| 00:05 | <Hixie> | annevk42: is there anywhere i can log issues for cssom/cssom-view ? |
| 01:44 | Hixie | redefines contact information using the "declarative" style instead of the "imperative" style (along with fixing it at the same time) |
| 01:44 | <Hixie> | easier to understand my foot. |
| 04:20 | jwalden | wonders how much longer people will incorrectly argue about the complexity of an ordered set implemented as a hash table |
| 07:43 | <hsivonen> | ah the moral high ground of awareness but not solving the problem and alt="Comic" |
| 08:12 | Philip` | can't quite work out the point of Opera Unite |
| 08:12 | <Dashiva> | Hold that thought |
| 08:13 | <Madness> | I still like face gestures better. |
| 08:13 | hsivonen | wonders if there's a royalty-free spec |
| 08:14 | <Philip`> | It seems the only real advantage is that it runs the server-side stuff on your own computer, which means you don't have to pay someone for hosting (disk space and CPU) (though Opera still has to pay for all the bandwidth) |
| 08:14 | <hsivonen> | or is it HTTP and not some kind of new p2p protocol? |
| 08:14 | <Philip`> | but that doesn't seem a particularly fundamental difference |
| 08:14 | <Madness> | Philip`, when I first saw it, it reminded me of dropbox. |
| 08:15 | <Philip`> | hsivonen: It's HTTP (with a possibly-custom protocol for proxying) - you write a widget with a web server that handles GETs and POSTs etc |
| 08:15 | <Philip`> | (as far as I can tell) |
| 08:19 | <hsivonen> | the Opera Unite video reminds me of the Du bist Terrorist video |
| 08:19 | <hsivonen> | (the graphical style and narration) |
| 08:25 | <annevk42> | Hixie, the CSS WG does have a bug tracking system |
| 08:59 | <jgraham> | Hmm Mozilla are proposing restricting when __proto__ can be used to mutate an object's prototype. I'll be interested to see how that works out |
| 09:01 | <fakeolliej> | jgraham: their performance argument isn't really valid as it's an issue specific to their implementation, much like there are issues that only occur in implementations like JSC |
| 09:01 | <hsivonen> | MikeSmith: is there a particular reason to change the order of assertions vs. RELAX NG? |
| 09:01 | <hsivonen> | jgraham: is this on an Ecma list somewhere? |
| 09:02 | <jgraham> | fakeolliej: I wondered if that might be the case. But I wonder if they have done a detailed web compat analysis. I understand that __proto__ is really needed if you are not IE |
| 09:02 | <jgraham> | hsivonen: es-discuss |
| 09:02 | <hsivonen> | jgraham: thanks |
| 09:03 | <fakeolliej> | jgraham: the irony of crockford saying that people will only use "safe browsers" that "don;t support __proto__" [paraphrased] is not lost on me |
| 09:03 | <jgraham> | https://mail.mozilla.org/pipermail/es-discuss/2009-June/009523.html |
| 09:03 | <hsivonen> | uh oh. I have an <isindex> Heisenbug |
| 09:03 | <fakeolliej> | jgraham: because time and time again people have chosen "site working" over security, speed, etc |
| 09:04 | <jgraham> | fakeolliej: Indeed. I wonder if I should ask/say any of this on this list |
| 09:04 | <fakeolliej> | jgraham: we can ask hixie to tell us how many sites use __proto__ |
| 09:04 | <MikeSmith> | hsivonen: no, no current reason I can think of. When I asked you about that before, it was because at that time, I thought maybe your reason for having legacy.rnc was to control the order in which the obsolete-element errors get reported. But I realize now that the reason it be able to report errors in the subtree. |
| 09:05 | <hsivonen> | MikeSmith: ok |
| 09:07 | <hsivonen> | oops. not a Heisenbug |
| 09:08 | <hsivonen> | fortunately |
| 09:14 | hsivonen | wishes the public-html threads didn't go off on tangents based on analogies |
| 09:15 | <jgraham> | hsivonen: You made ti to the compasses I see ;) |
| 09:15 | <hsivonen> | jgraham: yes |
| 09:19 | <hsivonen> | did anyone ship DOMTokenLish.has() yet? |
