| 00:31 | <Hixie> | dave_levin: webkit is correct on both mac and windows |
| 00:31 | <Hixie> | dave_levin: on mac, context menus appear on mouse down, on windows they appear on mouse up. |
| 00:34 | <dave_levin> | Hixie: Thx. That makes sense. |
| 00:35 | <dave_levin> | It also makes sense why mouseup occurs for FF on OSX after the contextmenu. |
| 00:35 | <dave_levin> | But a little weird with respect to consistency of events. |
| 00:35 | <Hixie> | yeah |
| 00:35 | <Hixie> | that's one reason to have the contextmenu event, so that the authors don't have to worry about it |
| 00:35 | <Hixie> | though of course many unnecessarily will |
| 00:36 | <dave_levin> | I think Chrome had some bug on this on Windows where the mouseup occurred after the context menu and it caused some problem. |
| 00:36 | <dave_levin> | Anyway, that makes sense now. Thx. |
| 00:38 | <Hixie> | np |
| 00:38 | <Hixie> | is anyone actually implementing contextmenu per spec yet? |
| 00:38 | <Hixie> | or does it just fire the event without actually doing anything other than showing the UA menu? |
| 00:43 | <dave_levin> | Hixie: Honestly, I wasn't trying to impl the context menu per spec. I was simply trying to resolve this issue about mouseup. |
| 00:43 | <wilhelm> | Hixie: See CORE-16665 and CORE-16567. |
| 00:43 | <dave_levin> | So I was looking at the spec to see what it had to say about it. |
| 00:43 | <Hixie> | ah |
| 00:45 | <dave_levin> | Sorry, I don't understand the CORE-16665. Is that a bug number? Where is the bug db? (I tried bugs.whatwg.org.) |
| 00:45 | <dave_levin> | Whoops, wrong conversation. My mistake. |
| 00:45 | <dave_levin> | nevermind me. |
| 00:46 | <wilhelm> | dave_levin: Opera bugs on oncontextmenu. Our BTS is not open, unfortunately. |
| 00:47 | <dave_levin> | wilhelm: Thx for the explanation. |
| 04:47 | <MikeSmith> | dglazkov: you around? |
| 04:54 | <dglazkov> | MikeSmith: partially |
| 04:55 | <MikeSmith> | dglazkov: was just curious if you know if there's any up-to-date list online of who the members of the Google Chrome team |
| 04:56 | <dglazkov> | mmm... The only thing I can think of is TechCrunch's list, but I don't know if it's kept up to date |
| 04:58 | <dglazkov> | are you working on your Christmas cards? ;) |
| 04:59 | <MikeSmith> | heh |
| 04:59 | <MikeSmith> | no, just curious, as usual.. I'm a natural nosy busybody :) |
| 09:09 | Hixie | chuckles at a roy's message to http-wg |
| 09:10 | <Hixie> | finally something he and i agree on -- html4 says bogus things |
| 09:10 | <Hixie> | (though i don't think we should just ignore it if we disagree with it, even if html2 says more convenient things) |
| 10:19 | <gsnedders> | Philip`: You are meeting me tonight at T19:00:00Z/PT5m |
| 10:19 | <gsnedders> | Oddly, the wifi on the train has only just started working |
| 10:19 | <gsnedders> | It was always dodgy north of Edinburgh before, but this time it simply wasn't working at all. |
| 10:20 | <gsnedders> | Hixie: I left a station this morning with semaphores! |
| 10:20 | <Hixie> | britain is living in the dark ages |
| 10:20 | <gsnedders> | Hixie: (per our discussion at TPAC :)) |
| 10:20 | <gsnedders> | Hixie: Now I've got to Edinburgh I've got electric signals hereon |
| 10:21 | annevk42 | hopes Philip` has an atomic clock |
| 10:22 | <gsnedders> | annevk42: I did give him a time range of five minutes |
| 10:22 | <Philip`> | gsnedders: What if I'm 300.5 seconds from the designated time? |
| 10:23 | <gsnedders> | Philip`: Then you are late |
| 10:23 | <Philip`> | gsnedders: That makes it sound like my fault, but it may be that my clock is a second slow |
| 10:24 | <gsnedders> | Philip`: I'm not entirely sure about what ISO 8601 says about precision |
| 10:25 | <Philip`> | Giving a precise time plus a precise range around that time is not enough to avoid the problem that clocks are imprecise |
