| 07:09 | <zcorpan> | "In this case the "proposed changes" will very likely be the null set if the CfC passes." |
| 07:09 | <zcorpan> | very likely? |
| 07:10 | <zcorpan> | "In this case the "proposed changes" will very likely be the null set if the CfC passes." |
| 07:10 | <zcorpan> | very likely? |
| 07:11 | <zcorpan> | it sounds a bit as if the chairs haven't reviewed the cp properly yet to be absolutely sure whether it would result in no changes or not |
| 07:11 | <zcorpan> | it sounds a bit as if the chairs haven't reviewed the cp properly yet to be absolutely sure whether it would result in no changes or not |
| 07:14 | <zcorpan> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1377 seems kinda fragile in opera |
| 07:14 | <zcorpan> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1377 seems kinda fragile in opera |
| 07:15 | <zcorpan> | sometimes it disappears, sometimes it gets a random color, sometimes a table border is rendered but no cells, and sometimes it renders as it should... |
| 07:15 | <zcorpan> | sometimes it disappears, sometimes it gets a random color, sometimes a table border is rendered but no cells, and sometimes it renders as it should... |
| 07:27 | <zcorpan> | hmm seems it just gets a random color and sometimes is fully transparent |
| 07:27 | <zcorpan> | hmm seems it just gets a random color and sometimes is fully transparent |
| 07:38 | <othermaciej> | zcorpan: it's very clear that it will result in no changes (as least to me) |
| 07:39 | <othermaciej> | zcorpan: it's very clear that it will result in no changes (as least to me) |
| 07:40 | <zcorpan> | othermaciej: ok |
| 07:41 | <zcorpan> | othermaciej: ok |
| 12:35 | <bga> | youtube discriminante Opera |
| 12:35 | <bga> | for Opera docuemt.write(<embed> |
| 12:36 | <bga> | youtube discriminante Opera |
| 12:36 | <bga> | for Opera docuemt.write(<embed> |
| 12:36 | <bga> | => dont work w/o js |
| 12:36 | <bga> | => dont work w/o js |
| 12:36 | <bga> | for ff - just <embed> |
| 12:36 | <bga> | for IE - <object> |
| 12:37 | <bga> | for ff - just <embed> |
| 12:37 | <bga> | for IE - <object> |
| 12:37 | <gsnedders> | bga: That's all to do with the click-to-activate plugin nonsense. |
| 12:37 | <gsnedders> | bga: That's all to do with the click-to-activate plugin nonsense. |
| 12:42 | <Dashiva> | gsnedders: Speaking of that |
| 12:43 | <Dashiva> | gsnedders: Speaking of that |
| 12:43 | <Dashiva> | How do you click to activate when there's no window? |
| 12:43 | <Dashiva> | How do you click to activate when there's no window? |
| 12:43 | <Dashiva> | E.g. the chat sounds in gmail |
| 12:44 | <Dashiva> | E.g. the chat sounds in gmail |
| 12:46 | <gsnedders> | Dashiva: the plugin can still run, and produce sounds, while not active |
| 12:46 | <gsnedders> | Dashiva: the plugin can still run, and produce sounds, while not active |
| 12:48 | <bga> | gsnedders speaking about sound. What "Enable Sound in Web pages" in f12 disable? Only <bgsound>? or <audio> too? |
| 12:49 | <bga> | gsnedders speaking about sound. What "Enable Sound in Web pages" in f12 disable? Only <bgsound>? or <audio> too? |
| 12:50 | <gsnedders> | bgsound, I believe. |
| 12:51 | <gsnedders> | bgsound, I believe. |
| 12:51 | <bga> | :( |
| 12:51 | <bga> | it should disable <audio> too |
| 12:52 | <bga> | :( |
| 12:52 | <bga> | it should disable <audio> too |
| 12:52 | <bga> | and <embed midi/wav> |
| 12:52 | <bga> | and <embed midi/wav> |
| 12:52 | <bga> | embed/iframe |
| 12:53 | <bga> | embed/iframe |