| 09:20 | <MikeSmith> | are tokens in HTML5 allowed to contain character references? |
| 09:20 | <hsivonen> | MikeSmith: element and attribute names aren't |
| 09:21 | <MikeSmith> | hsivonen: but tokens in attribute values are, right? |
| 09:21 | <hsivonen> | MikeSmith: the character references are expanded before the token is created, so no in that sense |
| 09:22 | <Philip`> | http://devfiles.myopera.com/articles/608/jsrest-url.htm indicates __proto__ appears on something like 1% of pages, unless I'm misinterpreting the numbers |
| 09:22 | <MikeSmith> | hsivonen: well, I mean serialized attribute values -- e.g., in what the spec defines as "space-separated tokens" |
| 09:23 | <MikeSmith> | or rather "set of space-separated tokens" |
| 09:23 | <hsivonen> | MikeSmith: you can have character references in the serialization, yes |
| 09:23 | <MikeSmith> | k |
| 09:23 | hsivonen | was talking about different tokens |
| 09:24 | <Philip`> | http://philip.html5.org/data/proto.txt |
| 09:25 | <annevk42> | Philip`, Microsoft uses it in their compat layers for various properties |
| 09:26 | <Philip`> | The Prototype JS library seems to use it for something |
| 09:28 | <hsivonen> | grr. <isindex> is an evil special case for explicit memory management |
| 09:28 | <hsivonen> | aarrgh |
| 09:29 | <hsivonen> | now I'm annoyed at both <isindex> and at C++ not having a garbage collector |
| 09:30 | <hsivonen> | <isindex> is fast appoaching the most-buggy feature of the V.nu parser |
| 09:30 | <hsivonen> | if it isn't already |
| 09:30 | <hsivonen> | *approaching |
| 09:32 | <hsivonen> | should I add refcounting for this edge case or add a special hack with 'here be dragons' comments... |
| 09:32 | <hsivonen> | somehow I don't like the idea of refcounting just for this one special corner case |
| 09:32 | <hsivonen> | anyway, yay for unit tests |
| 09:33 | <Philip`> | Is it very unlikely that any other feature will ever trigger the same edge case? |
| 09:33 | <hsivonen> | Philip`: extremely unlikely |
| 09:33 | <hsivonen> | Philip`: <isindex> is the only parser macro |
| 09:34 | <hsivonen> | where attribute names get referenced from more than one attribute holder |
| 09:34 | <hsivonen> | everywhere else custom attribute name objects never leave their original holder |
| 09:35 | <hsivonen> | I guess I'll implement ownership transfer for this case instead of burning cycles refcounting all over |
| 09:37 | <hsivonen> | in case anyone is interested, the problem is that non-well-known attributes are copied from the isindex token to the generated <input> element and then get freed from both |
| 09:38 | <hsivonen> | I'll change things so that they get moved rather than copied |
| 09:38 | <hsivonen> | so that the isindex token no longer has attributes when its attributes are freed |
| 11:15 | <zcorpan> | hmm firefox sets tabindex="0" on video and audio elements |
| 11:37 | <hsivonen> | I have a hard time believing that the masses of Web authors could be able to deal with the codecs parameter |
| 11:40 | <zcorpan> | hsivonen: in canPlayType or in HTTP Content-Type? |
| 11:41 | <hsivonen> | zcorpan: anywhere |
| 11:41 | <hsivonen> | zcorpan: mostly in <source> |
| 11:42 | <zcorpan> | hsivonen: what do you suggest instead? |
| 11:43 | <zcorpan> | look at file extension? |
| 11:43 | <hsivonen> | I don't have a better suggestion |
| 11:43 | <hsivonen> | codecs=theora,vorbis vs. codecss=dirac,vorbis is sane |
| 11:43 | <hsivonen> | the MP4 profile stuff is just too weird |
| 11:44 | <hsivonen> | (IIRC last time I asked, even Apple QuickTime guys weren't sure what the right values for their encoder output would be) |
| 11:44 | <hsivonen> | I guess that's solved by only using it for theora vs. dirac and not using encumbered ISO stuff :-) |
| 11:45 | <hsivonen> | which remins me that we should really be beating the marketing drum about the lack of crazy AVC levels and profiles in Theora |
