| 00:00 | <Hixie> | in prose, at least |
| 00:00 | <gsnedders> | Hixie: Web Encodings |
| 00:00 | <gsnedders> | Or rather Encoding as it seems to now be called. |
| 00:00 | <Hixie> | like, change it within the spec? isn't that up to anne? |
| 00:02 | gsnedders | wonders if <!doctype html><meta charset=UTF-8><title>t</title> is conforming |
| 00:02 | <gsnedders> | "Authors must use the utf-8 encoding and must use the "utf-8" label to identify it." |
| 00:02 | <gsnedders> | Hixie: From my reading of this, the normative name of the encoding is lowercase. |
| 00:03 | <gsnedders> | i.e., the encoding is called "utf-8" and not "UTF-8", so if you're calling it "UTF-8" you're doing it wrong |
| 00:03 | <Hixie> | gsnedders: i just define UTF-8 = utf-8 and move on with my life :-) |
| 00:04 | <gsnedders> | (As for that snippet, the label isn't per-se case-insensitive, so prima-facie <meta charset=UTF-8> is against that.) |
| 00:04 | <gsnedders> | (I also don't get what's wrong with <meta charset=utf8>) |
| 00:06 | <gsnedders> | Like, I'm perfectly fine with wanting to get rid of unicode-1-1-utf-8 as a label. But utf8 or utf-8 really doesn't make a damned bit of difference, and trying to force everyone to use one is likely just going to cause confusion. |
| 00:07 | <gsnedders> | annevk-cloud: ^^ |
| 00:21 | <SimonSapin> | gsnedders: The "get an encoding" (from a label) algorithm lower-cases the input before looking up |
| 00:22 | <SimonSapin> | is there anything that says authors should use the canonical "name" rather than any "label" ? |
| 00:23 | <gsnedders> | SimonSapin: Yes, that's getting an encoding from a label. |
| 00:23 | <gsnedders> | SimonSapin: That's not authors-must-use-this-label. |
| 00:24 | <SimonSapin> | is the latter a requirement? |
| 00:24 | <gsnedders> | "Authors must use the utf-8 encoding and must use the "utf-8" label to identify it." |
| 00:25 | <gsnedders> | The label as defined is that literal string of characters. Nothing says you're allowed to use that ASCII case-insensitively. Nothing says you're allowed to use the "utf8" label. |
| 00:25 | <SimonSapin> | hum, ok |
| 00:25 | <gsnedders> | To me, at least, that's fucking stupid. It makes huge numbers of documents non-conforming for I'm-not-sure-what-good. |
| 00:26 | <gsnedders> | s/what-good/what good/ |
| 00:26 | <SimonSapin> | agreed |
| 00:26 | <gsnedders> | Also the fact the encoding is now named "utf-8" when almost everywhere actually calls it "UTF-8" is a completely needless change. |
| 00:26 | <SimonSapin> | personally I just don’t care about being "conforming" to that kind of requirements, but meh |
| 00:30 | <SimonSapin> | data:text/html;charset=utf8,<script>document.write(document.characterSet)</script> is UTF-8 upper case in Gecko and Blink |
| 00:30 | <gsnedders> | Non-conforming! |
| 00:41 | gsnedders | starts writing JS and realizes how long it's been since he's written much JS to run in the browser |
| 00:41 | gsnedders | has spent too long in JS shells :) |
| 00:45 | <gsnedders> | Do data URIs in iframes load async? |
| 01:01 | <dekiss> | Hixie here? |
| 01:01 | <dekiss> | you gave me to work on one bug but I am confused :S |
| 01:01 | <dekiss> | I don't really understand what I need to do |
| 01:02 | <dekiss> | should I come up with a way for how can websites like caniuse.com be connected to the html spec? |
| 01:02 | <dekiss> | for checking browser's implementation of certain features |
| 01:19 | <gsnedders> | https://github.com/gsnedders/encoding-spec-names is a rough TC to check this |
| 01:48 | <dekiss> | is it ok to place <a> on the same level as div? |
| 01:51 | <dekiss> | google.com has 28 errors |
| 01:52 | <dekiss> | wikipedia 14 |
| 01:53 | <dekiss> | guys, can you depend on sites caniuse.com to officialy check what is the browsers implementation of certain features? |