| 13:01 | <Dashiva> | gsnedders: Experience suggests it doesn't |
| 13:01 | <Dashiva> | gsnedders: Experience suggests it doesn't |
| 13:01 | <Dashiva> | There's no chat sounds in gmail in Opera |
| 13:02 | <Dashiva> | There's no chat sounds in gmail in Opera |
| 14:02 | zcorpan | fiddled a bit more with http://simon.html5.org/specs/quirks-mode#the-percentage-height-quirk |
| 14:03 | zcorpan | fiddled a bit more with http://simon.html5.org/specs/quirks-mode#the-percentage-height-quirk |
| 15:01 | <wilhelm> | Which browsers support object-fit these days? |
| 15:02 | <wilhelm> | Which browsers support object-fit these days? |
| 15:12 | <zewt> | if you have two WebIDL overloads, void foo(interface1 a) and void foo(interface2 b), and you call with an object Z that implements neither interface, exactly what happens? |
| 15:13 | <zewt> | if you have two WebIDL overloads, void foo(interface1 a) and void foo(interface2 b), and you call with an object Z that implements neither interface, exactly what happens? |
| 15:13 | <zewt> | it seems like TypeError is supposed to be thrown from http://dev.w3.org/2006/webapi/WebIDL/#es-interface, but it's not clear how it gets *there* in the first place |
| 15:14 | <zewt> | it seems like TypeError is supposed to be thrown from http://dev.w3.org/2006/webapi/WebIDL/#es-interface, but it's not clear how it gets *there* in the first place |
| 15:16 | <zewt> | (not sure where overload selection is actually defined; 3.2.6 doesn't seem to do it) |
| 15:16 | <zewt> | (not sure where overload selection is actually defined; 3.2.6 doesn't seem to do it) |
| 15:19 | <zewt> | oh, hidden away in 4.4.1.1; the word "overload" needs to show up in the TOC for it somewhere |
| 15:19 | <zewt> | oh, hidden away in 4.4.1.1; the word "overload" needs to show up in the TOC for it somewhere |
| 16:13 | <Ms2ger> | zewt, ah, now I see why you were asking :) |
| 16:13 | <Ms2ger> | zewt, ah, now I see why you were asking :) |
| 17:22 | <kennyluck_> | I don't understand how would text/html in DOMParser take 10 years to finish → https://bugzilla.mozilla.org/show_bug.cgi?id=102699 |
| 17:23 | <kennyluck_> | I don't understand how would text/html in DOMParser take 10 years to finish → https://bugzilla.mozilla.org/show_bug.cgi?id=102699 |
| 17:23 | <kennyluck_> | I thought there was some technical debate about this feature, though I don't see any. |
| 17:23 | <kennyluck_> | in the bug, at least. |
| 17:23 | <kennyluck_> | I thought there was some technical debate about this feature, though I don't see any. |
| 17:23 | <kennyluck_> | in the bug, at least. |
| 17:23 | <annevk> | nobody interested in doing it is usually the reason |
| 17:24 | <annevk> | nobody interested in doing it is usually the reason |
| 17:42 | <Ms2ger> | Also, the fact that our old HTML parser has been unmaintained for a decade or so |
| 17:42 | <Ms2ger> | Also, the fact that our old HTML parser has been unmaintained for a decade or so |
| 17:42 | <annevk> | kennyluck_: btw, have you heard back yet on lists.w3.org hosting a copy of whatwg⊙wo? |
| 17:43 | <annevk> | kennyluck_: btw, have you heard back yet on lists.w3.org hosting a copy of whatwg⊙wo? |
| 17:44 | <kennyluck_> | annevk, no. You could ask MikeSmith. |
| 17:45 | <kennyluck_> | annevk, no. You could ask MikeSmith. |
| 17:50 | <bga> | i wonder why ppl dont accept pt and cm units in css |
| 17:51 | <bga> | i wonder why ppl dont accept pt and cm units in css |
| 17:51 | <bga> | font-size: 0.5cm |
| 17:51 | <bga> | font-size: 0.5cm |
