| 00:52 | <erlehmann> | is there a good way to discourage someone to invent a new scheme for a use case that can be appropriately handled by data URIs? |
| 00:53 | <zewt> | i suggest http://image.made-in-china.com/2f0j00tMsaevNhAPcl/Rubber-Mallet-Wood-Hande-1lb-TSH0902-.jpg |
| 02:01 | <abarth> | Hixie: do you have a couple minutes for an HTML parser question? |
| 02:12 | <erlehmann> | abarth, JUST ASK |
| 02:12 | <abarth> | yeah? |
| 02:12 | <abarth> | i'm writing it up in an email to whatwg |
| 02:16 | <erlehmann> | abarth, YOU ARE A VERY CLEVER MAN |
| 02:16 | <abarth> | :) |
| 02:16 | <zewt> | caffeine |
| 04:14 | <Hixie> | abarth: yeah, that's on my radar |
| 04:14 | <Hixie> | abarth: see if you can get henri to comment :-) |
| 04:14 | <abarth> | Hixie: thanks |
| 04:15 | <abarth> | what's the point of going through the HTML states for the text integration points? |
| 04:15 | <abarth> | to get the formatting elements reconstructed? |
| 04:17 | <Hixie> | off the top of my head i can't recall, but probably |
| 04:17 | <Hixie> | also to enter things like select mode for <select>s in cells, etc |
| 04:17 | <abarth> | for the time being, i'm probably going to implement what flangagan suggests in http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-October/033542.html |
| 04:18 | <abarth> | i'm trying to update the webkit impl to match the changes to the spec that happened after we did the first iteration |
| 04:18 | <abarth> | and this would be a regression |
| 04:18 | <Hixie> | i can't tell off the top of my head if that's the right thing, but i can try to prioritise it tomorrow if you like |
| 04:18 | <abarth> | if that's not too much trouble, it would be helpful |
| 04:19 | <abarth> | i'd rather fix it the right way :) |
| 04:19 | <Hixie> | abarth: yeah |
| 04:19 | <Hixie> | abarth: can you work out what gecko does? |
| 04:19 | <abarth> | i'll investigate and email the list if I figure it out |
| 04:19 | <Hixie> | abarth: (i haven't tried, but that info would be useful. i'll try to reverse-engineer it if neither you nor hsivonen comment before i get to it) |
| 04:19 | <Hixie> | cool |
| 04:20 | <abarth> | wow, they're still auto generating their parser from java |
| 04:27 | <abarth> | so, the control flow in the moz parser is really far from the spec |
| 04:27 | <abarth> | they have a bunch of branches all over the place for inforiegncontent |
| 04:27 | <abarth> | it looks like they call flushCharacters unconditionally |
| 04:27 | <Hixie> | i thought the spec had been changed specifially to be more like gecko :-) |
| 04:27 | <Hixie> | they might have the branches backwards from what the spec does, but i imagine the basic idea is the same |
| 04:28 | <Hixie> | or rather, inside out |
| 04:28 | <Hixie> | i.e. instead of if (foo) { switch } else { switch }, have switch { case: if (foo) else; case: if (foo) else; } |
| 04:28 | <abarth> | http://mxr.mozilla.org/mozilla-central/source/parser/html/javasrc/TreeBuilder.java#3161 |
| 04:29 | <abarth> | as an example |
| 04:29 | <abarth> | the if (isInForeign()) comes after the flushCharacters call |
| 04:30 | <abarth> | or, consider the passage that starts http://mxr.mozilla.org/mozilla-central/source/parser/html/javasrc/TreeBuilder.java#1186 |
| 04:30 | <abarth> | the if (!isInForeignButNotHtmlIntegrationPoint()) has been pushed into the states |
| 04:34 | <abarth> | yeah, so we just need to flush right away |
| 04:34 | <abarth> | rather than holding onto the pending characters until we get the </table> |
| 04:34 | <abarth> | let me try in code and then I'll send some spec text to the list |
| 04:36 | <Hixie> | abarth: this would be a generic change, not just for foreign content? |