| 01:53 | <dekiss> | I am not sure how serious is that |
| 01:53 | <dekiss> | I think the web lacks official "tool" for tracking browsers impementation of certain features |
| 02:05 | <Hixie> | dekiss: yeah, the idea is to have some sort of mechanism whereby the html spec can have automatically updated annotations saying which parts are implemented and which aren't |
| 02:06 | <Hixie> | dekiss: i don't mind if we rely on caniuse.com, or something else |
| 02:06 | <Hixie> | or even merge many things together |
| 02:07 | <dekiss> | I think this will be very nice |
| 02:07 | <dekiss> | oh I really think html lacks this so much and will be of much use |
| 02:08 | <dekiss> | Hixie but honestly I am not sure if it is good ieda to rely on some websites, I don't know who is behind those websites, if it is trsutable person-organization that it is ok I think |
| 02:14 | <Hixie> | dekiss: it's what web devs rely on already |
| 02:14 | <Hixie> | dekiss: if it turns out to be unreliable, we change it |
| 02:14 | <Hixie> | dekiss: better something than nothing |
| 02:14 | <Hixie> | dekiss: and better anything than what we have now |
| 02:15 | <dekiss> | Hixie you are right on this |
| 02:18 | <Domenic_> | <!--"SHALL", "SHALL NOT",--> hmm |
| 02:18 | <Domenic_> | do we not like those? |
| 02:38 | <SimonSapin> | Hixie: does HTML have a lot of ad-hoc syntaxes like this? http://www.whatwg.org/specs/web-apps/current-work/multipage/common-microsyntaxes.html#rules-for-parsing-dimension-values |
| 02:39 | <SimonSapin> | A Servo contributed was tempted to implement it with the CSS tokenizer… and I think it might be equivalent |
| 02:39 | <SimonSapin> | except for CSS comments |
| 02:40 | <SimonSapin> | So, would it be useful to you if CSS Syntax defined a "tokenize without comments" primitive, that other specs can use? |
| 02:40 | <SimonSapin> | TabAtkins: ^ |
| 02:40 | <SimonSapin> | (TabAtkins: didn’t SVG want this as well?) |
| 02:52 | <Hixie> | Domenic_: it tends to make me use the passive voice, so i avoid it |
| 02:52 | <Hixie> | Domenic_: i had some difficulties with spec text i'd written that used shall |
| 02:52 | <Hixie> | Domenic_: so i stopped using it |
| 03:24 | <Itprotj> | MikeSmith: Sorry, you were saying there was no documentation for the "Google-translate-customization" tag? |
| 03:29 | <Itprotj> | MikeSmith: If it helps this is the full tag I use <meta name="google-translate-customization" content="a940027e7f997750-6e3cf3faa4816ec1-g9f154a7bf834cf45-13" /> It provides a Google Translate box to choose different languages for the website |
| 03:32 | <dekiss> | are those custom meta tags part of the specification? |
| 03:32 | <Itprotj> | According to "http://wiki.whatwg.org/wiki/MetaExtensions"; the tag is in the status of Proposal |
| 03:33 | <dekiss> | hmm let me see |
| 03:33 | <Itprotj> | Which makes me think it should be validating correctly |
| 03:34 | <dekiss> | all are proposal |
| 03:35 | <dekiss> | I am not sure if it is good idea custom meta tags to be added to the specification |
| 03:36 | <dekiss> | because if Google for example decide to change the meta tag, or imagine 5000 companies decide to change the tags |
| 03:36 | <dekiss> | it willo involve a lot of work to change the specification and there might be constant work |
| 03:36 | <dekiss> | to track custom meta tags changes |
| 03:37 | <Itprotj> | What would my options be then? |
| 03:37 | <dekiss> | how you mean |
| 03:37 | <dekiss> | well browser implement then unnoficially |
| 03:37 | <dekiss> | they are like semi official |
| 03:38 | <dekiss> | I mean widely suported it is publicly known, but officially not in the HTML spec |
| 03:38 | <dekiss> | I think it is ok for you to use anything that is out of the html specification, but you should be aware that not all browsers might support it |
