| 02:40 | <JonathanNeal> | Is there a native function like instanceof that works across windows? |
| 02:50 | <caitp> | Array.isArray comes to mind |
| 02:50 | <caitp> | but I can't think of anything else |
| 02:50 | <caitp> | unless we have Map.isMap() and friends now |
| 03:18 | <JonathanNeal> | caitp: I was curious because checking the class of something comes up a lot in polyfills, but writing an abstraction library for this gets tiresome and makes the scripts less readable. |
| 03:20 | <caitp> | well, you could use Symbol.toStringTag for a not-super-thorough test |
| 03:21 | <caitp> | but I think in general the goal is to avoid testing the class of objects |
| 03:27 | <JonathanNeal> | We have polyfills that require an argument to be a function. We want to ensure that without someOtherWindowsFn instanceof Function failing. |
| 03:28 | <caitp> | couldn't you just `ASSERT(typeof fn === "function")` |
| 03:28 | <caitp> | ? |
| 03:29 | <JonathanNeal> | typeof /regex/ is unfortunately ‘function’ in some older browsers. |
| 03:30 | <caitp> | well I guess you'd need like |
| 03:30 | <caitp> | an old-crazy-ua-specific path for that |
| 03:31 | <caitp> | if you wanted to support those |
| 03:43 | <JonathanNeal> | ua? that would be unreliable. |
| 05:56 | <Hixie> | roc: hey, the case i was talking about recently where firefox has been violating the spec for years after making something up unilaterally because the previous spec was disliked is now preventing chrome from following the spec :-) |
| 05:59 | <roc> | HTML menus? |
| 05:59 | <Hixie> | yeah |
| 05:59 | <Hixie> | see sanjay's mail to whatwg just now |
| 05:59 | <Hixie> | not clear how many sites are affected. maybe we'll have to use another element |
| 06:01 | <roc> | looks minor to me ... but what do you want me to do about it? |
| 06:02 | <Hixie> | well it means we probably won't be able to use <menuitem>, which means we can't do what mozilla wanted |
| 06:02 | <Hixie> | i mean i only made it <menuitem> in the first place because of mozilla's feedback |
| 06:03 | <roc> | why can't we make it a non-void element? |
| 06:03 | <Hixie> | because that's a poor authoring aethestic |
| 06:04 | <roc> | OK, but that still sounds minor to me |
| 06:04 | Hixie | shrugs |
| 06:04 | <Hixie> | that's what i said about the things you were complaining about in that same thread too |
| 06:04 | <Hixie> | i guess it's all relative |
| 06:05 | <roc> | I do not see PNaCl + Pepper are in the same league as this |
| 06:06 | <Hixie> | i think this is worse, but ok |
| 06:06 | <roc> | "invent an entirely new execution environment and platform API" vs "make <menuitem> not void" |
| 06:06 | <Hixie> | this actively breaks interoperability of a published spec |
| 06:06 | <Hixie> | the other one is just an experimental technology |
| 06:06 | <Hixie> | that doesn't break anything |
| 06:07 | <roc> | which is shipping, on by default, and actively evangelized |
| 06:07 | <Hixie> | sure |
| 06:07 | <Hixie> | so's linux |
| 06:07 | <Hixie> | so what? |
| 06:08 | <Hixie> | all vendor-specific platforms "invent an entirely new execution environment and platform API" |
| 06:09 | <Hixie> | i mean, PNaCl is no worse than, say, ActiveX |
| 06:09 | <Hixie> | or iOS |
| 06:10 | <Hixie> | (indeed in many ways it's better, since it's open source) |
| 06:10 | <Hixie> | (and hardware-agnostic) |
| 06:11 | <roc> | it's available for Web pages to use. |
| 06:11 | <Hixie> | so was ActiveX |
| 06:11 | <roc> | so you think that ActiveX was a fine thing? |
| 06:11 | <Hixie> | so are Windows .exes or linux Elfs |
| 06:11 | <Hixie> | i think these are all proprietary technologies, and thus all irrelevant to the Web |
| 06:12 | <Hixie> | anyone who decides to target a proprietary technology deserves what they get when they find themselves with a limited market or when the vendor discontinues the product |