| 17:51 | <bga> | look identical in all UAs |
| 17:51 | <bga> | not related your device dpi |
| 17:51 | <gsnedders> | bga: IE6. |
| 17:52 | <bga> | look identical in all UAs |
| 17:52 | <bga> | not related your device dpi |
| 17:52 | <gsnedders> | bga: IE6. |
| 17:52 | <bga> | ie6 what? |
| 17:52 | <bga> | ie6 what? |
| 17:52 | <bga> | dont support it? |
| 17:53 | <bga> | dont support it? |
| 17:55 | <gsnedders> | bga: Scaling didn't work with absolute units in IE6, for a11y reasons was discouraged |
| 17:56 | <gsnedders> | bga: Scaling didn't work with absolute units in IE6, for a11y reasons was discouraged |
| 17:56 | <bga> | ok. so when IE6 will die we now can use cm? |
| 17:57 | <bga> | ok. so when IE6 will die we now can use cm? |
| 17:57 | <bga> | many projects doesnt support IE6 today |
| 17:57 | <gsnedders> | bga: I guess so. |
| 17:57 | <bga> | many projects doesnt support IE6 today |
| 17:57 | <gsnedders> | bga: I guess so. |
| 18:05 | <kennyluck_> | I never try hard to understand the debates about physical units in CSS. There are so many long long threads about them on www-style... |
| 18:05 | <kennyluck_> | I never try hard to understand the debates about physical units in CSS. There are so many long long threads about them on www-style... |
| 18:29 | <zewt> | cm units would be very strange on large monitors if they actually worked |
| 18:30 | <zewt> | cm units would be very strange on large monitors if they actually worked |
| 18:31 | <zewt> | annevk: re utf32--i wouldn't quote that; it reads like nothing but unicode people trying to pretend utf-16 wasn't a gross mistake |
| 18:31 | <zewt> | annevk: re utf32--i wouldn't quote that; it reads like nothing but unicode people trying to pretend utf-16 wasn't a gross mistake |
| 18:31 | <zewt> | (it talks about "industrial-strength text", which ignores the fact that 98% of code isn't that) |
| 18:32 | <zewt> | (it talks about "industrial-strength text", which ignores the fact that 98% of code isn't that) |
| 18:41 | <bga> | zewt check it http://dl.dropbox.com/u/5332813/test/screenDiagonal/index.xhtml |
| 18:41 | <bga> | zewt check it http://dl.dropbox.com/u/5332813/test/screenDiagonal/index.xhtml |
| 19:03 | <annevk> | zewt: if utf-16 was a mistake, why wouldn't utf-32 be? |
| 19:03 | <annevk> | zewt: if utf-16 was a mistake, why wouldn't utf-32 be? |
| 19:06 | <StoneCypher> | because 16's got varying width codepoints and 32 doesn't |
| 19:06 | <StoneCypher> | the issue is about whether you have constant time random access to string contents |
| 19:07 | <StoneCypher> | because 16's got varying width codepoints and 32 doesn't |
| 19:07 | <StoneCypher> | which is a lot more important than most people realize |
| 19:07 | <StoneCypher> | the issue is about whether you have constant time random access to string contents |
| 19:07 | <StoneCypher> | which is a lot more important than most people realize |
| 19:10 | <annevk> | doesn't help with grapheme clusters |
| 19:10 | <annevk> | doesn't help with grapheme clusters |
| 19:10 | <StoneCypher> | yeah, it does |
| 19:10 | <annevk> | euh no |
| 19:10 | <StoneCypher> | those are distinct codepoints, iirc (they aren't distinct characters) |
| 19:11 | <StoneCypher> | yeah, it does |
| 19:11 | <annevk> | euh no |
| 19:11 | <StoneCypher> | those are distinct codepoints, iirc (they aren't distinct characters) |
| 19:11 | <StoneCypher> | the germane issue is, it lets you find whatever the label is of the granular lookup in constant time, without a string scan |
