| 06:16 | <annevk> | Domenic: heh |
| 06:16 | <annevk> | GPH-Hickory: did you review the CSS @charset tests? I recall some of them were bogus, not sure if those got fixed |
| 06:17 | <GPHemsley> | annevk: They seemed like they may be from what I recall about encodings, but I didn't check against the Encoding spec. |
| 06:17 | <GPHemsley> | Especially the Shift_JIS one |
| 09:07 | <hemanth> | MeOw |
| 10:07 | <hsivonen> | Oh great. Eric J. Bowman asked an attorney who "doesn't grok" the subject matter of the question |
| 10:10 | <SteveF_> | FYI new HTML data set on http://webdevdata.org/ |
| 10:32 | <annevk> | hsivonen: that guy |
| 10:35 | <hemanth> | annevk: meow |
| 10:40 | <annevk> | hemanth: ? |
| 10:41 | <hemanth> | annevk: that's a way to greeting :) |
| 10:41 | <hemanth> | *of |
| 10:41 | <annevk> | I see, well good morning to you then :-) |
| 10:41 | <hemanth> | annevk: ^_^ I'm still digging for internal states :/ |
| 10:51 | <JakeA> | Gutted I missed the conversion about offence a couple of days back. Can we have it again so I can take part? |
| 10:52 | <jgraham> | JakeA: That depends. Were you offended that you missed out? |
| 10:57 | <JakeA> | jgraham: I have a moderate discomfort about it that I believe others should cater for |
| 11:06 | <jgraham> | Ah, what is civilisation if not the avoidance of moderate discomfort in others. We must reopen the debate post-haste. |
| 11:06 | <JakeA> | :D |
| 11:53 | <JakeA> | annevk: I'm struggling with https://github.com/slightlyoff/ServiceWorker/issues/588#issuecomment-67662246 - do we have other APIs that are 'snapshots' but also have methods to change state? |
| 11:54 | <annevk> | JakeA: it seems fine to have a method on a snapshot, but it shouldn't affect the snapshot itself |
| 11:54 | <annevk> | JakeA: so returning a new client as mounir suggested makes sense |
| 11:55 | <JakeA> | annevk: from a usability point of view, I do think people will look for methods on the snapshot |
| 11:55 | <annevk> | JakeA: yeah I don't see why we wouldn't have methods that essentially forward |
| 11:57 | <JakeA> | Thanks! |
| 12:58 | <GPHemsley> | hsivonen: It doesn't appear that any action was ever taken on this: http://lists.w3.org/Archives/Public/public-css-testsuite/2012Dec/0000.html |
| 12:58 | <GPHemsley> | Might want to follow up |
| 13:03 | <Ms2ger> | GPHemsley, you almost sound surprised |
| 15:44 | <Domenic> | hsivonen: I was really tempted to write a reply whose only content was: "> To elaborate, my attorney doesn't grok this stuff \n\n I thought that was the case." |
| 16:24 | <MikeSmith> | Domenic: heh |
| 16:25 | <MikeSmith> | that guy keeps outdoing himself |
| 16:26 | <MikeSmith> | Taking the absurdity to its full potential |
| 16:27 | <rubys> | Dominic: I did manage to get IE11 preview installed; saw no change to how it handles URLs. See http://intertwingly.net/blog/2015/01/08/Ununzippable-Modern-IE#c1420802707 for the gory details |
| 16:27 | <jgraham> | Isn't that the W3C motto? |
| 16:27 | rubys | notes that jgraham equates web with absurity |
| 16:27 | <rubys> | absurdity* |
| 16:28 | <MikeSmith> | I think Jeff confused that with the W3C motto about backasswardsness |
| 16:29 | <MikeSmith> | s/Jeff/jgraham |
| 16:30 | MikeSmith | tries to come up with some other typos to dig his hole further |
| 16:30 | rubys | notes that MikeSmith confused jgraham with jeff jaffe |
| 16:30 | <MikeSmith> | nope |
| 16:30 | <rubys> | :-P |
| 16:30 | <MikeSmith> | Jeff Bridges |
| 16:31 | <MikeSmith> | either the Dude or as in True Grit |
| 16:38 | hemanth | says hello to jspm by fetching random XKCD comics https://github.com/hemanth/hello-jspm Influenced by @glenmaddern |
| 16:39 | <hemanth> | Domenic, meow, I'm still digging those internal states...:| |