| 06:12 | <Hixie> | or at least, should expect what they get, i dunno about "deserves" |
| 06:13 | <Hixie> | though again, in the case of PNaCl I guess it's open source so even if the vendor abandons it it's not necessarily dead |
| 06:14 | <Hixie> | in contrast, <menuitem> has been shipping in violation of the spec for years and now is preventing chrome from implementing a spec |
| 06:14 | <Hixie> | that seems relatively worse |
| 06:14 | <roc> | back in the day we worked pretty hard to oppose the spread of ActiveX into the Web, in particular by not supporting it even when it would have been expedient to do so. All in the name of Web standards. I'm shocked by this revisionism coming from you. |
| 06:16 | <Hixie> | what revisionism? the problem with ActiveX has always been that it's a single-vendor technology stack. That's what the difference between ActiveX and Web Standards _is_ |
| 06:17 | <Hixie> | this is consistent with e.g. what i wrote in 2007: http://lists.w3.org/Archives/Public/public-html/2007Apr/0632.html |
| 06:18 | <roc> | I don't think you can simultaneously agree that ActiveX was bad for the Web but PNaCl is just fine. |
| 06:18 | <roc> | maybe you're not even trying to do that |
| 06:18 | <Hixie> | i'm saying that ActiveX wasn't the Web |
| 06:19 | <Hixie> | it's a proprietary platform like .NET or Win32 or PNaCl or iOS |
| 06:19 | <roc> | it's not so easy to draw that line. |
| 06:19 | <Hixie> | one vendor in control of technology = proprietary |
| 06:19 | <Hixie> | that's a pretty easy line to draw |
| 06:20 | <roc> | can a vendor just add a new DOM API, say "that's proprietary" and we should all be comfortable with that? |
| 06:20 | <Hixie> | we shouldn't be comfortable with any proprietary platform |
| 06:20 | <Hixie> | are you comfortable with iOS? |
| 06:20 | <roc> | no |
| 06:21 | <roc> | hang on |
| 06:21 | <roc> | in our previous discussion I was arguing that asm.js was a standards-based alternative to PNaCl and therefore more desirable |
| 06:21 | <roc> | are you agreeing with that now? |
| 06:22 | <roc> | anyway, I gotta go home. If there's something specific you want Mozilla to do about <menuitem> let me know. |
| 06:22 | <Hixie> | well there's two questions there. There's "which technology do I think should become a standards-based multi-vendor technology", and there's some question about a standard vs a proprietary platform |
| 06:22 | <Hixie> | not sure exactly what the second question is |
| 06:23 | <Hixie> | for the first one, i think if the choices are asm.js or PNaCl, I don't have an opinion on which is better technology. |
| 06:23 | <Hixie> | i would be fine with PNaCl being a multi-vendor-supported standard. |
| 06:23 | <Hixie> | ditto asm.js. |
| 08:58 | <foolip> | annevk: what are your thoughts on the "WHATWG/W3C collaboration proposal" thread, given that your spec is the suggested test balloon? |
| 09:16 | <annevk> | foolip: I don't know, but I'm happy for rubys to try it out |
| 09:18 | <foolip> | ok, just curious |
| 10:18 | <SimonSapin> | darobin: I’m confused, is https://specs.webplatform.org/url/webspecs/develop/ a mirror, a snapshot, a fork? |
| 10:19 | <darobin> | SimonSapin: annevk and rubys have some agreement to try to work out something on neutral ground; that's the neutral ground |
| 10:20 | <SimonSapin> | hum, ok |
| 10:20 | <annevk> | afaik rubys uses that because he thinks Microsoft might review that |
| 10:20 | <annevk> | I don't have anything to do with it |
| 10:20 | <SimonSapin> | seems very meta-xkcd-927 at first |
| 10:20 | <annevk> | yeah I'm not a fan |
| 10:21 | <annevk> | but rubys seems to believe in hosting it in as many places as possible... and there's nothing really preventing that from happening |