| 10:25 | <gsnedders> | I do, of course, have a copy of ISO 8601:2004 with me |
| 10:25 | <gsnedders> | Philip`: Well, just have an atomic clock and you'll be fine |
| 10:27 | <Philip`> | There's the big grasshopper clock visible from King's, but apparently that's only accurate once every five minutes |
| 10:27 | <gsnedders> | But won't you be at the computer lab? |
| 10:28 | <Philip`> | At some point I will be |
| 10:28 | <Philip`> | but I will probably move |
| 10:29 | <gsnedders> | Philip`: But when? :) |
| 10:29 | <Philip`> | and once I arrive, assuming we're meeting at the place I assume we're meeting at, and I find I am late, I can point to the clock as a demonstration that time is more interesting when you don't worry so much about precision |
| 10:31 | <Philip`> | (http://en.wikipedia.org/wiki/Image:Corpus_clock_pol.jpg - it's a little bit peculiar) |
| 10:31 | <gsnedders> | Philip`: We are meeting by the post box outside King's |
| 10:31 | <Hixie> | i love that clock |
| 10:31 | <Hixie> | and especially the play on words |
| 10:32 | <gsnedders> | where in Corpus is that? |
| 10:33 | <gsnedders> | On a slightly different note, anyone got any protips for interviews? |
| 10:33 | <Philip`> | It's on the corner at http://maps.google.co.uk/?ie=UTF8&ll=52.203716,0.117607&spn=0.00021,0.00063&t=h&z=21 if I'm not mistaken |
| 10:34 | gsnedders | wonders whether the bandwidth here is enough for google maps |
| 10:34 | <gsnedders> | on the corner of King's parade? |
| 10:34 | <gsnedders> | ah, OK |
| 10:34 | <gsnedders> | there |
| 10:34 | <Hixie> | gsnedders: giving or receiving? |
| 10:34 | <gsnedders> | Hixie: receiving, for Cambridge uni |
| 10:34 | <Hixie> | ah, uh, good luck |
| 10:35 | <Hixie> | my main advice would be to avoid that side of it! |
| 10:35 | <gsnedders> | Hixie: Sadly, I'm all ready on the train on the way down for the two interviews :) |
| 10:36 | <Hixie> | well then my second piece of advice would be correct grammar :-P ("already") |
| 10:36 | <Philip`> | One tip I've heard is to not lie on your personal statement |
| 10:36 | <Philip`> | but it's probably too late for that now |
| 10:36 | <Hixie> | since if you are "all ready" then you don't need us :-P |
| 10:36 | gsnedders | has never quite learnt the distinction between all right and alright |
| 10:37 | <Philip`> | Hixie: I don't think it's really a play on words - the thing was called a grasshopper escapement because it looked a bit like a grasshopper, so the clock is just literalising the analogy |
| 10:38 | gsnedders | is going to have to go and have a look at it now |
| 10:38 | <gsnedders> | Philip`: Yeah, I probably ought to look at what the hell I put on my personal statement |
| 10:38 | <yecril71> | How would writing a special ID getter for tables help solve the problem of in-table identifier references? |
| 10:41 | <Hixie> | Philip`: still funny :-) |
| 10:42 | <yecril71> | I think the implementation should store a list of elements with a given ID internally. |
| 10:42 | <gpy> | hallo |
| 10:42 | <gpy> | whats up |
| 10:42 | <yecril71> | That would make some particularly insane pages slow, but in general it should not cause problems. |
| 10:42 | <gpy> | whats new |
| 10:44 | <Philip`> | yecril71: A lot of the web is particularly insane :-) |
| 10:47 | <Philip`> | gsnedders: I have been told that "There are a lot of general truths about the process you can helpful convey (e.g. that interviewers will try to see the best in candidates, not to confront them or to catch them out, that they are simply looking for constructive dialogue, etc.)" and also "you must not offer applicants any prohibited substances, alcoholic or otherwise" |
| 10:48 | <yecril71> | A lot of the Web is insane in an ordinary way. |
| 10:48 | <yecril71> | However, it cannot be particularly insane. |