| 12:15 | <Philip`> | hsivonen: You mean the lack of any way to know in advance that you'll get smooth playback in a wide range of devices when using Theora? :-) |
| 12:17 | <jgraham> | Philip`: That applies to all codecs, for some sutaible definiton of "wide range" |
| 12:19 | <jgraham> | (but yeah Vorbis plays on fewer devices than mpeg-something) |
| 12:19 | <jgraham> | s/vorbis/theora/ |
| 12:20 | <Philip`> | jgraham: When there are levels and profiles, any device can say "I support this set of profiles" and any video can say "I can be played on this set of profiles" and if the intersection is non-empty then you know in advance it'll work, and that seems to be independent of the wideness of the range |
| 12:20 | <Philip`> | and I assume that's the idea of the MPEG stuff, and it's entirely lacking in Theora |
| 12:21 | <Philip`> | (I remember DivX had a similar profile system too) |
| 12:21 | <jgraham> | In practice I expect authors to put in on <source> that wworks on desktop (maybe two) and one that wworks onm the iPhone |
| 12:22 | <jgraham> | The profile system seems like a plausible solution to a technical problem that is highly likely to fail on the web |
| 12:22 | <jgraham> | s/plausible/plausible sounding/ |
| 12:23 | <Philip`> | Oh, I'd agree it seems unlikely to actually work, because it requires authors to understand and care about lots of details that they have no motivation to understand or care about |
| 12:24 | <jgraham> | Right, so not solving the problem doesn't seem like a big deal |
| 12:24 | <jgraham> | Or at least not solving it in that way |
| 12:25 | <jgraham> | In fact it can be a good thing because it leaves the oppertunity open for a solution that does work |
| 12:26 | <Philip`> | It just seems potentially dangerous to use as a marketing claim, because it could be easily twisted into saying Theora doesn't have support for a wide range of mobile devices |
| 12:31 | gsnedders | wonders what this whole Unite thing is |
| 12:33 | <Dashiva> | Are the profiles really usable that way, though? |
| 12:33 | <Dashiva> | I seem to remember there being about fifteen different ones related to h264 |
| 12:33 | <annevk42> | hsivonen, yeah, I'm not convinced at all that the whole codecs thing and canPlayType will work for anything but the simplest of cases |
| 12:37 | <hsivonen> | Dashiva: as far as I can tell, the h264 profiles are a nightmare for interop |
| 12:38 | <hsivonen> | Dashiva: if you are doing something other than encoding with QuickTime, you can spend a week tweaking parameters so that you interop with anything you can think of |
| 12:45 | <Dashiva> | While still retaining the same quality? |
| 12:45 | <Philip`> | gsnedders: It's something that's top secret and you mustn't talk about it else Opera will fire you |
| 12:45 | <gsnedders> | Philip`: But, but, but I haven't started work at Opera! :P |
| 12:45 | gsnedders | knows as much about Unite as is publicly available! |
| 12:50 | <jgraham> | gsnedders: That somewhat contradicts your earlier statement |
| 12:52 | <gsnedders> | jgraham: You assume I haven't read anything inbetween those two statemnets |
| 12:52 | <jgraham> | gsnedders: No I assume you haven't read /everything/ in between those two statements |
| 12:53 | <gsnedders> | Well, that's probably true |
| 13:04 | <Philip`> | Most of the suggested uses of Unite seem to be overestimating the uptime of a typical person's home computer |
| 13:06 | <gsnedders> | I't all these engineers assuming everyone is like them |
| 13:07 | <Philip`> | Is "I't" Klingon? |
| 13:08 | <Philip`> | I don't even have a high-uptime home computer myself - I've just got some headless servers that can't run graphical browsers, and some laptops that are switched off or are offline for at least half the day |
| 13:09 | <Philip`> | so they're not exactly great for hosting a web site or sharing music from |