| 10:23 | <darobin> | SimonSapin: see 927 in https://specs.webplatform.org/docs/ |
| 10:25 | <SimonSapin> | darobin: reading. I landed on https://specs.webplatform.org/ from Twitter, and didn’t find /docs/ at first |
| 10:25 | <darobin> | SimonSapin: yeah, I need to make the link far more prominent |
| 10:26 | <SimonSapin> | I’m curious how rubys thinks that publishing in many places helps |
| 10:26 | <darobin> | I think annevk may have been exaggerating the "as many places as possible" |
| 10:27 | <SimonSapin> | I’m still curious how rubys thinks that publishing in more than one place helps |
| 10:27 | <annevk> | darobin: well his personal site, WHATWG, webplatform.org, w3.org, presumably some github.w3.org, ... |
| 10:28 | <caitp> | ubiquity is its own reward |
| 10:28 | <SimonSapin> | darobin: what does "transversality theorem me" mean? |
| 10:28 | <darobin> | annevk: you'll have to see with him, but I don't think there's anything on a GH site or on w3.org, and I'm guessing it's getting off his personal site |
| 10:29 | <darobin> | SimonSapin: huwah? |
| 10:29 | <SimonSapin> | darobin: or "pitchfork me", to take the Unicode name |
| 10:29 | <darobin> | oooooooh |
| 10:29 | <darobin> | "fork me" :) |
| 10:29 | <darobin> | yeah, that amused me but it should probably go |
| 10:29 | <SimonSapin> | https://en.wikipedia.org/wiki/⋔ redirects to https://en.wikipedia.org/wiki/Transversality_theorem |
| 10:30 | <darobin> | though I like your reading |
| 10:30 | <darobin> | yes yes, I had seen that initially |
| 10:30 | <SimonSapin> | though I have no idea what that is |
| 10:30 | <SimonSapin> | “Let f:X\rightarrow Y be a smooth map between manifolds […]” |
| 10:31 | <jgraham> | SimonSapin: Isn't that just an explaination of the Git data model? ;) |
| 10:32 | <SimonSapin> | probably :) |
| 10:35 | <darobin> | lol |
| 12:06 | <Ms2ger> | jgraham++ |
| 12:18 | <rubys> | annevk: please let me know if you disagree with (or feel that anything important is missing) from my recent email to public-w3process (which oddly hasn't shown up in the web archives yet). |
| 12:27 | <annevk> | rubys: not the email from 18min ago to www-archive? |
| 12:31 | <rubys> | ah, that would explain why it didn't show up in w3-process :-) |
| 12:32 | <rubys> | yes: http://lists.w3.org/Archives/Public/www-archive/2014Nov/0056.html |
| 12:33 | <annevk> | yeah seems fine |
| 12:33 | <rubys> | cool |
| 12:37 | <rubys> | TBH, I don't plan to work on it for a few days, as process/politics wears me down and I need a few days of coding to recharge my batteries. :-) |
| 12:42 | <annevk> | I wonder what https://github.com/domokit/mojo is |
| 12:42 | <annevk> | rubys: tell me about it :-( |
| 12:42 | <annevk> | rubys: I hope the next time we can discuss something technical |
| 12:43 | <rubys> | I should have a draft merge done in the next few days. I'm sure that there will be a lot broken in the process. |
| 12:44 | <rubys> | In fact, I plan to have it today, unless my wife has other plans for me :-) |
| 12:44 | <annevk> | rubys: there's still a couple of outstanding bikeshed regressions too |
| 12:44 | <rubys> | Perhaps it would be a good idea to open some tracking bugs on these? |
| 12:45 | <annevk> | rubys: people did |
| 12:45 | rubys | checking |
| 12:45 | <annevk> | rubys: https://www.w3.org/Bugs/Public/show_bug.cgi?id=27383 and https://www.w3.org/Bugs/Public/show_bug.cgi?id=27368 |
| 12:46 | <rubys> | Thanks! I'll fix both. |
| 12:49 | <annevk> | What password manager do people use? Is LastPass any good? |
| 12:53 | <iuuso> | i'm using lastpass currently, and its usability is certainly great but i'm planning on switching over to KeePass. |
| 12:55 | <iuuso> | if you want to store your passwords in a cloud where you have no control over them then lastpass is fairly good for you :) |