| 16:43 | <caitp> | you're like the WHATWG's pet cat it's awesome |
| 16:45 | <jgraham> | It's pretty weird. "meow" is used to accuse people of being nasty ("catty") so it doesn't sit well as a greeting |
| 16:46 | <Ms2ger> | Is it? |
| 16:46 | <Ms2ger> | TIM |
| 16:46 | <Ms2ger> | Today I Meowed? |
| 16:46 | <Ms2ger> | TIL* |
| 16:46 | <jgraham> | Usually in speech rather than print I guess, but yes |
| 16:49 | <Domenic> | rubys: hmm, what's your IE version? I could swear I got different results, on the page that displayed the components |
| 16:50 | <rubys> | Dominic: look at the top of https://github.com/webspecs/url/commit/68210260199db858c9eb349d643916b0d5d564f9 |
| 16:53 | <rubys> | Dominic: if you want to see what your version produces for all of the tests, go to http://intertwingly.net/projects/pegurl/urltest |
| 16:53 | <rubys> | if you want me to update my results based on a different version, simply provide me with the hex ticket number that page produces |
| 16:57 | <Domenic> | rubys: what is the version shown in "About Internet Explorer" though; it's not part of the UA string. Mine is 11.0.9879.0 |
| 16:57 | <Domenic> | rubys: ticket http://intertwingly.net/projects/pegurl/urltest |
| 16:57 | <Domenic> | er |
| 16:58 | <Domenic> | wow, copying and pasting out of this vm is not working |
| 16:58 | <rubys> | b795e48410c0d1ba0c4764e9c30f4e1e |
| 16:58 | <Domenic> | yes, that one :P |
| 16:59 | <rubys> | just a sec, I'll give you my ie version |
| 16:59 | <rubys> | 11.0.9841.0 |
| 16:59 | <rubys> | I'm downloading a new VM. They claim to have fixed the unzip problem. |
| 17:05 | <hemanth> | ^_^ |
| 17:08 | <hemanth> | is there a way to intervene promise, say we have fetch(url).then(); I need to show a spinner till the promise resolves...fetch(url).tillThen().then().catch()...so on... |
| 17:09 | <Ms2ger> | showSpinner(); fetch(url).then(hideSpinner) |
| 17:11 | <hemanth> | Ms2ger, yeah, currently doing a similar thinge, tillThen() sound dumb ? |
| 17:11 | <Ms2ger> | I don't see the point |
| 17:11 | <hemanth> | I really liked the way protractor's expect is modified to expect a promise, but still the user need not really resolve it with then... |
| 17:14 | <rubys> | Dominic: are you *SURE* that was with IE? The User agent looks awfully suspicious. |
| 17:14 | <rubys> | See: https://github.com/webspecs/url/commit/ed1a053212aae906a4edcce94d7c1681b9b001bd |
| 17:25 | hemanth | paws at caitp |
| 17:32 | <hemanth> | pawing is saying hey, in cat; meow is saying that i'm here, clawing is when you are angry...( gosh...why i'm talking/typing so much... alias hemanth='hemanth -s') |
| 17:37 | <caitp> | it's all good man |
| 17:38 | astearns | on the internet, no one knows you're a dog who identifies as a cat |
| 17:40 | <hemanth> | astearns, I'm hemanth's dog, I have drunk the magic portion and turned into a cat, will be a dog soon, by the time the cat would be back from it's walk |
| 17:42 | <Domenic> | rubys: yes, they have purposefully obfuscated their user agent to avoid detection. |
| 17:43 | <annevk> | www-tag continues to amaze in uniting all the trolls |
| 17:43 | <Domenic> | rubys: "Edge/12.0" is the only remaining IE detection |
| 17:44 | <rubys> | Dominic: cool. OK, I've updated https://url.spec.whatwg.org/interop/test-results/ |
| 17:44 | <rubys> | there is a *lot* less "ie" on that page |
| 17:45 | <Domenic> | :D |
| 17:46 | <Domenic> | I am curious about the rows that have no entries, e.g. (when I select "all current user agents") "http://f:0/c"; |
| 17:46 | <Domenic> | all current browsers, rather |
| 17:47 | <rubys> | I'm confused. You are curious about results for which there are no current browsers that disagree? |
| 17:48 | <Domenic> | i thought those were filtered out of the table, I guess, but now that I think about it I don't know where I got that impression. |