| 19:11 | <StoneCypher> | the germane issue is, it lets you find whatever the label is of the granular lookup in constant time, without a string scan |
| 19:11 | <gsnedders> | annevk: It helps because you can then check general category of each codepoint in constant time in the string |
| 19:12 | <gsnedders> | annevk: It helps because you can then check general category of each codepoint in constant time in the string |
| 19:12 | <StoneCypher> | it's much more general than that |
| 19:12 | <StoneCypher> | you can act in constant time |
| 19:12 | <StoneCypher> | it's much more general than that |
| 19:12 | <StoneCypher> | you can act in constant time |
| 19:12 | <StoneCypher> | you can take substrings, you can replace characters, you can do searches, etc |
| 19:12 | <StoneCypher> | without fixed-width granular units you are forced to behave in a fashion equivalent to a linked list |
| 19:12 | <annevk> | well you can't do a lot of that because of grapheme clusters |
| 19:12 | <StoneCypher> | again |
| 19:13 | <StoneCypher> | you can take substrings, you can replace characters, you can do searches, etc |
| 19:13 | <StoneCypher> | without fixed-width granular units you are forced to behave in a fashion equivalent to a linked list |
| 19:13 | <StoneCypher> | you're speaking at a higher level of organization than is appropriate |
| 19:13 | <annevk> | well you can't do a lot of that because of grapheme clusters |
| 19:13 | <StoneCypher> | again |
| 19:13 | <StoneCypher> | you're speaking at a higher level of organization than is appropriate |
| 19:13 | <StoneCypher> | there are layers in unicode |
| 19:13 | <StoneCypher> | there are layers in unicode |
| 19:13 | <StoneCypher> | this is about finding positions in the codepoint stream, not interpreting them |
| 19:13 | <StoneCypher> | this is about finding positions in the codepoint stream, not interpreting them |
| 19:13 | <StoneCypher> | if you want to discuss interpretation, you *still* get a significant complexity increase from locating the things you later interpret |
| 19:13 | <annevk> | if you want to do text manipulation, you should look at the text, not code points |
| 19:13 | <StoneCypher> | and what do you believe the text is made of |
| 19:14 | <StoneCypher> | if you want to discuss interpretation, you *still* get a significant complexity increase from locating the things you later interpret |
| 19:14 | <annevk> | if you want to do text manipulation, you should look at the text, not code points |
| 19:14 | <StoneCypher> | and what do you believe the text is made of |
| 19:14 | <StoneCypher> | this is an issue of datastructure access time |
| 19:14 | <StoneCypher> | this is an issue of datastructure access time |
| 19:14 | <StoneCypher> | addressing what's in the datastructure is missing the point |
| 19:14 | <StoneCypher> | addressing what's in the datastructure is missing the point |
| 19:16 | <annevk> | not sure what you're saying |
| 19:17 | <annevk> | not sure what you're saying |
| 19:17 | <gsnedders> | annevk: If you want to do text manipulation, you have to look at the text, and to look at the text, you have to look at its underlying representation, so you have to look at its codepoints. |
| 19:17 | <gsnedders> | annevk: If you want to do text manipulation, you have to look at the text, and to look at the text, you have to look at its underlying representation, so you have to look at its codepoints. |
| 19:17 | <gsnedders> | annevk: So performance of looking at the codepoints is relevant for text manipulaton |