| 12:57 | <iuuso> | but i would suggest to take a glance at keepass. from the security perspective it's considered (as far as i know) safer than most common alternatives. |
| 13:47 | <foolip> | annevk: I use http://www.passwordstore.org which is backed by GnuPG and Git |
| 13:49 | <foolip> | storing the Git repo in public would reveal which passwords you have but not what they are, but to avoid both I just keep it all out of the "cloud" |
| 14:30 | <hsivonen> | is there no search for the current WebKit source tree (as opposed to searching the changesets?) |
| 14:30 | <annevk> | hsivonen: perhaps through https://github.com/WebKit/webkit |
| 14:32 | <caitp> | github's search tools are a bit lacking |
| 14:32 | <caitp> | iirc they have a source browser online, but without search utilities |
| 14:32 | <caitp> | so you can often make use of mxr or similar for blink and try to dig up stuff in webkit |
| 14:33 | <annevk> | hsivonen: https://github.com/WebKit/webkit/blob/master/Source/WebCore/platform/text/TextEncoding.cpp#L105 |
| 14:42 | <hsivonen> | annevk: a quick look suggests that WebKit agrees with IANA case and Gecko agrees with IANA case except for gbk and gb18030 |
| 14:48 | <hsivonen> | annevk: I fail to see value in changing the case where Gecko, WebKit and Blink agree even if Trident doesn't |
| 14:49 | <annevk> | I can't find the issue where this was originally discussed |
| 14:49 | <annevk> | hsivonen: does this mean we should have inconsistent results for TextEncoder.prototype.encoding and characterSet? |
| 14:50 | <annevk> | hsivonen: I guess I can add some mapping table to DOM for characterSet |
| 15:09 | <annevk> | hsivonen: if you could give feedback on https://www.w3.org/Bugs/Public/show_bug.cgi?id=27190 that'd be great btw |
| 15:14 | <Domenic> | annevk: I <3 LastPass |
| 15:42 | <wanderview> | JakeA: fyi, TrainedToThrill is mostly working on our SW project branch as of today... still some issues to sort out before we can get it into nightly behind a pref, though |
| 16:04 | <Domenic> | wanderview: \o/ |
| 16:04 | <JakeA> | whoa that's really cool |
| 16:05 | <wanderview> | and our nightly builds for the branch are broken, but I'm trying to fix |
| 16:07 | <Domenic> | annevk: thoughts on https://github.com/whatwg/resources.whatwg.org/issues/4#issuecomment-64668094 ? I'm not a big fan of inlining CSS into my JS files but will do it if you think it's best. |
| 16:09 | <annevk> | Domenic: let's inline until everyone does bikeshed |
| 16:10 | <Domenic> | sgtm |
| 16:22 | <Domenic> | annevk: it works on https://notifications.spec.whatwg.org/ \o/ |
| 16:49 | <annevk> | Hmm, does ArrayBuffer require new or not? |
| 16:49 | <annevk> | Throws in Chrome, does not throw in Firefox or IE |
| 16:50 | <annevk> | Or perhaps only Uint8Array throws |
| 16:51 | <Hunter14895> | I hope I am not outta my realm here (newbie) have some rather general questions. |
| 16:52 | <Hunter14895> | Is a grid absolutely necessary as I attempt my hand as responsive web designs? |
| 16:54 | <hemanth> | annevk, it can be with or without new |
| 16:55 | <hemanth> | new ArrayBuffer( ...argumentsList ) or ArrayBuffer( length ) |
| 16:56 | <annevk> | hemanth: seems to throw in Safari/Chrome |
| 16:57 | <hemanth> | hmm, not yet implemented fully? |
| 17:00 | <Hixie> | ooh, i like zcorpan's ideo |
| 17:07 | <SteveF_> | Hunter14895: suggest asking on irc://irc.freenode.net/html5 |
| 17:08 | <Domenic> | annevk: hemanth: per spec it should throw. |
| 17:08 | <Hunter14895> | @SteveF_ thanks bud |
| 17:13 | <annevk> | Domenic: hmm, I thought we recently changed our ArrayBuffer behavior, weird that it would still not align with the spec |
| 17:13 | annevk | prepares a fix for web-platform-tests |
| 17:16 | <annevk> | MikeSmith: https://github.com/w3c/web-platform-tests/pull/1419 |
| 17:16 | MikeSmith | looks |