| 04:37 | <abarth> | oh, i mean just for foreign content |
| 04:37 | <abarth> | basically, http://www.whatwg.org/specs/web-apps/current-work/multipage/tree-construction.html#parsing-main-inforeign |
| 04:37 | <abarth> | would check whether we're in the table text mode |
| 04:37 | <abarth> | and processing a non-character token |
| 04:38 | <abarth> | and then invoke the "anything else" clause of http://www.whatwg.org/specs/web-apps/current-work/multipage/tree-construction.html#parsing-main-intabletext |
| 04:38 | <abarth> | i guess an easier way |
| 04:39 | <abarth> | would be to force the HTML path when in the "in table text" |
| 04:39 | <abarth> | but have the "reprocess the token" in http://www.whatwg.org/specs/web-apps/current-work/multipage/tree-construction.html#parsing-main-intabletext |
| 04:39 | <abarth> | check whether it should go down the foreign content path |
| 04:39 | <Hixie> | abarth: it scares me to hardcode the table mode specifically there |
| 04:39 | <Hixie> | abarth: what about cell mode? select mode? etc |
| 04:39 | <abarth> | hum... |
| 04:40 | <Hixie> | abarth: i feel a more generic solution would be something i'd feel more confident was solid |
| 04:40 | <Hixie> | abarth: this is just a gut feeling though |
| 04:40 | <Hixie> | abarth: i'll have to look closer tomorrow |
| 04:41 | <abarth> | i guess the real strangeness is that we're in the "in table" model when the current node isn't a <table> |
| 04:41 | <abarth> | normally we'd get moved to another insertion mode |
| 04:41 | <abarth> | rather than staying in the table mode |
| 04:42 | <Hixie> | right; like i said in one of the mails in the earlier thread about this, this is a result of moving foreign content out of being a node |
| 04:42 | <Hixie> | er |
| 04:42 | <Hixie> | mode |
| 04:42 | <Hixie> | which was hsivonen's idea |
| 04:42 | <Hixie> | hence wanting his input :-) |
| 04:42 | <abarth> | yes, which is why i'm running into this problem in my patch that deletes the "in foreign content" mode :) |
| 04:45 | <Hixie> | indeed |
| 04:45 | <Hixie> | i think the change makes sense in general |
| 04:45 | <Hixie> | but this was an unexpected sideeffect and i'm worried there are others i've missed too |
| 07:02 | <zcorpan> | so the next XSS thing will be injecting <intent> tags |
| 07:29 | <webben> | may I please have an account on the WHATWG wiki? (creating new accounts is restricted to administrators) |
| 08:18 | <hsivonen> | I wonder if Glenn Adams is aware of http://en.wikipedia.org/wiki/Sega_v._Accolade#Trademark |
| 08:24 | <hsivonen> | webben: what user id and email would you like to use? |
| 08:40 | <abarth> | hsivonen: any thoughts on that parsing issue? |
| 08:41 | <abarth> | http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-December/034155.html |
| 08:41 | <abarth> | hsivonen: i need to go to sleep now, but any insight you can provide on the list would be appriciated |
| 08:41 | <abarth> | thx |
| 08:50 | <webben> | hsivonen: benjaminhawkeslewis , bhawkeslewis⊙gc - please |
| 08:50 | <webben> | hsivonen: oh - BenjaminHawkesLewis - sorry |
| 08:51 | <webben> | oh wait you can have spaces can't yo |
| 08:51 | <webben> | if hyphens are allowed Benjamin Hawkes-Lewis otherwise just BenjaminHawkesLewis |
| 08:51 | webben | is difficult. |
| 08:56 | <hsivonen> | webben: mediawiki claims to have sent you email |
| 08:57 | <hsivonen> | webben: the user id is BenjaminHawkesLewis |
| 09:29 | <annevk> | david_carlisle: there are certainly differences (that's the whole reason for this spec) |
| 09:30 | <annevk> | zewt: "correct" has never been defined for encodings, and since this is about legacy, it makes more sense to match the majority of implementations I think |