| 03:42 | <Itprotj> | Okay it is just a concern that I would like to stay HTML5 validated whilst also keeping a Translate box for overseas viewers |
| 03:50 | <dekiss> | I think that is good, I will try to make 100% valid website in the next few days |
| 03:51 | <dekiss> | however in the past I think I was inable to do so, because it was pretty much immposible for me to do some things which I wanted to make, without making the site not 100% valid |
| 03:52 | <dekiss> | I really think there should be massive campaign to influence developers to drop support for old browsers and bad browsers |
| 03:52 | <dekiss> | this came to ridicilous level now |
| 03:52 | <dekiss> | it is not a way to go, I think |
| 03:53 | <dekiss> | ideally browsers makers should make the browsers to autoupdate |
| 03:53 | <dekiss> | also Microsoft including Internet Explorer in the Windows, which had many BIG mistakes it was so sad |
| 03:54 | <dekiss> | finally Internet Explorer 11 is ok I think, at least I hear so and my pages look ok on it |
| 03:55 | <dekiss> | there should not be 1000 libraries to make this job easy to developers, this is not a way to go, it is slowing the internet progress. It should be done other way. Also I am against any libraries |
| 03:55 | <dekiss> | like jQuery |
| 03:56 | <dekiss> | recently I want a lot of developers who don't know Javascript or DOM but know jQuery |
| 03:56 | <dekiss> | heh |
| 03:56 | <dekiss> | not a way to go |
| 04:00 | <dekiss> | I don't want to be understood wrongly but I must say this, I think the guy who makes jQuery is really good programmer talented, but to dedicate life to something that is not the right way to do it I can't understand, I really think he should be involved in html, css, ecmascript development |
| 04:00 | <dekiss> | and other things that matter |
| 04:00 | <dekiss> | I can't work in something that I don't believe it is a good thing to do |
| 04:00 | <dekiss> | and will do world and people good - progress |
| 04:33 | <dekiss> | I make this statements after I see 9 from 10 job posts on freelance websites are - Company offers position for Wordpress plugin fix developper |
| 04:33 | <dekiss> | :D |
| 07:27 | Ms2ger | should get foolip some chocolate too :) |
| 07:28 | <Ms2ger> | <!DOCTYPE HTML> <script>document.write('<meta charset="ISO-8859-' + '2">')</script> |
| 07:28 | <Ms2ger> | Ugh. |
| 07:29 | <annevk-cloud> | Where is that from? |
| 07:30 | <Ms2ger> | https://critic.hoppipolla.co.uk/r/592 |
| 07:31 | <annevk-cloud> | I cannot access that if not signed into GitHub. Curious |
| 07:34 | <annevk-cloud> | Oh gsnedders, I recommend filing a bug rather than ranting on IRC. You might find it more relaxing ;) |
| 07:35 | annevk-cloud | wonders what foolip did |
| 07:56 | <foolip> | annevk-cloud: not me! |
| 08:00 | <Ms2ger> | annevk, reviews |
| 08:02 | <Ms2ger> | wefo, I'm sure the spec explains what clearRect does: whatwg.org/html |
| 08:14 | <wefo> | Not clearly. |
| 08:17 | <TabAtkins> | SimonSapin: SVG was going to see if they can just allow comments, and thus use the full CSS tokenizer. (They probably can.) |
| 08:17 | <Ms2ger> | TabAtkins, this is HTML, though |
| 08:17 | <TabAtkins> | SimonSapin: HTML *might* also be able to accept comments, but I'd also be fine with defining a tokenizer mode that disallowed comments. |
| 08:18 | <TabAtkins> | (Or rather, that emitted the / and * as delims.) |
| 08:18 | <TabAtkins> | Ms2ger: Right, but I'm unsure if there'd actually be any compat requiring /**/ comments to invalidate the attribute or whatever. |
| 08:20 | <Ms2ger> | Otoh, it seems like a lot of spec/impl/test work that would use time that could be better spent elsewhere |