| 17:49 | <Domenic> | cool, all is well |
| 17:49 | <rubys> | adding more filters is only a matter of code. :-) |
| 17:50 | <Domenic> | i also love the infrastructure you've set up where i can provide results by navigating to a page; it worked really nicely here |
| 17:51 | <Domenic> | https://url.spec.whatwg.org/interop/test-results/c6c0953a53?select=current is a fun one. you can just see all four engineers making different decisions. |
| 17:53 | <rubys> | dominic: a related bug on that one: https://www.w3.org/Bugs/Public/show_bug.cgi?id=26446 |
| 17:54 | <Domenic> | wow, the fact that Firefox fails above 2^31 makes things even more fun. |
| 17:55 | <rubys> | now if only we could get the four engineers to participate in a discussion as to what the correct results should be |
| 17:56 | <caitp> | in TCP, the port field is a 16 bit value, so there isn't much reason to support infinitely high ones is there? |
| 17:57 | <rubys> | caitp: another way to phrase the question: "what should impls do for unsupported values"? |
| 18:00 | <caitp> | port = port >= 0x10000 ? 80 : port |
| 18:00 | <caitp> | idk |
| 18:01 | <rubys> | that would indeed be a fifth option. Here's four others: https://url.spec.whatwg.org/interop/test-results/c6c0953a53?select=current |
| 18:02 | <rubys> | At the moment, the spec agrees with Firefox. I personally like Chrome's choice here better. |
| 18:04 | <caitp> | i wasn't really being serious about falling back on 80, 0 seems pretty reasonable |
| 18:06 | <rubys> | I care that we get people to converge on a single answer, I don't particularly care what that answer is. |
| 18:08 | <Domenic> | rubys: I think the spec doesn't agree with Firefox for values > 2^31, according to bug 26446 -_- |
| 18:09 | <Domenic> | I like the uniformity of having everything be strings, but it'd probably be easier to get agreement on some restriction to <= 0x10000. Still need to figure out what the behavior is outside that range though. |
| 18:09 | <rubys> | The bug indeed does suggest that the spec shouldn't disadvantage implementations that store port numbers as integers |
| 18:13 | <caitp> | < 0x10000, not <= :p |
| 22:12 | <Domenic> | Is it possible when parsing a HTML page in browsers to get instances of Element that are not instances of HTMLElement? |
| 22:14 | <jgraham> | Yes: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3360 |
| 22:14 | <jgraham> | (svg, mathml) |
| 22:14 | <Domenic> | I meant, own-instances. Just Element, not any subclass. |
| 22:15 | <Domenic> | Looks like createElementNS can do it |
| 22:16 | <Domenic> | document.createElementNS("http://example.com";, "baz:foo").constructor === Element |
| 22:16 | <jgraham> | Right, but that's not "when parsing a HTML page" |
| 22:16 | <Domenic> | ah yeah I guess I forgot the second part of my question |
| 22:17 | <jgraham> | I don't think the parser can create things that are bare Elements |
| 22:18 | <jgraham> | Oooh, maybe <math> doesn't have a special subclass |
| 22:23 | <Hixie> | you can trick the parser into inserting elements in neither svg, mathml, or html namespaces |
| 22:23 | <Hixie> | in which case it'll be from Element |
| 22:23 | <Hixie> | also, i don't think any of hte browsers implement MathML's DOM, so they're all just Element too |
| 22:26 | <jgraham> | with <svg><foo> or something more subtle? |
| 22:27 | <jgraham> | Not that at least |
| 22:27 | <Hixie> | maybe innerHTML on a non-HTML non-SVG non-MathML element? |
| 22:27 | <Hixie> | i forget teh details |
| 22:27 | <jgraham> | Oh innerHTML, I hadn't thought of that |
| 22:27 | <jgraham> | I guess that still counts as the parser |
| 22:35 | <gsnedders> | jgraham: that just creates an svg:foo element |