| 13:09 | <beowulf> | i think it's a good idea for showing stuff quickly, when I'm at my computer |
| 13:10 | <beowulf> | saves me the bother of ftp'ing |
| 13:10 | <beowulf> | did i just say that? how lazy have i become? |
| 13:11 | <Philip`> | I suppose it could be useful in situations where someone wants to send me a file by MSN and it doesn't work because I'm using a crazy client, and if they were using Opera then they could share it that way and I could download it trivially |
| 13:12 | <Philip`> | except those people don't use Opera, so that's not going to help much |
| 13:14 | <Philip`> | (I just upload files with the same process as I use for copying local files, i.e. navigating to the right directories in Krusader and pressing F5 to copy between them, and it's no different if some of the directories are over FTP/SFTP/etc) |
| 13:14 | <annevk42> | sshfs ftw |
| 13:15 | annevk42 | uses cd/cp |
| 13:16 | <Philip`> | Things like "Fridge" and "Web Server" and "Photo Sharing" don't make much sense unless you're keeping your computer up all day, else people will get frustrated when they're down, assuming I'm not failing to notice some clever magic in Unite that makes it work when you're offline |
| 13:16 | <Philip`> | annevk42: I tried sshfs for a while but it always seemed a bit unreliable, particularly if I disconnected without unmounting it cleanly first |
| 13:17 | <Philip`> | and I like doing file management in Krusader much more than in bash :-) |
| 13:18 | <jgraham> | I don't think there is any special magic |
| 13:19 | <beowulf> | my example was nonsense |
| 13:19 | <jgraham> | (although I am far from being an expert on unite so I am avoiding saying much lest it be construed as accurate) |
| 13:20 | <gsnedders> | jgraham: But you work for Opera :P |
| 13:20 | gsnedders | has new headphones |
| 13:20 | <jgraham> | gsnedders: I noticed |
| 13:21 | <gsnedders> | jgraham: What? My new headphones? Where's the camera? |
| 13:21 | Philip` | is happy to treat Unite in the same way he treats widgets, i.e. by ignoring it entirely because he doesn't see how it would be of any value to him |
| 13:23 | <jgraham> | gsnedders: It is possible you misinterpreted my statement. It is also possible I track your post. |
| 13:23 | <gsnedders> | jgraham: I did work out what you meant, but that's boring in comparison |
| 13:24 | <beowulf> | what's the difference with mobile widgets and offline applications? |
| 13:26 | gsnedders | wonders how much SEK he should get before leaving |
| 13:27 | <gsnedders> | jgraham: Do you know whether the return tickets for the coach to/from Skavsta have a time limit on them? |
| 13:29 | <jgraham> | gsnedders: I don't recall |
| 13:30 | <jgraham> | gsnedders: There are plenty of cash machines here that will give you SEK |
| 13:30 | <gsnedders> | jgraham: You must be joking :P |
| 13:31 | <jgraham> | gsnedders: Well unless you have a silly bank account that charges you for withdrawls abroad there is little point in getting much cash |
| 13:31 | <gsnedders> | jgraham: As of a few months ago every bank/buidling society here does, AFAIK :( |
| 13:31 | <jgraham> | gsnedders: Even the nationwide? They didn't used to |
| 13:32 | <gsnedders> | jgraham: They do since the 1 June, I think the date was |
| 13:32 | <gsnedders> | 1 May, actually |
| 13:33 | <gsnedders> | "We pass on a Visa charge for processing transactions in certain currencies other than sterling. The fee is currently 0.84% and it will be included in the sterling amount shown on your statement." |
| 13:33 | <gsnedders> | Oh, there's no charge in some countries. |
| 13:33 | <gsnedders> | WTF? |
| 13:33 | <jgraham> | gsnedders: Sweden - No fee charged |
| 13:33 | <gsnedders> | Indeed |
| 13:33 | <gsnedders> | That wasn't the implication of the letter they sent out several months ago |
| 13:34 | <jgraham> | It seems like there is no fee in EU countries and a few others |