| 09:30 | <annevk> | ms2ger: I haven't, I probably should |
| 09:50 | <hsivonen> | hmm. http://blogs.adobe.com/jd/ has been quiet lately |
| 09:54 | <david_carlisle> | annevk: on differences, Guessed as much:-) there are other things in your spec unique to that (such as the various way implementations alias) but it would be good I think in a final version to be able to highlight any differences from the unicode site.. 'twould be easier for the reader to diff them if your tables were in an easier to process form such as XML :-) |
| 09:55 | <annevk> | I was planning on exposing at least JSON mapping for all the single-octet encodings |
| 09:55 | <annevk> | also useful for tests |
| 09:56 | <david_carlisle> | annevk: thanks |
| 10:12 | <annevk> | Ms2ger: just generated the tables again, does look cleaner now |
| 10:58 | <webben> | hsivonen: That works. Cheers :) |
| 11:35 | <annevk> | Ms2ger: updated spec |
| 11:35 | <annevk> | seems Opera fixed a few things too |
| 11:36 | <annevk> | this encoding alignment might actually work :) |
| 11:52 | <annevk> | Ms2ger: btw, you did not fix it for all encodings |
| 11:52 | <annevk> | Ms2ger: e.g. "macintosh" still has it |
| 12:01 | <annevk> | Ms2ger: nevermind |
| 12:15 | <Ms3ger> | annevk: good, I was confused why my change to the macintosh decoder didn't work out :) |
| 12:22 | <hsivonen> | where is encoding name comparison specced these days? |
| 12:22 | <hsivonen> | reality-based comparison, that is |
| 12:26 | <annevk> | that is still HTML5 |
| 12:26 | <annevk> | it's trim leading and trailing whitespace, ASCII case-insensitive compare |
| 12:26 | <annevk> | iirc |
| 12:26 | <annevk> | it will become part of the Encoding Standard |
| 12:27 | <annevk> | Ms3ger: Opera apparently tackled a similar issue for the windows-* encodings which has not yet made it into a release |
| 12:27 | <annevk> | Ms3ger: you might want to do the same |
| 12:28 | <annevk> | Ms3ger: see e.g. http://dvcs.w3.org/hg/encoding/raw-file/tip/Overview.html#windows-1255 |
| 12:32 | <annevk> | http://annevankesteren.nl/2010/8-bit-labels#ibm864 is kind of interesting |
| 12:32 | <annevk> | Firefox and IE versus WebKit and Chrome |
| 12:32 | <annevk> | seems like something WebKit did wrong |
| 12:32 | <annevk> | Opera has not implemented that one at all... |
| 12:37 | <Ms3ger> | annevk: oh, Opera changed those already? I'll have a look |
| 12:42 | <annevk> | Ms3ger: well, for the next public release |
| 12:42 | <Ms3ger> | Yeah |
| 12:53 | <Ms3ger> | Hmm, it doesn't seem terribly clear in DOM4 that you need to call EventListener.handleEvent |
| 13:01 | <Ms3ger> | Looks like I'll have to learn more about this file format than I'd like... |
| 13:05 | <annevk> | Ms3ger: not changing octet 25 in ibm864 would also be good |
| 13:17 | <annevk> | why is ibm864 such a mess? |
| 13:17 | <annevk> | meh, I'll just make some choices and let people complain |
| 13:30 | <woef> | If I want to use the itemscope in xhtml5 (or whatever I should call it), do I write itemscope="itemscope" ? |
| 13:30 | <woef> | ("xml serialization of html5" I believe) |
| 13:31 | <gsnedders> | woef: yes |
| 13:31 | <gsnedders> | woef: either that or itemscope="" |
| 13:31 | <woef> | okay, cool :) |
| 13:32 | <woef> | Is there a place where I can verify how Google/others will read this data (to verify if I implemented it well?) |
| 13:34 | <annevk> | validator.nu ? |
| 13:39 | <annevk> | so JSON can be just be a [] literal right? |
| 13:41 | <woef> | http://www.google.com/webmasters/tools/richsnippets |
| 13:41 | <woef> | No idea if validator.nu does something similar. |
| 13:43 | <gsnedders> | annevk: yes |
| 13:44 | <gsnedders> | annevk: JSON can be an object literal or an array literal |