| 19:18 | <gsnedders> | annevk: So performance of looking at the codepoints is relevant for text manipulaton |
| 19:18 | <annevk> | are you saying perf would be drastically impacted because of utf-16? |
| 19:18 | <annevk> | surrogates are really easy to calculate and find too |
| 19:19 | <annevk> | are you saying perf would be drastically impacted because of utf-16? |
| 19:19 | <annevk> | surrogates are really easy to calculate and find too |
| 19:19 | <StoneCypher> | yes, performance is radically impacted by imposing non-constant time access on strings. |
| 19:20 | <StoneCypher> | yes, performance is radically impacted by imposing non-constant time access on strings. |
| 19:21 | <zewt> | annevk: utf-32 is a reasonable internal format for strings; utf-16 is just garbage |
| 19:22 | <zewt> | annevk: utf-32 is a reasonable internal format for strings; utf-16 is just garbage |
| 19:22 | <gsnedders> | And UTF-8 is even worse. |
| 19:22 | <annevk> | given that memory has been an argument in the past, I'd argue utf-32 is prolly worse, and utf-8 would be better |
| 19:22 | <zewt> | utf-16 combines the cons of both utf-8 (variable length) and utf-32 (not ASCII-compatible) |
| 19:23 | <gsnedders> | And UTF-8 is even worse. |
| 19:23 | <annevk> | given that memory has been an argument in the past, I'd argue utf-32 is prolly worse, and utf-8 would be better |
| 19:23 | <zewt> | utf-16 combines the cons of both utf-8 (variable length) and utf-32 (not ASCII-compatible) |
| 19:23 | <annevk> | see e.g. some posts from roc on what browsers from scratch would use |
| 19:23 | <annevk> | see e.g. some posts from roc on what browsers from scratch would use |
| 19:23 | <gsnedders> | annevk: Yes, perf and memory usage are directly opposite here |
| 19:24 | <gsnedders> | annevk: Yes, perf and memory usage are directly opposite here |
| 19:24 | <zewt> | i'd expect new implementations to use transparent optimization (eg. UCS-2 for strings that fit in it, expanding strings to UTF-32 when necessary, without application-level code having to know the difference) |
| 19:24 | <zewt> | i'd expect new implementations to use transparent optimization (eg. UCS-2 for strings that fit in it, expanding strings to UTF-32 when necessary, without application-level code having to know the difference) |
| 19:24 | <gsnedders> | annevk: UTF-8 undeniably has the lowest memory usage and the worst performance; UTF-32 has the highest memory usage and best performance. |
| 19:24 | <gsnedders> | zewt: Probably ASCII too. |
| 19:24 | <zewt> | "performance" is a bit more complex than that |
| 19:24 | <annevk> | yeah, best performance for what? |
| 19:24 | <zewt> | eg. if you're just copying and comparing strings (very common), lowest memory usage equals highest performance |
| 19:25 | <gsnedders> | annevk: UTF-8 undeniably has the lowest memory usage and the worst performance; UTF-32 has the highest memory usage and best performance. |
| 19:25 | <annevk> | certainly not for 16-bit code unit indexing :p |
| 19:25 | <gsnedders> | zewt: Probably ASCII too. |
| 19:25 | <zewt> | "performance" is a bit more complex than that |
| 19:25 | <annevk> | yeah, best performance for what? |
| 19:25 | <zewt> | eg. if you're just copying and comparing strings (very common), lowest memory usage equals highest performance |
| 19:25 | <annevk> | certainly not for 16-bit code unit indexing :p |
| 19:26 | <zewt> | another problem with utf-16 is that it leads to lots of subtle bugs when surrogates appear, since they're so rare people never test it |