| 13:34 | gsnedders | has to pay for his Visa card though, as that's from Lloyds TSB (as they are, as far as I could tell, the only place to give under 18s a debit card) |
| 13:34 | <gsnedders> | (Whereas I only have a cash card for the Nationwide) |
| 13:35 | <Philip`> | Hasn't anybody written a non-Flash file upload system that has progress bars? (Surely you could get the server to record how much it's received already, and send that back to the client via XHR or something, but I've only ever heard of ones that do it in Flash) |
| 13:37 | <gsnedders> | There are such things |
| 13:38 | <Philip`> | http://www.lonelycode.com/2009/06/09/django-file-upload-progress-bar-without-flash/ - oh, okay |
| 13:38 | <gsnedders> | http://bluga.net/projects/uploadProgressMeter/ too |
| 13:41 | <Philip`> | "It uses the upload progress meter patch and extension from: http://pdoru.from.ro/"; |
| 13:41 | <Philip`> | You have to install a patch and recompile PHP? |
| 13:41 | <Philip`> | That really doesn't sound good |
| 13:44 | <gsnedders> | You can't normally do anything until the request has been fully received |
| 13:45 | <Philip`> | Seems like the language is somewhat lacking in flexibility |
| 13:46 | <gsnedders> | Who ever said PHP was a good language? |
| 13:46 | <Philip`> | I presume the people who designed it thought it was |
| 13:47 | <Philip`> | else they would have stuck with Perl |
| 13:47 | <gsnedders> | Wait, PHP was designed? |
| 13:47 | <Philip`> | s/designed/implemented/ |
| 15:58 | <gsnedders> | takkaria: yt? |
| 16:54 | <takkaria> | gsnedders: hi |
| 17:29 | <gsnedders> | takkaria: What do you think of Manchester (university)? |
| 17:38 | <mpt> | Is there a referencable summary/explanation anywhere of the principle that the accuracy of metadata is proportional to its visibility? |
| 17:44 | <tantek> | mpt I'm fairly certain this blog post is (one of?) the first to introduce/describe the principle: http://tantek.com/log/2005/06.html#d03t2359 "Principles of visibility and human friendliness" |
| 17:45 | <mpt> | thanks tantek, that's pretty much what I'm looking for |
| 17:47 | <tantek> | mpt - more microformats principles (in addition to visibility and human friendliness) documented here: http://microformats.org/wiki/principles |
| 18:09 | <hober> | mpt: related is ruby's postulate (accuracy of metadata inversely proportional to the square of the distance between the data and the metadata) |
| 18:09 | <mpt> | ah, that's the one I was thinking of |
| 18:09 | <mpt> | well, it's the form I was thinking of, but a slightly different idea |
| 18:09 | <hober> | from http://www.intertwingly.net/blog/2004/09/23/Copy-and-Paste |
| 18:10 | <tantek> | hober do you have a citation for that? we've been stating that for quite some time in the microformats community re: minimizing impact of DRY violations etc. |
| 18:10 | <hober> | that link is the source |
| 18:13 | <tantek> | thanks hober - crossed messages :) |
| 18:16 | <mpt> | tantek, http://intertwingly.net/slides/2004/devcon/68.html apparently dates from a few months earlier than that post (but is missing any context) |
| 18:17 | <hober> | visible metadata principle also relates to 'people are lazy': http://www.well.com/~doctorow/metacrap.htm#2.2 |
| 18:18 | <mpt> | hm, no, actually, that slide is from a month later |
| 18:19 | <tantek> | note that per the Ruby postulate: "The accuracy of metadata is inversely proportional to the square of the distance between the data and the metadata." |
| 18:19 | <tantek> | if the metadata IS the data itself (c.f. microformats, visible data, DRY), then the distance is 0 |
| 18:19 | <hober> | indeed |
| 18:20 | <tantek> | inverse of 0 squared = infinity |
| 18:20 | <tantek> | ergo the most accurate metadata is to not have separate metadata at all and just use the data as metadata. |
| 18:23 | <hober> | right. given the 'people are lazy' thing, there's no reason to expect people to update content separated at all from the content they /have to/ update. |