| 13:59 | <bga> | gsnedders can you recomend me sources of hight order stuff? Programming langs, modern math. Some ppl in twitter, blogs etc |
| 14:01 | <gsnedders> | bga: What are you actually looking for? |
| 14:02 | <smaug____> | heycam|away: what should happen if non-object is passed as dictionary |
| 14:02 | <smaug____> | say new Event("foo", 1) |
| 14:04 | <bga> | gsnedders new papers, researches |
| 14:04 | <gsnedders> | bga: For that sort of stuff, I'm the wrong person to ask. |
| 14:04 | <bga> | tired to grep forums to find something new |
| 14:04 | <bga> | heh |
| 14:19 | <annevk> | david_carlisle: http://dvcs.w3.org/hg/encoding/raw-file/tip/single-octet-encodings.json |
| 14:19 | <annevk> | david_carlisle: it's also linked from the specification |
| 14:20 | <annevk> | david_carlisle: from now I'll generate new tables from that source to make sure they stay in sync |
| 14:20 | <annevk> | not entirely sure I want to automate the entire specification, does not seem worth it |
| 14:20 | <annevk> | and it will not work nicely when non single-octet encodings are added to the list of encodings |
| 14:25 | <david_carlisle> | annevk: thanks |
| 14:26 | <david_carlisle> | can't you just have a bigger array for the two byte encodings? or do you mean you'd want some kind of sparse format |
| 14:29 | <annevk> | david_carlisle: they're not two byte, they're multi |
| 14:35 | <david_carlisle> | annevk: yes sure but they can all be thought of as a mapping from some finite range of integers to a the codepoint range so can all be represented by an array of integers as you have so far, can;t they? (admittedly that doesn't capture how the bytes represent the encoding number but) so for example utf-8 would just be an idenity as the numeric part is an identity map, the actual utf-8 bit... |
| 14:35 | <david_carlisle> | ...layout wouldn't fit the layout in your current json, perhaps that's what you meant originally |
| 14:38 | <hsivonen> | woef: Validator.nu doesn't check for suitability for consumption by Google |
| 14:48 | <annevk> | david_carlisle: I think some of the multi octet encodings have state |
| 14:48 | <annevk> | david_carlisle: also not sure error handling works for them, e.g. for UTF-8 |
| 14:48 | <annevk> | david_carlisle: but I need to look at them more closely basically |
| 14:48 | <annevk> | david_carlisle: for now I'm happy with just doing the single-octet stuff |
| 15:08 | <annevk> | where did Ms2ger go? |
| 15:08 | <annevk> | or Ms3ger for that matter |
| 15:09 | <jgraham> | Maybe he is metamorphosing into Ms4ger |
| 15:09 | <jgraham> | On the internet, no one knows you're a butterfly |
| 15:59 | <zewt> | annevk: well, there's often no majority (gecko+presto vs. trident+webkit), it's existing content that matters, of course ... anyway, i'm guessing at least the windows-* discrepencies should be easy to get fixed |
| 16:00 | <zewt> | seems pretty obvious to treat IE as the authority on those mappings |
| 16:02 | <annevk> | if the four others did the same I opted for following them instead |
| 16:03 | <zewt> | smaug____: new Event("foo", 1) throws TypeError, step 1 of http://dev.w3.org/2006/webapi/WebIDL/#es-dictionary |
| 16:03 | <smaug____> | yup |
| 16:03 | <smaug____> | just adding that to the implementation |
| 16:04 | <annevk> | zewt: I did favor trident/webkit over presto/gecko; same for trident/gecko |
| 16:04 | <annevk> | zewt: and WebKit is not quite the same when it comes to this |
| 16:04 | <annevk> | so I have both Chromium and WebKit |
| 16:04 | <zewt> | fun stuff |
| 16:05 | <zewt> | i wonder if there really isn't much content on "arabic" on those codepoints trident assigns mappings to |