| 08:21 | <TabAtkins> | GPHemsley: Yes, they *should* increment the list-item counter. That said, wk/b do *not* implement that counter. We don't actually use CSS Counters at all for ul/ol, because they're too slow. |
| 08:22 | <TabAtkins> | Ms2ger: Right, that's certainly true, and it would be easy on my side to avoid it by defining a comment-free mode. I suspect that with that change, switching to using the CSS tokenizer rather than specialized microsyntaxes would be very little change, if any. (But I'd have to actually look at the spec text, which I haven't done yet.) |
| 08:23 | <Ms2ger> | Well, implementing a comment-free mode is still Work(tm) |
| 08:23 | <Ms2ger> | That'd reduce testing time and possible web compat, but increase QA time to verify that nothing actually changed |
| 08:25 | <Ms2ger> | Anyway :) |
| 08:26 | Ms2ger | wonders if Operians will come to fosdem |
| 08:28 | <Ms2ger> | TabAtkins, "if i have 200x100 <img>; if i set size if img to 400; this case what is the height?; 200? 100?" |
| 08:29 | <Ms2ger> | Looks like it's 200, but do you have a spec pointer? :) |
| 08:38 | <MikeSmith> | jgraham: doing the last bit of that PR now |
| 08:41 | <Ms2ger> | TabAtkins, I think http://dev.w3.org/csswg/css-images/#default-sizing |
| 08:51 | <MikeSmith> | jgraham: done |
| 08:51 | <MikeSmith> | you want me to merge it? |
| 08:59 | <MikeSmith> | marcosc: curious about what's "background actor stuff" |
| 08:59 | <marcosc> | MikeSmith: more context? |
| 08:59 | <marcosc> | not sure what you mean |
| 09:05 | <jgraham> | MikeSmith: Sure and thanks |
| 09:08 | <marcosc> | MikeSmith: oh, you mean in the roadmap |
| 09:08 | marcosc | had to google "background actor stuff" |
| 09:08 | <marcosc> | :) |
| 09:09 | <marcosc> | MikeSmith: we are going to write a blog post to explain why we are doing what we are doing, I think. There is a Moz workweek happening next week, so I'm sure a lot that stuff will get clarified and updated. |
| 09:10 | <Ms2ger> | Ah, the jst work week |
| 09:10 | <marcosc> | yep |
| 09:12 | <annevk-cloud> | On my way to Heathrow for that and some informal service worker thingie this week |
| 09:12 | <Ms2ger> | Good flight |
| 09:12 | <annevk-cloud> | To prevent jet lag I am already tired |
| 09:13 | <annevk-cloud> | Thanks Ms2ger |
| 09:14 | <MikeSmith> | marcosc: ok |
| 09:29 | <zcorpan> | "Tip for admins: instead of sub1.example.com use sandbox.sub1.example.com, which will limit impact of the cookie bomb to .sub1.example.com zone." is incorrect, isn't it? http://homakov.blogspot.fi/2014/01/cookie-bomb-or-lets-break-internet.html |
| 09:40 | <annevk-cloud> | Yeah, we use public suffix for Effective TLD determination |
| 11:25 | <Lachy_> | ffffffn9#\M:gI |
| 11:27 | <marcosc> | looks like Lachy_ passed out on his keyboard ... again |
| 11:51 | <MikeSmith> | he typed the hex code for black there first |
| 11:55 | <MikeSmith> | or maybe he was writing a self description and he only got as far as the f in "foxy" |
| 11:55 | <Lachy> | WTF? How did that happen? |
| 11:57 | <MikeSmith> | wait it seems like he is actually typing an autobiography |
| 11:57 | <jgraham> | Lachy: Either you say on your keyboard (in which case we won't ask), or you are hurredly changing your bank password ;) |
| 11:59 | <jgraham> | *sat |
| 12:00 | <Lachy> | It's my login password for my computer. Not sure how it got entered into here. My computer seemed was frozen at the time I tried entering it. |
| 12:00 | <Ms2ger> | Or you blame the cat |
| 12:00 | <Lachy> | at least, part of it is |
| 12:04 | <MikeSmith> | I think most security experts recommend against using IRC to log in to your computer |
| 12:04 | <jgraham> | It turns out that most "security experts" are NSA shills, so you should probably do the opposite of what those guys say |
| 12:05 | <Ms2ger> | Oh, okay |
| 12:05 | <Ms2ger> | hunter2 |
| 12:07 | <darobin_> | rofl |