| 18:29 | <Philip`> | tantek: In that case it's not metadata at all, it's just data :-p |
| 18:30 | gsnedders | thought the inverse of 0^2 was undefined :P |
| 18:32 | <Philip`> | Depends on how you define it |
| 18:33 | <Philip`> | If you define it to be defined, then it's not undefined |
| 18:44 | <tantek> | gsnedders - you can define it via Ruby's formula and limits. Calculus 101. |
| 18:57 | <gsnedders> | "Ruby's formula"? Googling it comes up with nothing useful… |
| 18:58 | <jgraham> | gsnedders: The one above... |
| 18:58 | <gsnedders> | Oh, just that |
| 18:59 | <jgraham> | (Although I would really be impressed o see evidence for the squared number rather than just an assertion :) ) |
| 19:00 | <jgraham> | (if people are going to take it seriously) |
| 19:08 | <tantek> | jgraham , indeed, which is specifically why I chose the word "hypotheses" for this blog post which also mentions inverse (geometric) proportionality: http://tantek.com/log/2007/02.html#d19t1813 |
| 19:26 | <takkaria> | gsnedders: uh, I'm not impressed with manchester uni |
| 19:26 | <takkaria> | gsnedders: depends what you're studying, mind |
| 19:26 | <takkaria> | gsnedders: life and physical sciences are excellent |
| 19:26 | <takkaria> | gsnedders: everything else is dross |
| 19:50 | <Wolfman2000> | Afternoon. I'd like confirmation please: there is officially no such concept as frames in HTML5, right? |
| 19:51 | <Hixie> | <iframe> is allowed in HTML5 |
| 19:51 | <Hixie> | <frame> and <frameset> is not |
| 19:51 | <Wolfman2000> | ...wonder why <iframe> is allowed when <object> should be able to replace it |
| 19:51 | <Hixie> | (browsers will continue to support them of course) |
| 19:51 | <Dashiva> | Iframe does more |
| 19:51 | <Wolfman2000> | more of what is my question |
| 19:51 | <Hixie> | <iframe> always has a nested browsing context; <object> might not |
| 19:52 | <Wolfman2000> | ...then I've got an interesting situation. |
| 19:55 | <Wolfman2000> | http://i42.tinypic.com/2w3bcz9.png I'm nearly done re-building/preserving a website for someone with HTML5. Originally, this page was in three frames: the top, the nav, and the rest. I'm trying to keep roughly the same layout as previously, with the disadvantage that once new content is loaded, the scroll bar goes back to the top. I don't want to break accessibility, but......well, I'm a little unsure where to proceed from he |
| 19:55 | <Wolfman2000> | re. |
| 19:57 | <Hixie> | which scrollbar? |
| 19:57 | <Wolfman2000> | the common scrollbar |
| 19:57 | <Wolfman2000> | basically: when you click on a kanji symbol, information shows up on top |
| 19:58 | <Hixie> | oh i see |
| 19:58 | <Hixie> | just use an iframe |
| 19:58 | <Hixie> | :-) |
| 19:58 | <Wolfman2000> | doesn't that also break accessibility? |
| 19:59 | <othermaciej> | I don't know of any reason iframes would break accessibility |
| 19:59 | <othermaciej> | my understanding is that screen readers know how to traverse into iframes |
| 20:00 | <Wolfman2000> | ...I am going to feel awkward using an iframe, but......I can understand the advantages. The question becomes...how do I adapt my framework properly? |
| 20:02 | Wolfman2000 | is using CodeIgniter, PHP, with semi-dynamic content stored in a database. Little unsure how to keep everything working at this point. |
| 20:28 | <dbaron> | Hixie, for what it's worth, I removed support for fallback background colors from Gecko (since the WG dropped the current approach to the feature), so the change to Acid2 is no longer needed |
| 20:30 | gsnedders | grumbles at Adobe again |
| 20:36 | <annevk42> | Hixie, the other day there was a comment in IRC regarding elements with a content model of "Text"; apparently it does not link to a definition and is only defined for "the HTML syntax" |
| 20:39 | <Hixie> | dbaron: excellent, thanks |