| 19:27 | <zewt> | another problem with utf-16 is that it leads to lots of subtle bugs when surrogates appear, since they're so rare people never test it |
| 19:42 | <annevk> | so if you give |
| 19:42 | <annevk> | FE FF 00 7A D8 AA 00 7A 00 7A |
| 19:42 | <annevk> | to WebKit |
| 19:42 | <annevk> | it only displays a single "z" |
| 19:43 | <annevk> | so if you give |
| 19:43 | <annevk> | FE FF 00 7A D8 AA 00 7A 00 7A |
| 19:43 | <annevk> | to WebKit |
| 19:43 | <annevk> | it only displays a single "z" |
| 19:43 | <annevk> | but if you go through the code points, there's actually 4, and the second is D8AA, not FFFD |
| 19:43 | <annevk> | (it's FFFD in Gecko) |
| 19:44 | <annevk> | but if you go through the code points, there's actually 4, and the second is D8AA, not FFFD |
| 19:44 | <annevk> | (it's FFFD in Gecko) |
| 19:44 | <kennyluck_> | That's a silly bug in WebKit. It doesn't display any text run after an isolated surrogate. |
| 19:44 | <annevk> | Opera mimics Opera |
| 19:44 | <kennyluck_> | That's a silly bug in WebKit. It doesn't display any text run after an isolated surrogate. |
| 19:44 | <annevk> | Opera mimics Opera |
| 19:44 | <annevk> | euh Opera mimics Gekco |
| 19:44 | <annevk> | Gecko |
| 19:44 | <annevk> | geez |
| 19:44 | <Velmont> | Would be nice to mimic Opera though. |
| 19:44 | <Velmont> | Nice and recursive. |
| 19:44 | <Ms2ger> | Opera did it first |
| 19:45 | <annevk> | euh Opera mimics Gekco |
| 19:45 | <annevk> | Gecko |
| 19:45 | <annevk> | geez |
| 19:45 | <Velmont> | Would be nice to mimic Opera though. |
| 19:45 | <Velmont> | Nice and recursive. |
| 19:45 | <Ms2ger> | Opera did it first |
| 19:45 | <Velmont> | So it is recursive then. Just with one indirection :P |
| 19:45 | <Velmont> | So it is recursive then. Just with one indirection :P |
| 19:45 | <annevk> | Ms2ger: it's not like Gecko could up with something innovative like that |
| 19:46 | <annevk> | Ms2ger: it's not like Gecko could up with something innovative like that |
| 19:47 | <annevk> | and of course, IE is the same as Webkit, minus the silly bug |
| 19:47 | <Ms2ger> | Gecko is perfectly capable of coming up with innovative things |
| 19:47 | <Ms2ger> | Did I say innovative? I meant ridiculous |
| 19:48 | <annevk> | and of course, IE is the same as Webkit, minus the silly bug |
| 19:48 | <Ms2ger> | Gecko is perfectly capable of coming up with innovative things |
| 19:48 | <Ms2ger> | Did I say innovative? I meant ridiculous |
| 19:48 | <annevk> | man not even UTF-16 decoding is interoperable |
| 19:48 | <annevk> | what is wrong with this world |
| 19:48 | <annevk> | man not even UTF-16 decoding is interoperable |
| 19:48 | <Ms2ger> | Surprise! |
| 19:48 | <annevk> | what is wrong with this world |
| 19:48 | <Ms2ger> | Surprise! |
| 19:51 | <annevk> | oh, @bz_moz made a public appearance at SXSW |
| 19:51 | <annevk> | I wonder if there's a video of that |
| 19:52 | <annevk> | oh, @bz_moz made a public appearance at SXSW |
| 19:52 | <annevk> | I wonder if there's a video of that |
| 19:52 | <Ms2ger> | I hope so |
| 19:53 | <Ms2ger> | dbaron had an interesting talk too |
| 19:53 | <Ms2ger> | dbaron had an interesting talk too |
| 19:55 | <scott_gonzalez> | MikeSmith: 4 or 5 files should be fine, assuming they're not massive. |
| 19:55 | <scott_gonzalez> | I'll ping you tomorrow about it. Thanks. |
| 19:56 | <scott_gonzalez> | MikeSmith: 4 or 5 files should be fine, assuming they're not massive. |
| 19:56 | <scott_gonzalez> | I'll ping you tomorrow about it. Thanks. |