| 17:20 | <MikeSmith> | annevk: I see now that test as already not executing as expected in Chrome, for that very reason |
| 17:20 | <MikeSmith> | Uncaught TypeError: Constructor ArrayBuffer requires 'new' |
| 17:20 | <annevk> | MikeSmith: yes |
| 17:21 | <annevk> | MikeSmith: I wonder if this changed recently |
| 17:21 | <MikeSmith> | maybe so |
| 17:22 | <MikeSmith> | I also wonder who now reviewed this initially.. |
| 17:22 | <annevk> | zcorpan I think |
| 17:22 | <annevk> | thought it was you for a second |
| 17:22 | <annevk> | zcorpan: ^^ |
| 17:23 | <annevk> | whoa, I wonder how https://github.com/w3c/web-platform-tests/graphs/contributors happened |
| 17:23 | <MikeSmith> | oh man I get some crazy behavior in chrome now when it does run |
| 17:23 | <annevk> | what did I do between 2010 and 2012 |
| 17:24 | <MikeSmith> | my chrome is trying to download dozens of single-byte-raw.py files to my machine |
| 17:24 | <MikeSmith> | annevk: that must because you updated your e-mail settings in github |
| 17:25 | <annevk> | MikeSmith: I did yeah, was curious what that would do |
| 17:25 | <annevk> | MikeSmith: heh |
| 17:26 | <Ms2ger> | Oh, I missed my 1000th commit |
| 17:26 | <MikeSmith> | annevk: looks like you made tons of changes to an Overview.src.html file |
| 17:27 | <MikeSmith> | oh, the dom spec |
| 17:27 | <annevk> | MikeSmith: maybe the old ranges repository? |
| 17:27 | <MikeSmith> | must be |
| 17:27 | <Ms2ger> | I imported the entire DOM history at one point |
| 17:27 | <MikeSmith> | ah |
| 17:29 | <MikeSmith> | if somebody else can try http://w3c-test.org/submissions/1419/encoding/single-byte-decoder.html to confirm I'm not insane, that'd be great |
| 17:30 | <MikeSmith> | christ |
| 17:30 | <Ms2ger> | 262 Pass |
| 17:30 | <Ms2ger> | 95 Fail |
| 17:30 | <Ms2ger> | 144 Timeout |
| 17:31 | <MikeSmith> | Safari is doing the download thing too |
| 17:32 | <MikeSmith> | Ms2ger: yeah I don't see that. I see instead "This site is attempting to download multiple files. Do you want to allow this?" |
| 17:33 | <MikeSmith> | and if I say yes all hell breaks loose |
| 17:33 | <Ms2ger> | Not here |
| 17:33 | <MikeSmith> | I wonder if it's some OS X thing |
| 17:33 | <MikeSmith> | Ms2ger: you not running on a mac? |
| 17:34 | <Ms2ger> | Nope, fortunately not :) |
| 17:35 | <MikeSmith> | heh |
| 17:36 | <MikeSmith> | webkit doesn't even prompt me it just downloads 237 "Unconfirmed NNNNNNN.crdownload" files |
| 17:39 | <Hixie> | man, the threads that anne are on about the w3c process are just more evidence of one of my least favourite parts about the w3c: the existence of process |
| 17:39 | <MikeSmith> | r+ed that change and merged it because clearly it's the right thing but overall the behavior of the test is whacked in chrome and webkit on OS X |
| 17:43 | <annevk> | Hixie: :-( |
| 17:45 | <Hixie> | annevk: you should have listed the existence of process in your list of things that should change ;-) |
| 17:46 | <annevk> | Hixie: in the list of things I was unhappy with, W3C Process was the first if I remember correctly |
| 17:46 | <Hixie> | aah. i read it as just it being the precise process you didn't like, not any process at all |
| 17:46 | <annevk> | Hixie: there's always some kind of process |
| 17:47 | <annevk> | Hixie: e.g. I have to file bugs and make sure I'm not too brief if I want to change HTML :p |
| 17:51 | <Hixie> | annevk: that's not a process. i mean, you and i could come up with a better way to do it if you wanted. like, e-mail, or github issues, or whatever we found more effective. |
| 17:51 | <Hixie> | annevk: a process is something that is expensive to change that isn't just what people want to do. |
| 17:52 | <Ms2ger> | Okay, I'd like a process |
| 17:52 | <Ms2ger> | One that stopped people from using github issues |