| 10:48 | <yecril71> | "A lot" and "particularly" are mutually exclusive. |
| 10:49 | <Philip`> | yecril71: Not necessarily - each of the many pages is insane in its own particular unique way |
| 10:50 | <yecril71> | But you cannot generalize this statement that way. |
| 10:51 | <yecril71> | "Particular" does not have such an easy plural because it is a comparative attribute. |
| 10:53 | <jgraham> | gsnedders: tip: even if you don't know the answer say what you are thinking / how you would tackle the problem |
| 10:53 | gsnedders | returns from having wifi vanish |
| 10:53 | <annevk42> | g<space>problem |
| 10:53 | <yecril71> | I think, in the current wording, an OL must be inside a LI. |
| 10:54 | <annevk42> | yecril71, that's correct (and has been for ages) |
| 10:54 | <yecril71> | I fail to understand Aaron�s problem. |
| 10:55 | <gsnedders> | jgraham: I think my main problem will be not seizing up with nervousness |
| 10:56 | <aaronlev> | annevk42: thanks for the response |
| 10:56 | <yecril71> | OL is flow content but it does not follow that it can *contain* any flow content. |
| 10:56 | <aaronlev> | annevk42: i guess i needed more coffee, it was pretty obvious in the end |
| 10:56 | <yecril71> | Take a good shot then :-) |
| 10:56 | <aaronlev> | hehe |
| 10:57 | <aaronlev> | "Before you hit Send, remember that Anne is reading this list" |
| 10:58 | <aaronlev> | Content model: flow content subset |
| 10:58 | <aaronlev> | but yeah, it's correct as it is |
| 11:00 | <Philip`> | jgraham: When I tried doing that, I was thinking in completely the wrong direction and got horribly lost in irrelevant messy details :-p |
| 11:00 | <Philip`> | but I suppose that's still better than sitting there and saying "uh" |
| 11:01 | <jgraham> | Philip`: You got in right :) |
| 11:02 | <annevk42> | aaronlev, np |
| 11:02 | <aaronlev> | annevk42: shouldn't contenteditable="true" make an element labelable? |
| 11:04 | <annevk42> | WYSIWYG editing seems pretty separate from form input to me |
| 11:04 | <annevk42> | at this point WYSIWYG editing is also mostly experimental, that is, most of the work still has to be implemented in script |
| 11:05 | <jgraham> | gsnedders: Tip 2: If you have some vauge idea of what your interviewers are interested in and some knowledge about it, it doesn't hurt to try to get them talking about it |
| 11:06 | <jgraham> | (or, in general, try to sound interested) |
| 11:07 | <gsnedders> | jgraham: Do you per-chance know Prof. Clarke? |
| 11:07 | <jgraham> | No |
| 11:08 | <jgraham> | (one way to sound interested is to ask them what thy work on) |
| 11:08 | jgraham | disclaims any bad outcomes from these tips |
| 11:10 | <Philip`> | jgraham: On the other hand, that might just end up wasting interview time that would be much more usefully spent talking about the interviewee rather than about the interviewer |
| 11:10 | annevk42 | doesn't have much experience with such interviews |
| 11:11 | <Philip`> | (From what I can tell, the interviewers do actually care that you're competent, and not just a nice person who they'd be happy to spend the next three years meeting twice a term, because they know that insufficiently competent people won't survive the course and it's detrimental to everybody) |
| 11:13 | <aaronlev> | annevk42: but it something is contenteditable then you might need to label it |
| 11:14 | <aaronlev> | annevk42: as far as the user is concerned it acts like a form control, and it often needs a label |
| 11:14 | <jgraham> | Philip`: Often they ask at some point if you have any questions. I'm not really suggesting going in and just asking them stuff |
| 11:14 | <annevk42> | aaronlev, really? I haven't really seen that happening in the wild, but maybe I'm missing something |
| 11:15 | <jgraham> | I assume they will go through whatever questions they have regardless |