| 16:06 | <annevk> | for multi-octet encodings it seems I should define algorithms for encode and decode |
| 16:06 | <annevk> | and then cleverly test them |
| 16:07 | <zewt> | (why do people randomly like to call bytes "octets"? heh) |
| 16:07 | <zewt> | they're multibyte encodings, damn it :P |
| 16:07 | <annevk> | zewt: http://en.wikipedia.org/wiki/Octet_(computing) |
| 16:07 | <zewt> | i know what the word means; in the real world we call them bytes |
| 16:07 | <hober> | zewt: because on older architectures byte length differed |
| 16:07 | <zewt> | hober: doesn't matter in 2011 (or 2001) |
| 16:08 | <annevk> | zewt: see http://dvcs.w3.org/hg/encoding/raw-file/tip/Overview.html#notes for general rules I used |
| 16:08 | <annevk> | zewt: it might be worth going with byte and defining that as eight bits |
| 16:08 | <annevk> | zewt: email WHATWG? |
| 16:09 | <zewt> | annevk: i'm guessing data will be wanted about the contentious mappings (if lots of pages are using those trident arabic mappings, that's probably what should be used, even if it's currently the minority...) |
| 16:10 | <zewt> | as long as there's data to back it, i'd be surprised if there was much pushback from browsers for adding them |
| 16:10 | <annevk> | yeah |
| 16:10 | <gsnedders> | zewt: If you're dealing with mainframes it does sometimes still matter. |
| 16:12 | <annevk> | zewt: that's one of the things mentioned |
| 16:12 | <zewt> | gsnedders: (i leave that to that field to translate terminology; the rest of the world has moved on) |
| 16:14 | <zewt> | mapping ascii to windows-1252 makes me sad inside |
| 16:58 | <gsnedders> | WTF? Thunderbird has started using Chrome for https links, though nowhere is Chrome set as the default. |
| 17:00 | <jgraham> | Google's marketing has become so agrressive that they now have Mozilla promoting them? :p |
| 17:00 | <martndemus> | Thunderbird knows whats best =D |
| 18:35 | <TabAtkins> | Hm. I'm not sure what mailing list to talk about this on. |
| 18:35 | <TabAtkins> | So, I'm in a discussion with our spreadsheet team. |
| 18:35 | <TabAtkins> | Spreadsheets are, in essence, a giant <table>. This is, predictably, absolutely shit for performance. They're trying to find ways around it, to the point of considering implementing all their rendering in WebGL instead. |
| 18:37 | <TabAtkins> | I'd like to discuss ways to fix this in HTML/CSS/JS. Possibilities include: a "just trust me" table layout mode that uses less constraints (allows overflow, etc.) to bring performance back down from quadratic+; a "virtual table" object that draws a simplistic rendering of empty cells and calls you to fill them in with real DOM when users stop scrolling. |
| 18:37 | <TabAtkins> | Possibly other ideas. |
| 18:37 | <TabAtkins> | So, where to bring this up? |
| 18:37 | <TabAtkins> | I'm thinking whatwg might be the best place. |
| 18:38 | <gsnedders> | TabAtkins: SVG might work decently — but Fx doesn't support selection of SVG, does it? |
| 18:38 | <TabAtkins> | gsnedders: I don't know if SVG would work. If we support the same spreadsheet size as Excel, that's 2^8 columns and 2^16 rows. |
| 18:38 | <TabAtkins> | Excel can do fast scrolling on that because it just cheats for the most part until you stop scrollilng. |
| 18:39 | <TabAtkins> | But 2^24 DOM nodes will kill a browser regardless of whether it's HTML or SVG, I think. |
| 18:40 | <Ms2ger> | table-layout: fixed? |
| 18:41 | <TabAtkins> | I'm pretty sure table-layout:fixed isn't "fixed" enough. |
| 18:41 | <TabAtkins> | You still need to do layout to determine the height of rows, for example. |
| 18:41 | <Ms2ger> | I guess we need table-layout: fixedforrealthistime |