| 17:52 | <annevk> | Hixie: it's true that things are easily changed |
| 17:52 | <annevk> | Hixie: I wasn't aware of that constraint on usage of the term |
| 17:53 | <Hixie> | "process" in this kind of context usually implies some sort of fixed set of steps or rules that are non-trivial to change |
| 17:53 | <Hixie> | as opposed to "personal preference" |
| 17:54 | <Hixie> | i mean, anyone can have their own personal process, sure, but it's not a Formal Process if everyone else can just ignore it :-) |
| 18:14 | <hsivonen> | annevk: yeah, if we don't want to require changes to characterSet in WebKit, Blink and Gecko, the result would have to be inconsistent with TextDecoder returning in lower case |
| 18:17 | <wanderview> | JakeA: do we have a spec issue for supporting Cache on main thread and other workers? I want to add a note |
| 18:18 | <annevk> | hsivonen: I don't really care either way, but I guess I can make such a change for characterSet, it does seem safer |
| 18:18 | <annevk> | hsivonen: but I guess Gecko should still change gbk / gb18030 then, if we want to make the compatibility names consistent |
| 18:19 | <wanderview> | we won't be able to support Cache in workers based on data:uris or created from a sandboxed iframe... although maybe those are gecko specific things |
| 18:22 | <JakeA> | wanderview: https://github.com/slightlyoff/ServiceWorker/issues/535 https://github.com/slightlyoff/ServiceWorker/issues/297 |
| 18:22 | <Ms2ger> | gsnedders, so is being named after a fish a requirement for Scottish first ministers? |
| 18:23 | <wanderview> | thanks... I searched but couldn't find them for some reason |
| 18:23 | <wanderview> | JakeA: ah... the second one there does not have the cache github issue tag |
| 18:34 | <JakeA> | wanderview: fixed, cheers! |
| 18:34 | <wanderview> | thanks! |
| 19:53 | <rubys> | TabAtkins: did something change recently in bikeshed? Building the whatwg URL spec, I'm now getting: |
| 19:53 | <rubys> | FATAL ERROR: The selector '#hosts-(domains-and-ip-addresses)' returned an error: |
| 19:53 | <rubys> | Expected selector, got <DELIM '(' at 7> |
| 19:53 | <rubys> | The input line in question is: |
| 19:53 | <rubys> | <h2 id="hosts-(domains-and-ip-addresses)">Hosts (domains and IP addresses)</h2> |
| 19:54 | <TabAtkins> | Interesting. This wasn't a problem before, with that same ID? |
| 19:54 | <rubys> | no |
| 19:54 | <TabAtkins> | (the issue is that I'm not doing sufficient escaping to make all IDs valid id selectors) |
| 19:55 | <rubys> | https://url.spec.whatwg.org/#hosts-%28domains-and-ip-addresses%29 |
| 19:56 | <rubys> | workaround: change the source to do the percent encoding? |
| 19:56 | <TabAtkins> | Hm, I must have made some change. Let me bisect. |
| 19:56 | <TabAtkins> | Don't workaround, just gimme a sec. |
| 19:56 | <rubys> | k |
| 19:58 | <gsnedders> | Ms2ger: nah, only SNP leaders. I'm not aware of any for the first three. :) |
| 19:58 | <TabAtkins> | Hm, the bad commit just changes the timing of the section-link generator. |
| 20:02 | <Domenic> | probably changed where it happens in the pipeline in relation to encoding stuff? |
| 20:05 | <TabAtkins> | No, encoding stuff is already long-since done, or still a long ways away. |
| 20:08 | <TabAtkins> | Oh! It's triggering on the ToC. Interesting. |
| 20:08 | <TabAtkins> | That shouldn't be happening anyway, obvs. |
| 20:09 | <TabAtkins> | But now the error makes sense. |
| 20:10 | <Hixie> | annevk: you around? |
| 20:25 | <Hixie> | annevk: when i sent a URLUtils object's input to null, is it supposed to keep its old state? |
| 20:27 | <Hixie> | oh nm i misread the spec |
| 20:39 | <TabAtkins> | rubys: Fixed. The new position, though, reveals some markup errors - you've got two spots like `<a title=foo><a>foo</a>`. |