| 12:18 | <gsnedders> | annevk-cloud: Nah, ranting on IRC is more relaxing. Less productive, but more relatxing. |
| 12:18 | <gsnedders> | annevk-cloud: I will file a bug, however :) |
| 12:20 | <wefo> | jgraham: It depends on what advice they give. |
| 12:20 | <wefo> | If they give the advice to use the "cloud" in any manner, they are NSA shills. |
| 12:20 | <wefo> | If they give the advice to smack every cellphone you see, they are not. |
| 12:21 | <wefo> | Just the thought that my computer might have shipped with custom hardware/firmware makes me angry. |
| 12:21 | <wefo> | It's so dishonest. |
| 12:22 | <wefo> | If at least it was snooping on plaintext communication over the network. |
| 12:22 | <wefo> | Then they would just be cunts. |
| 12:22 | <wefo> | Now they are below subhuman. |
| 12:24 | <jgraham> | wefo: You might accidentially have taken something I said seriously |
| 12:26 | <wefo> | Yeah, silly me for not laughing everything away. |
| 12:26 | <wefo> | I guess I meant to say: "ya lol dat b dope ur a tinfoil hat lulz u fink ur so importent lol r u high dawg" |
| 12:27 | <Ms2ger> | Good good |
| 12:42 | <MikeSmith> | somebody forgot to ring the bell for the 4001st w-p-t commit |
| 12:42 | Ms2ger | rings |
| 13:22 | <zcorpan> | heycam|away: pls fix https://www.w3.org/Bugs/Public/show_bug.cgi?id=22646 kthxbye |
| 13:24 | <zcorpan> | MikeSmith: how is ffffff black? |
| 13:25 | <MikeSmith> | because it is? |
| 13:25 | <MikeSmith> | either that or it's white |
| 13:25 | <zcorpan> | it's white in css at least |
| 13:26 | <MikeSmith> | clearly I meant white |
| 13:26 | <MikeSmith> | read between the lines, man |
| 13:26 | <MikeSmith> | that's what the NSA pays you for!! |
| 13:27 | <jgraham> | I can see that MikeSmith will get himself killed at the next Zebra crossing |
| 13:27 | <zcorpan> | yeah but i also get paid for acting like i don't |
| 13:32 | <xvoom> | hi everybody |
| 13:48 | Ms2ger | wonders why Google thinks the title of http://hoppipolla.co.uk/410/results.html is "Web Platform Tests Results - Mozilla XPath Documentation" |
| 13:54 | <zcorpan> | oops, looks like i missed to git add a helper file to resolve-url |
| 14:02 | <zcorpan> | https://critic.hoppipolla.co.uk/r/597 |
| 14:04 | <jgraham> | Hmm, so Ubuntu decided I wanted to open http URLs in Thunderbird |
| 14:04 | <jgraham> | That didn't work so well |
| 14:19 | <jgraham> | zcorpan: What was the final word on snffing video? |
| 14:19 | <zcorpan> | jgraham: hah |
| 14:19 | <jgraham> | :( |
| 14:19 | <zcorpan> | there is no spoon |
| 14:20 | <jgraham> | So you have to get the Content-Type header right to ensure that a supporting browser will play the video at the moment? |
| 14:20 | <jgraham> | (also: I did your review) |
| 14:35 | <zcorpan> | right |
| 14:36 | <zcorpan> | yep, addressed it. now you have MORE lines to review. might be recursive |
| 14:43 | <jgraham> | zcorpan: Not recursive at all :) Merged |
| 14:44 | <zcorpan> | great, thanks! |
| 17:09 | <jgraham> | Dear spotify: When your "discover" feature shows that you have a good grasp of the kinds of music I like, why are you wasting prime screen estate advertising playlists featuring Milley Cirus and One Direction at me? Also, why are you advertising at all when I am paying you money not to show me adverts (other than via the aforementioned "Discover" feature, which is actually useful and so doesn't count). |
| 17:15 | <gsnedders> | jgraham: Because how can one not love One Direction and Milley Cirus? They are obviously the best thing ever. |
| 18:26 | <JonathanNeal> | SteveF, http://www.jonathantneal.com/blog/introducing-subhead/ |
| 19:00 | <TabAtkins> | jgraham: Because spotify has learned the lessons of cable TV, which is that people forgot that they're paying you and accept advertising anyway? |
| 19:05 | <SteveF> | JonathanNeal: cool will read |