| 18:41 | <TabAtkins> | table-layout:static |
| 18:42 | <Ms2ger> | But you only get to add that if you spec normal tables first :) |
| 18:42 | <TabAtkins> | I REFUSE |
| 18:42 | <Ms2ger> | Then I REFUSE to implement it |
| 18:43 | <Ms2ger> | Which shouldn't bother you, I wouldn't implement it anyway |
| 18:43 | <gsnedders> | Children, no fighting. |
| 18:43 | <Ms2ger> | gsnedders, remind me, are you younger or older than me? :) |
| 18:43 | <gsnedders> | Ms2ger: I believe I am still the youngest active person in this channel. |
| 18:43 | <gsnedders> | Ms2ger: (19) |
| 18:44 | <Ms2ger> | Good |
| 18:44 | <Ms2ger> | So be quiet when adults talk ;) |
| 18:44 | <gsnedders> | Ms2ger: Pfff. |
| 18:44 | <gsnedders> | Ms2ger: I've been an adult since I was 16! |
| 18:45 | <Ms2ger> | Scot? |
| 18:45 | <gsnedders> | Ms2ger: Yup. |
| 18:45 | <gsnedders> | (I then regressed and became a child for a while again while in Sweden) |
| 18:45 | <Ms2ger> | Heh |
| 18:46 | <gsnedders> | It was quite amusing going from an adult to being a child. Don't get to do that often. :) |
| 18:57 | <Ms2ger> | annevk5, why would not changing octet 25 in ibm864 be good? |
| 19:30 | <zewt> | changing the ASCII overlap is never good |
| 19:34 | <Ms2ger> | Firefox is outnumbered 3-to-1 |
| 19:46 | <zewt> | all browsers are outnumbered, they're all doing different things |
| 19:47 | jcranmer | wonders why he remains in this channel |
| 19:47 | <zewt> | are you asking us? |
| 19:48 | <gsnedders> | jcranmer: Because we're cool? |
| 19:48 | <zewt> | Ms2ger: if IE is handling the ASCII range for that encoding sanely, then I'd hope other browsers could align at least for those |
| 19:48 | <jcranmer> | I only passively follow HTML development, and even that is beginning to wear thin |
| 19:48 | jcranmer | settles for "inertia" |
| 19:48 | Ms2ger | whacks jcranmer |
| 19:49 | <zewt> | (i don't really care about the others for such an obscure encoding, but having as close to everything as possible align for ASCII is big) |
| 19:49 | <Ms2ger> | zewt, did you look at the data? |
| 19:49 | <jcranmer> | let's go to EBCIDIC? |
| 19:50 | <zewt> | i've only looked at the tables |
| 19:50 | <Ms2ger> | http://annevankesteren.nl/2010/8-bit-labels#ibm864 |
| 19:50 | <Ms2ger> | So what would you recommend for 0x25? |
| 19:51 | <zewt> | according to http://dvcs.w3.org/hg/encoding/raw-file/tip/Overview.html#ibm864, trident maps [0,7f] to identity for that encoding |
| 19:51 | <zewt> | which is what all mbcs and 8bit encodings should do, so that's what I'd recommend |
| 19:51 | zewt | squints at funkytable |
| 19:52 | <zewt> | looks like it (unsurprisingly) says the same thing |
| 19:56 | <zewt> | based on those tables, it should be possible to align all of these encodings that way |
| 19:56 | <zewt> | assuming other browsers don't dig in their heels for some obscure reason |
| 19:56 | <zewt> | (and ignoring multibyte encodings for now, of course) |
| 20:02 | <zewt> | how could anyone ever use 864 in Firefox, with it remapping %? |
| 20:03 | <zewt> | that's well into crazyland |
| 20:11 | <karega|aniasis> | does anyone know any good design examples of horizontal accordions |
| 20:12 | <Ms2ger> | Are there any? |
| 20:14 | <karlcow> | yeah a few karega|aniasis http://www.fallonemusic.com/images/Hohner_accordeons_2_XL.jpg |
| 20:15 | <karlcow> | also http://www.skripta-paris.com/810-9621-large/japanese-album-accordion-moleskine-pocket-a6.jpg |
| 20:15 | <karega|aniasis> | accordions for web content |
| 20:16 | <erlehmann> | are those not http? |
| 20:16 | <karega|aniasis> | yes, but I mean accordions for web based content layout |
| 20:17 | <karega|aniasis> | not the instrument |
| 20:17 | <karlcow> | erlehmann: those are http pipes(lining) ;) |