| 11:15 | <aaronlev> | annevk42: i have |
| 11:15 | <aaronlev> | annevk42: for example, some vendors use a rich text field to create something like an email "To: " entry field |
| 11:15 | <annevk42> | aaronlev, we currently also do not have disabled, readonly, etc. for editable regions; as I said, it's mostly experimental |
| 11:15 | <aaronlev> | it will have autocomplete and use different colors or an underline to show correct/incorrect emaiol addresses |
| 11:16 | <aaronlev> | annevk42: but if it's experimental why support it? |
| 11:16 | <annevk42> | aaronlev, they should be using <input type=email multiple> there at some point :) |
| 11:16 | <annevk42> | aaronlev, to allow for innovation |
| 11:16 | <aaronlev> | annevk42: that's just one case of a general example you know that :P |
| 11:16 | <annevk42> | aaronlev, so we can see what people would do if rich text editing works |
| 11:16 | <annevk42> | aaronlev, and then we'd simplify the most common cases, such as email widgets :) |
| 11:16 | <aaronlev> | you can't have type="foo" for every kind of validating input someone might want |
| 11:17 | <annevk42> | indeed, that's why we ECMAScript and plenty of APIs |
| 11:17 | <aaronlev> | i don't see what you gain by not allowing contenteditable items to be disabled or have a label |
| 11:17 | <annevk42> | the gain is spec, QA, and implementor resources for other features |
| 11:17 | <gsnedders> | jgraham: She some theoretical astrophysicist. I have an interview with her tomorrow morning. |
| 11:17 | <annevk42> | that are more important |
| 11:18 | <aaronlev> | annevk42: and a loss to consistency from the user's poiint of view |
| 11:18 | <annevk42> | contenteditable mostly acts not like a typical form control at all though, so I'm not sure there's much consistency to start with |
| 11:19 | <aaronlev> | it acts like a form cotnrol in that you can focus it, enter data, and leave it |
| 11:19 | <aaronlev> | so i want to know what data i'm entering |
| 11:19 | <annevk42> | contenteditable is very much for editing existing data, not entering new data |
| 11:19 | <aaronlev> | annevk42: prove it :) |
| 11:20 | <aaronlev> | that's a shaky argument |
| 11:20 | <aaronlev> | i don't see any way you can back that up |
| 11:20 | Philip` | has written some code that uses <span contenteditable> as an automatically-resizing borderless <input> |
| 11:20 | <aaronlev> | when i compose a new email or doc i'm entering new data |
| 11:21 | <annevk42> | aaronlev, true |
| 11:21 | <jgraham> | gsnedders: Oh, Cathy Clarke |
| 11:21 | <jgraham> | She's really nice |
| 11:21 | <gsnedders> | jgraham: yeah, her |
| 11:21 | <jgraham> | Don't worry about what she works on |
| 11:21 | <jgraham> | obviously ;) |
| 11:21 | <Philip`> | gsnedders: Challenge her to refute the Time Cube |
| 11:22 | <annevk42> | aaronlev, any case, it seems to soon to assume it will be used just like form controls |
| 11:22 | <gsnedders> | jgraham: Why? Something ridiculously complex? :) |
| 11:22 | <annevk42> | aaronlev, been just a year or so that all browsers actually support it |
| 11:23 | <aaronlev> | annevk42: also, the email validation thing -- how does it know which emails are in the user's contact list? |
| 11:23 | <jgraham> | gsnedders: Aren't you applying for 50% comp-sci? I suppose there is no harm in asking about it |
| 11:24 | <aaronlev> | let's say it's trying to restrict |
| 11:24 | <aaronlev> | i dunno, i guess i'll wait to come up with good examples to show you what i mean |
| 11:24 | <jgraham> | She works on planet formation |
| 11:24 | <annevk42> | aaronlev, <input type=email multiple list=contacts><datalist id=contacts><option>annevk⊙oc<option>...</datalist> |
| 11:24 | <jgraham> | gsnedders: ^ |
| 11:24 | hendry | wonders how authors should embed ads in the future. sandboxed iframes? |