| 20:44 | <trevnorris> | is there a standardized test for something like https://url.spec.whatwg.org/#host-parsing ? |
| 20:45 | <Ms2ger> | github/w3c/web-platform-tests/ probably has some |
| 21:03 | <trevnorris> | Ms2ger: thanks |
| 21:05 | <Hixie> | ok dreamhost have informed me that they are updating ubuntu (and openssl) with an ETA of Q1 2015 |
| 21:28 | <rubys> | TabAtkins: thanks! |
| 21:33 | <Domenic> | I just sent a mail to public-script-coord and got back "******* About the W3C Mailing Lists *******" in response |
| 21:33 | <Domenic> | It's not in the archives |
| 21:33 | <Domenic> | ohhh |
| 21:33 | <Domenic> | i sent it to public-script-coord-request |
| 21:34 | <Domenic> | good times! |
| 21:41 | <TabAtkins> | Domenic: "spike"? |
| 21:41 | <Domenic> | TabAtkins: http://www.extremeprogramming.org/rules/spike.html |
| 21:42 | <TabAtkins> | ah, kk |
| 21:43 | <TabAtkins> | I understand the stream/pipe model better now. I assumed that piping was done directly by either the source or the target; instead, it's done by a transient (not fully reified) pipe function that handles both ends. |
| 21:43 | <Domenic> | yeah :) |
| 21:44 | <Domenic> | it lives on the source for convenience of syntax + the fact that we don't have anywhere (like, say, a module) to put it in |
| 21:48 | <TabAtkins> | And it does look like pipes are "push-like" - they aggressively pull data from the source and write it to the target, rather than waiting for someone to pull on the target first. |
| 21:55 | <Domenic> | well, you can't pull on generic targets, is the thing |
| 21:55 | <Domenic> | I guess the OS can |
| 21:57 | <TabAtkins> | Right. |
| 21:57 | <TabAtkins> | Purely-writable streams are just transform streams that only the OS has the readable side of. ^_^ |
| 21:58 | <TabAtkins> | All streams are transform streams in theory, since you can't produce information from nothing, and piping to /dev/null isn't very useful. |
| 22:02 | <Hixie> | foolip: ping |
| 22:04 | <Hixie> | TabAtkins: random question... what emotion is your ^_^ smiley supposed to convey? is it equivalent to :-), :-|, >:-(, -_- ? |
| 22:04 | <TabAtkins> | :) |
| 22:06 | <Hixie> | k, thanks |
| 22:09 | <Domenic> | TabAtkins: yeah, I spent a few days trying to make that explicitly the case in the API, but it ended up falling down in some pragmatic ways. i agree it is a nice conceptual model though. |
| 22:19 | <Domenic> | TabAtkins: if I make changes to my Bikeshed repo then type "bikeshed" somewhere else will it use my local copy? |
| 22:25 | <TabAtkins> | If you install it in the recommended way, yes. Otherwise, dunno |
| 22:28 | <Domenic> | we'll find out!! |
| 22:34 | <Hixie> | annevk: looks like your url spec allows scheme data to be encoded including "?", even though it defines "?" as the delimiter for the next component when parsing |
| 22:42 | <Hixie> | annevk: (or maybe the bug is allowing a query component on those urls? e.g. consider data:,? ) |
| 23:21 | <Mathuin> | Searching for cairocffi and IRC brought up a hit on this channel -- is this the right place to ask cairocffi-related questions? |
| 23:27 | <Ms2ger> | Not quite |
| 23:28 | <Mathuin> | can you recommend another one? |
| 23:43 | <MikeSmith> | SimonSapin: ↑ |
| 23:44 | <SimonSapin> | Hi Mathuin. Let’s try #cairo |
| 23:44 | <SimonSapin> | (What lead you to this channel?) |
| 23:46 | <Mathuin> | Actually something you mentioned in the channel about it got logged and searchable on google. :-) heading to #cairo now... :-) |
| 23:58 | <jamesr_> | Hixie: ^_^ ==> http://fc06.deviantart.net/fs45/i/2009/068/1/8/Emoticon_WP___04_by_elusive.jpg |
| 23:59 | <TabAtkins> | Its actual source is anime, where the "triangle eyes" indicate happiness. |
| 23:59 | <Hixie> | wouldn't that be ^u^ or something |
| 23:59 | <Hixie> | i thought it was supposed to be a cat :-P |