| 20:17 | <karega|aniasis> | no links to instruments |
| 20:17 | <erlehmann> | a web accordion seems like a weird instrument |
| 20:17 | <karlcow> | karega|aniasis: yes we are being saccarstic :) |
| 20:18 | <karega|aniasis> | it does, and that's why I'm not talking referring to it. |
| 20:18 | <karlcow> | karega|aniasis: I have no answer more than would you get on DuckDuckGo |
| 20:18 | <karega|aniasis> | karlcow, yes I know and I am going along with your sarcasm. |
| 20:18 | <karlcow> | karega|aniasis: you would have better chances I guess in #html5 |
| 20:24 | <bga> | too many questions about webkit's console.log |
| 20:24 | <bga> | plz do deep copy |
| 21:34 | <Hixie> | TabAtkins: yt? |
| 21:34 | <Hixie> | TabAtkins: can you take a look at http://lists.whatwg.org/htdig.cgi/help-whatwg.org/2011-December/000966.html ? |
| 21:34 | <jgraham> | Disadvantages of using Google as a calculator: inch to cm conversions lead to lots of penis-related search results |
| 21:35 | <jgraham> | In other news, bein able to do mental arithmetic is still good |
| 21:35 | <Hixie> | (or anyone else who is up to date on table model rendering rules) |
| 21:35 | <gsnedders> | jgraham: multiplying by 2.54 have that affect as well? |
| 21:36 | <zewt> | then you have to remember 2.54 :) |
| 21:37 | <jgraham> | Hixie: (I don't know what the spec says but the behaviour he wants seems very magical) |
| 21:37 | <Hixie> | well my real question is whether he should file the bug on gecko or webkit |
| 21:38 | <jgraham> | In particular http://btrem.com/tests/table1 seems like it much be the wrong expectation |
| 21:38 | <jgraham> | *must |
| 21:39 | <jgraham> | The other one I don't know. And I would give even odds that CSS doesn't know either :) |
| 21:39 | <Ms2ger> | gtalbot would probably be able to say something about those tests |
| 21:40 | <jgraham> | Or dbaron |
| 21:40 | Ms2ger | curses annevk5 |
| 21:54 | <Hixie> | hsivonen: ping |
| 21:56 | <jgraham> | Pretty sure that if you look at the distribution of when hsivonen is in the channel the probability he is around is tiny. Unless he is abroad or something. |
| 22:00 | <Hixie> | jgraham: always worth a try |
| 22:01 | <Hixie> | i guess i'll just have to work out this parser thing myself :-P |
| 22:02 | <gsnedders> | Hixie: You wrote it! :P |
| 22:02 | <Hixie> | yeah but that was months ago |
| 22:02 | <Hixie> | i've offloaded all my knowledge to off-site storage |
| 22:02 | <Hixie> | (the spec) |
| 22:03 | <Hixie> | any html5lib people here? |
| 22:04 | <gsnedders> | Myself and jgraham? |
| 22:04 | <gsnedders> | (Though certainly my knowledge of the parser isn't as great as it was a year ago…) |
| 22:05 | <gsnedders> | But jgraham's should be, if he's been doing his work properly. :P |
| 22:05 | <Hixie> | do you know if it's up to date with the spec? |
| 22:05 | <Hixie> | in particular the foreign content stuff |
| 22:07 | <gsnedders> | Not for any changes in the past six months |
| 22:08 | <Hixie> | k |
| 22:09 | <TabAtkins> | Hixie: His "test1" (using the separated border model) is wrong. There is no cell in the lower-right corner, so there's nothing for CSS to target and apply a border to. |
| 22:10 | <TabAtkins> | In the "test2" (using the collapsed border model, and applying border to <tr>), it's a WebKit bug. |
| 22:10 | <TabAtkins> | Which I believe is already filed. |
| 22:10 | <gsnedders> | Yeah, that's an ancient bug/ |
| 22:10 | <Hixie> | TabAtkins: k. i'll reply to the mail, unless you'd like to |
| 22:10 | <Hixie> | TabAtkins: (and thanks) |
| 22:10 | <TabAtkins> | I'm not subscribed to that list, so go ahead. |
| 22:10 | <Hixie> | k |