| 11:24 | <aaronlev> | annevk42: yikes |
| 11:24 | <aaronlev> | client side? |
| 11:25 | <gsnedders> | jgraham: 50% comp.sci., 25% maths and nat.sci. each — with a possible intention of switching to physical nat.sci. |
| 11:25 | <annevk42> | aaronlev, you can populate the markup through script, of course |
| 11:25 | <gsnedders> | jgraham: ah, interesting |
| 11:25 | <aaronlev> | annevk42: ok, i'll send better examples when i get them |
| 11:25 | <aaronlev> | that seems like a good challenge |
| 11:25 | <annevk42> | aaronlev, we had a feature <datalist data=foobar.xml></datalist> but that's been out for now |
| 11:25 | <annevk42> | s/been // |
| 11:25 | <annevk42> | (Opera still has it, though) |
| 11:26 | <annevk42> | aaronlev, I don't how it's more than a challange than implementing a complete email widget, that's much harder :) |
| 11:26 | <annevk42> | aaronlev, just do a simple fetch with XHR to some data on the server, and then insert a bunch of <option> elements, rather trivial I'd say |
| 11:26 | <aaronlev> | annevk42: email is just 1 example where you might need a label |
| 11:26 | <annevk42> | sure |
| 11:26 | <aaronlev> | annevk42: so i plan to send more cases to show there's no point in trying to stretch the spec over every use case |
| 11:27 | <aaronlev> | just let people insert a label, that's far easier than trying to cover everything an author wants to do |
| 11:27 | <jgraham> | gsnedders: If you say that you might switch to Physics I guess it helps to sound interested in that too :) |
| 11:27 | <annevk42> | aaronlev, we won't try to do that either |
| 11:27 | <gsnedders> | jgraham: I already have :) |
| 11:27 | <annevk42> | aaronlev, people can create custom labels with their custom controls; using ARIA to associate them together |
| 11:28 | <aaronlev> | yes |
| 11:28 | <aaronlev> | but I thought the idea was that the main uses of ARIA ultimately are superceded by HTML |
| 11:28 | <aaronlev> | so we take what we learn from it and improve HTML |
| 11:31 | <aaronlev> | ok, we will see what happens |
| 11:40 | <annevk42> | i'm sort of guessing we'll always need something like ARIA to fill certain gaps |
| 11:40 | <annevk42> | it would be nice if it looked different, but we're probably past that now |
| 11:40 | <annevk42> | s/sort of// |
| 11:43 | <annevk42> | http://lists.w3.org/Archives/Public/www-validator/2008Dec/0019.html should I have filed a bug report somewhere or is there one already? |
| 11:43 | annevk42 | thought about that too late |
| 11:45 | <jgraham> | gsnedders: Tip 3: Try to relax and have fun. They are going to ask you some interesting questions so try to enjoy thinking about them. The absolute worst that can happen is that you end up going to Edinburgh which is really not such a bad thing |
| 11:45 | <gsnedders> | jgraham: But I've already left Edinburgh! |
| 11:46 | <jgraham> | The interviews are not perfect; there were plenty of bright people doing a PhD with me who got rejected as undergrads and some not-so-bright people who got in and I supervised :) |
| 11:47 | <jgraham> | s/rejected/rejected from Cambridge/ |
| 11:48 | Philip` | isn't quite sure how interviews can actually be fun |
| 11:48 | <gsnedders> | Philip`: There is little prize for guessing what, at least in part, I'll be asking you about tonight :) |
| 11:49 | <jgraham> | gsnedders: Tip 4: Try to estimnate how many piano tuners there are in Chicago. It is not impossible you will be asked some similar type of order-of-magnitude question so it helps to have thought through one before |
| 11:49 | <Philip`> | For the same amount of effort as attending interviews, you could stay at home and play computer games |
| 11:49 | <jgraham> | gsnedders: Population of chicago is about 3 million |
| 11:49 | <Philip`> | gsnedders: Do you have any interviews today, or are they all tomorrow? |
