |
| 22:35 | <takkaria> | also, thunderbird has mysteriously hidden one so that it only shows up in the folder view when I look at only-unread stuff |
| 22:35 | <takkaria> | despite it not being unread |
| 22:40 | <zcorpan_> | my point being that it results in less mail to the list. :) |
| 22:40 | takkaria | nods |
| 22:41 | <zcorpan_> | if one doesn't remember what a mail was about, it probably wasn't important anyway ;) |
| 22:41 | <takkaria> | no. not on the grand scale of things |
| 22:41 | <colione> | tell the person with amnesia... |
| 22:42 | <takkaria> | normally someone else comes along and makes the same point, more or less eloquently, and I can be happy that I didn't waste bandwidth |
| 23:17 | <zcorpan_> | i'm sure i've read the logo joke somewhere before. where is it from, Hixie? |
| 23:58 | <mpt> | Philip`, the numbers survey has been done. <http://www.turbulence.org/Works/nums/prepare.html>; |
| 23:58 | <mpt> | Unfortunately it hasn't been updated since 2002 |