| 20:40 | <Hixie> | annevk42: what would it link to? |
| 20:40 | <Hixie> | annevk42: text is... text |
| 20:41 | <annevk42> | 1) apparently this is not clear to people; the concept of "Text" is also not explained in the section that explains content models 2) text is defined in the HTML writing section confusing matters |
| 20:42 | <Hixie> | annevk42: ah, fair enough. file a bug? |
| 20:42 | <Hixie> | pretty please :-) |
| 20:42 | <annevk42> | sugar on top? |
| 20:42 | <Hixie> | coarse cane sugar |
| 20:44 | <annevk42> | done :) |
| 20:46 | <Hixie> | thanks |
| 21:18 | <Philip`> | jgraham: I presume the squared is so that metadata isn't considered negatively accurate if it comes after the data instead of before |
| 22:22 | <remysharp> | please excuse the newbish question, but where does, if at all, querySelectorAll fit in with HTML5? From the specs I'm reading, it's not - is that correct? |
| 22:22 | <othermaciej> | remysharp: it's a separate spec, not part of HTML5 |
| 22:23 | <remysharp> | you don't have a link handy do you - or at least where I go ahuntin'? |
| 22:23 | <othermaciej> | remysharp: although, sometimes people use the term "HTML5" loosely to describe the whole current generation of client-side Web standards |
| 22:23 | <othermaciej> | http://www.w3.org/TR/selectors-api/ |
| 22:23 | <remysharp> | othermaciej: that's a lot of what I'm seeing |
| 22:23 | <remysharp> | perfect - thanks. |
| 22:24 | <remysharp> | on that topic - of the umbrella term of HTML5 - would you say that Web Sockets are part of HTML5 or not? |
| 22:24 | <remysharp> | given that they've been split out too? |
| 22:24 | <gsnedders> | They're still in the same source file even if the created docs are seperate :P |
| 22:25 | <remysharp> | I'm currently kinda putting it under the same umbrella, but I don't want to be misleading people I speak to |
| 22:25 | <othermaciej> | remysharp: WebSocket is officially a separate specificiation (actually, it will likely be more than one separate spec) |
| 22:25 | <othermaciej> | remysharp: it was once part of HTML5 proper |
| 22:25 | <othermaciej> | people do sometimes include it in the HTML5 blanket term |
| 22:26 | <remysharp> | do you think that's right / is there an opinion? |
| 22:26 | <othermaciej> | do I think what is right? |
| 22:26 | <remysharp> | putting it all under the blanket term |
| 22:26 | <gsnedders> | If you want to wrap it all up and keep it all warm, then it's right :P |
| 22:27 | <remysharp> | gsnedders: then that's what I'll do ;-) |
| 22:27 | <gsnedders> | remysharp: I, personally, would only call HTML 5 what is in the HTML 5 spec :) |
| 22:28 | <remysharp> | you're one of the peeps working on the spec (right?) |
| 22:28 | <gsnedders> | Hixie is the sole author/editor |
| 22:28 | <remysharp> | cripes - I didn't realise it was a one man job! |
| 22:28 | <remysharp> | fair enough. |
| 22:28 | <gsnedders> | He is employed to work on it full time though :) |
| 22:28 | <gsnedders> | (And has been for several years) |
| 22:29 | <remysharp> | I guess what I'm trying to check, is that I'm going to be speaking to a few people about the JS HTML5 APIs - |
| 22:29 | <remysharp> | I don't want to give the wrong information (or too much at least) |
| 22:29 | <remysharp> | and querySelectorAll is something I wanted to point out - |
| 22:29 | <remysharp> | but if it's not part of HTML5 - should people know or understand that? |
| 22:29 | <remysharp> | people == html/js authors |
| 22:29 | <remysharp> | (or the people I'm speaking to certainly!) |
| 22:30 | <gsnedders> | I'd just call it all stuff that's new to browsers :P |
| 22:30 | gavin_ | wonders whether he is related to remysharp |
| 22:30 | <remysharp> | My gut feeling is that I should just put it under the broad term of HTML5 - but I wanted to sound you chaps out |
| 22:31 | <beowulf> | i say they're seperate, otherwise people go looking for them in the html5 spec |