| 11:49 | <gsnedders> | jgraham: maybe around 100? |
| 11:49 | <gsnedders> | Philip`: all tomorrow |
| 11:50 | <Philip`> | gsnedders: 0 marks - needs to show working :-p |
| 11:50 | annevk42 | tries upgarding to Ubuntu 8.10 |
| 11:50 | <jgraham> | gsnedders: IIRC it is rather smaller than that but as Philip` says the point is how you got there not what your answer is |
| 11:50 | <annevk42> | someone else with a thinkpad had it working, hopefully I'm good |
| 11:51 | jgraham | finds an estimate that gets 150 |
| 11:51 | <gsnedders> | jgraham: IRC isn't in some ways the best medium for thinking allowed |
| 11:52 | <gsnedders> | jgraham: *aloud |
| 11:52 | <gsnedders> | jgraham: You end up being a bit of a flooder :) |
| 11:53 | <Philip`> | gsnedders: You could write multiple sentences on a single line, and then it wouldn't take more than a line to explain your thinking :-) |
| 11:54 | <jgraham> | gsnedders: Sure. Just as long as you /had/ a logical method |
| 11:54 | <gsnedders> | Hixie: Also only in Britain: I'm on a train for around 500 miles, running on diesel trains the entire way, under wires for around 450 miles :) |
| 11:55 | <gsnedders> | Philip`: Meh. |
| 11:55 | <gsnedders> | Oh, after 3 hours on this train we've finally had it announced that there is no trolley service |
| 11:56 | <jgraham> | aaronlev: Allowing @headers to point at the headers in a nested table seems bad |
| 11:57 | <jgraham> | aaronlev: Have you tried inverting the header association algorithm and found it to be difficult or are you just worried about the theoretical difficulty? |
| 11:58 | gsnedders | is rather sleepy |
| 12:18 | <gsnedders> | Is it bad being on Facebook on a train> |
| 12:28 | <aaronlev> | jgraham: i think every browser dev will invert it differently |
| 12:28 | <aaronlev> | it's not really useful in the current form, so i suggest changing it so it is |
| 12:28 | <aaronlev> | then we can see if that is correct |
| 12:51 | <MikeSmith> | gsnedders: did you see that the latest Mr. Last Week has some videos of you singing "The American Ruse" and "Looking at You"? |
| 12:52 | <gsnedders> | MikeSmith: Me singing it? |
| 12:52 | <MikeSmith> | yeah, man, you |
| 12:52 | <MikeSmith> | at least it looks like you |
| 12:52 | gsnedders | puts in earphones and wonders if it will work |
| 12:53 | <gsnedders> | No, nowhere near enough bandwidth |
| 12:53 | gsnedders | is listening to Decomposing Composers by Monty Python from Monty Python Sings |
| 12:56 | gsnedders | yawns and puts on Christian metal |
| 12:57 | <gsnedders> | Is it bad this 10-or-so year old girl reminds me of a certain girl at school every time she walks past? |
| 12:59 | <Philip`> | That depends on whether you are imagining doing something illegal to the girl |
| 13:00 | <gsnedders> | No, only something legal to the one she reminds me of |
| 13:01 | <gsnedders> | (And nothing that would be illegal to do to her) |
| 13:02 | <annevk42> | yay, done |
| 13:02 | <annevk42> | no network manager in the taskbar though |
| 13:02 | gsnedders | notes some of the things he says would be rather odd to see as the first line |
| 13:04 | gsnedders | wonders what the people in college are going to be like |
| 13:13 | <annevk42> | is lastweekinhtml5 ridiculing standardssuck.org? fun |
| 13:13 | <annevk42> | more viewers! |
| 13:15 | <gsnedders> | damnit! I should have ridiculed standardssuck.org on standardssuck.org! |
| 13:42 | <gsnedders> | Philip`: remember! :P |
| 13:42 | gsnedders | runs off |
| 13:45 | <Philip`> | Please nobody say anything for the next five hours, so that reminder doesn't scroll off the screen |
| 14:16 | Lachy | wonders what Philip` needs to remember? |
| 14:26 | <Philip`> | Lachy: I'm not quite sure |
| 14:26 | <Philip`> | Lachy: I think I'm probably meant to be meeting gsnedders or something |