| 22:15 | <astearns> | TabAtkins: then what does "the end of that row should be padded with empty cells" mean for his test1? |
| 22:17 | <TabAtkins> | It doesn't mean anything, because HTML4 has no conformance criteria in it. |
| 22:19 | <Ms2ger> | Sure does |
| 22:19 | <Ms2ger> | It requires quote marks for q |
| 22:19 | <TabAtkins> | Yup. |
| 22:19 | <TabAtkins> | That rounds to zero. |
| 22:25 | <Ms2ger> | annevk5, you'll need to help me reverse engineer our file format for encodings if you want me to fix anything in Gecko :) |
| 22:32 | <Hixie> | ok how the heck do you escape a ! in a ""-quoted string in bash |
| 22:32 | <Hixie> | \! turns into a literal backslash-bang |
| 22:33 | <zewt> | i usually just give up and use ' |
| 22:33 | <Hixie> | me too |
| 22:34 | <Hixie> | but sometimes i want inner single quotes |
| 22:34 | <zewt> | "a"'!'"b" |
| 22:34 | <jgraham> | I really hate bash |
| 22:34 | <jgraham> | That is all |
| 22:34 | <zewt> | (more important when you want expansions) |
| 22:34 | <Hixie> | zewt: oh you can do that? interesting |
| 22:34 | <Hixie> | funky |
| 22:34 | <Hixie> | zewt: thanks! |
| 22:35 | <zewt> | a rare universal among programmers; everyone hates bash, and uses it anyway |
| 22:36 | <Hixie> | i don't hate bash |
| 22:36 | <zewt> | you should |
| 22:47 | <heycam> | smaug____, a TypeError is thrown if the value passed in isn't an object: http://dev.w3.org/2006/webapi/WebIDL/#es-dictionary |
| 22:47 | <heycam> | smaug____, (oh I see zewt answered for you) |
| 22:47 | <heycam> | (I process my scrollback as I scroll!) |
| 22:47 | <zewt> | (since I asked the same thing yesterday or so :) |
| 22:48 | <heycam> | TabAtkins, you're right that Firefox doesn't support selection of SVG text; being worked on though https://bugzilla.mozilla.org/show_bug.cgi?id=655877 |
| 22:52 | <TabAtkins> | That was gsnedders, but OK. |
| 22:59 | <wiz-dumb> | Hello! Are there any developers from the html5lib project around? |
| 23:00 | <jgraham> | Yes, but I soon won't be |
| 23:00 | <jgraham> | (but please go ahead) |
| 23:00 | <wiz-dumb> | Just wanted to see if I could get a status check on an open issue regarding test/ directory missing from the 0.90 release |
| 23:01 | <wiz-dumb> | https://code.google.com/p/html5lib/issues/detail?id=134#c2 |
| 23:01 | <jgraham> | The status is that we need to make a new release |
| 23:01 | <wiz-dumb> | Trying to upgrade the version we have at the office to .90, but this is blocking |
| 23:01 | <jgraham> | but have utterly failed to do that |
| 23:02 | <wiz-dumb> | Ok, would it be possible to have the .90 release .zip updated to simply include the tests which existed when it was packaged? |
| 23:02 | <jgraham> | OK. Well I have to sleep now but if you are blocking on this that provides some motivation to actually get a new release together |
| 23:03 | <wiz-dumb> | ok. I added myself to the issue with the same comment. I look forward to an update. Good night! |
| 23:03 | <jgraham> | I don't want to spend time on that. But it is all in hg so you can get the tests yourself if you need them. That doesn't work so well if you depend on PyPI ofc |
| 23:05 | <wiz-dumb> | Ok, thanks for the help |
| 23:13 | <MikeSmith> | http://ariya.ofilabs.com/2011/12/introducing-esprima.html |
| 23:13 | <MikeSmith> | bravo Ariya |
| 23:13 | <MikeSmith> | http://www.esprima.org/demo/parse.html |
| 23:18 | <TabAtkins> | MikeSmith: That's pretty cool! |
| 23:18 | <MikeSmith> | yeah |
| 23:19 | <MikeSmith> | Ariya's a mad scientist |
| 23:21 | <MikeSmith> | incidentally, I didn't know about the Mozilla parser API |
| 23:21 | <MikeSmith> | but looking at that, it's pretty cool too |
| 23:21 | <MikeSmith> | https://developer.mozilla.org/en/SpiderMonkey/Parser_API |