| 22:31 | <remysharp> | which is what I was doing in fact. |
| 22:32 | <remysharp> | erm ... |
| 22:32 | <remysharp> | that all said though - |
| 22:32 | <remysharp> | http://dev.w3.org/html5/ - why is web sockets sitting in there? |
| 22:33 | <remysharp> | is it just a legacy location? |
| 22:43 | <othermaciej> | remysharp: I would prefer to have a separate term for the latest generation of Web standards, one that could encompass the many recent important Web API specs, ES5, and perhaps some of the interesting new CSS3 modules |
| 22:44 | <othermaciej> | remysharp: but I think it would not be a good use of time to correct everyone using the term HTML5 in a broad way |
| 22:44 | <remysharp> | perhaps it's another comet or ajax term |
| 22:44 | <othermaciej> | we've discussed ideas like "Web 5.0" or just "Web 5" to use as a buzzword |
| 22:44 | <remysharp> | though html5 isn't that sexy I guess.... |
| 22:44 | <othermaciej> | but someone has to actually use it first |
| 22:44 | <remysharp> | or the acronym isn't... :) |
| 22:45 | <remysharp> | web 5! my client's heads would explode!!! |
| 22:45 | <remysharp> | they're seeing blogs and referring to them as "two dot oh" |
| 22:45 | <remysharp> | lol |
| 22:45 | remysharp | can I lol at my own statements...I think not... |
| 22:47 | <remysharp> | so, backpeddling a bit - am I right in saying the storage api is also *not* part of HTML5 |
| 22:50 | <othermaciej> | technically it's not part of the actual HTML5 spec |
| 22:50 | <othermaciej> | it was once |
| 22:50 | <othermaciej> | now it is separate |
| 22:50 | <othermaciej> | though it is still part of the blanket term many people use |
| 22:50 | <othermaciej> | for example Google |
| 22:51 | <remysharp> | okay, good - sorry, it's just I'm going through the specs trying to wrap my head around what is and isn't HTML5. |
| 22:51 | <remysharp> | My eyes are opening now - cheers! 8-) |
| 22:51 | remysharp | that wasn't supposed to be shades... |
| 22:52 | <remysharp> | one last thing before I leave you guys be for another day? I've got a drag and drop question - is that cool for this channel? |
| 22:57 | <gsnedders> | remysharp: Yeah, sure, whatever. It's really just a social channel, where most of the people involved in HTML 5 hangout and discuss stuff, including HTML 5 |
| 22:57 | <remysharp> | cheers - http://jsbin.com/ohuzi |
| 22:57 | <remysharp> | needs Firefox 3.5b4 - with DnD support |
| 22:57 | <remysharp> | and a console to see what's going on |
| 22:58 | <remysharp> | The 'drop' event is never triggering and I can't for the life of me work out why... |
| 23:00 | <remysharp> | This is the working example I based my code from - which works, but I don't understand why some events we return false, and why we *have* to setData on the transfer prop: http://jsbin.com/uzima |
| 23:05 | <tantek> | hober, regarding metadata, and inaccuracy, here is an illustrative photo: http://www.flickr.com/photos/tantek/430429916/ |
| 23:09 | <hober> | tantek: excellent |
| 23:10 | <remysharp> | okay, so if there's anyone left active in this room - if I've got an authoring question (which is hitting walls here), do I shoot it to help⊙wo - is that the right mailing list? Cheers |
| 23:12 | <Hixie> | that's the right list, whether you'll get an answer depends on if anyone knows the answer :-) |
| 23:13 | <Hixie> | for the drag and drop events i really need to write a better intro |
| 23:13 | <gsnedders> | Hixie: Yeah, learn to write! |
| 23:13 | gsnedders | still can't decide what to do at uni… |
| 23:13 | gsnedders | is quite often tempted by English |
| 23:13 | <remysharp> | It's not so much the intro - it just needs a quick demo of code |
| 23:14 | <remysharp> | that doesn't use inline handlers |
| 23:14 | <remysharp> | (based on the only one I could find) |
| 23:15 | <Hixie> | that's what i mean by intro :-) |
| 23:17 | <remysharp> | Ah, then yeah - that would help a ton :-) |