| 14:38 | <JohnResig> | Lachy: hey, are you around? |
| 14:40 | <JohnResig> | Lachy: I'm trying to figure out what the most recent proposal is for contextual qSA - ":context ~ div"? |
| 14:41 | <Lachy> | JohnResig, hi |
| 14:41 | <Lachy> | sibling selectors won't work with :context (or :scope) with the current methods, since only descendant nodes can be matched |
| 14:42 | <JohnResig> | Lachy: so currently there is no proposal in which it'll work? |
| 14:42 | <Lachy> | that's something that will probably need to be addressed in the next version |
| 14:42 | <JohnResig> | Lachy: ok - we're discussing it again because the JS libraries are standardizing on a unified selector engine |
| 14:43 | <Lachy> | IIRC, this was discussed on the mailing list once before. I think the solution will have to be something like document.someNewSelectorMethod(":context~div", contextNode); |
| 14:43 | <annevk42> | might have been interesting to allow it in such a way if :scope had been there from the start |
| 14:43 | <annevk42> | Lachy, why not have the method on contextNode? |
| 14:44 | <Lachy> | annevk42, that may work |
| 14:44 | <annevk42> | seems more DOM-like |
| 14:48 | <Lachy> | annevk42, it might make things complicated though, because if div.foo(":context+div") matches elements outside of the context node, then does div.foo("div") do so as well? |
| 14:49 | <Lachy> | i.e. is the scope of elements limited for the method in any way? |
| 14:49 | <annevk42> | nope |
| 14:50 | <annevk42> | though i'm not sure if it's worth it given that the only use case is :context ~ and :context + |
| 14:50 | <Lachy> | ok, consider <body><div><p/></div><p/></body> What do each of these match? div.foo("p"); div.foo(":context+p"); div.foo("body");? |
| 14:51 | <annevk42> | p1,p2;p2;body |
| 14:52 | <Lachy> | that seems a little unintuitive |
| 14:52 | <Lachy> | well, at least, it's the complete opposite of how the existing methods work |
| 14:53 | <Lachy> | if we used the approach I suggested, then it's possible we could just extend the existing methods with an additional parameter, like document.querySelector(":context+div", contextNode); |
| 14:54 | <Lachy> | (though we'd need to consider compat issues before doing that) |
| 14:55 | <annevk42> | i suppose, anyway, are there compelling use cases? |
| 14:55 | <Lachy> | JohnResig, do you know if any JS libraries implement functionality similar to the requested Element.matchesSelector() proposal? |
| 14:57 | <JohnResig> | Lachy: yes |
| 14:57 | <JohnResig> | Lachy: many, if not all of them |
| 14:57 | <Lachy> | annevk42, I can't remember if already had use cases presented for it, or whether we needed to investigate them |
| 14:58 | <JohnResig> | Lachy: it's going to be in the new unified library we're working on |
| 14:58 | <Lachy> | really? I checked the JQuery docs and couldn't find any method that seemed to that. Which method is it? |
| 14:59 | <JohnResig> | Lachy: $(element).is("selector") |
| 14:59 | <Lachy> | where is that in the JQuery docs? |
| 15:00 | <JohnResig> | Lachy: http://docs.jquery.com/Traversing/is#expr |
| 15:01 | <Lachy> | thanks |
| 15:08 | <Lachy> | JohnResig, are your discussions with the other JS library devs public, or is there some documentation of what's going into the unified engine? |
| 15:08 | <JohnResig> | Lachy: the engine: http://github.com/jeresig/sizzle the discussion: http://groups.google.com/group/sizzlejs |
| 15:14 | <Lachy> | thanks |
| 17:33 | <jgraham> | Philip`: You may have had to remember something but its hard to be sure |
| 17:51 | <Philip`> | jgraham: Hmm, I don't recall anything important |
| 20:13 | <virtuelv> | Ok, I think I understand the reasoning behind unicode decomposition |
| 20:14 | <virtuelv> | but that stuff is giving me a severe headache |
| 20:15 | <Dashiva> | on macos, or in general? |
