| 00:03 | <Domenic_> | annevk: the way I see it you want raw deflate algorithm operating on ArrayBuffers, and then a nice convenience zip API on top of it that can make assumptions about formats and modes. |
| 00:03 | <Domenic_> | annevk: the convenience API should be able to open common renamed-zip formats like .docx; not sure what zip variant they use, but I imagine it provides a good outer limit on reasonable-things-to-support. |
| 00:04 | <annevk> | Yeah, should probably look into that... |
| 00:09 | <Domenic_> | getting deflate out the door fast though would bring many smiles to the world |
| 00:36 | <Domenic_> | annevk: why does https://github.com/annevk/url have issues turned off? |
| 00:37 | <annevk> | not sure |
| 00:38 | <annevk> | will have a look tomorrow, way tired |
| 00:38 | <Domenic_> | annevk: ok, will email you failing test case then? |
| 00:38 | <annevk> | sure |
| 03:06 | <heycam> | TabAtkins, did you switch Variables to your new spec generator? the "«token»"s show now as "<<token>>". |
| 09:52 | <Ms2ger> | krijn, logs seem to be stuck? |
| 10:28 | <krijn> | Ms2ger: on it! |
| 10:28 | <Ms2ger> | Thanks |
| 10:31 | <krijn> | Thanks for noticing |
| 10:33 | <Ms2ger> | Np |
| 12:46 | <annevk> | matijs: I recommend finding the answers to questions in specifications, not mailing lists ;) |
| 12:46 | <annevk> | matjas: ^^ |
| 12:46 | <annevk> | your nicks are confusingly similar |
| 12:47 | <matjas> | annevk: what mail in particularare you referring to? |
| 12:47 | <annevk> | matjas: latest tweet |
| 12:48 | <matjas> | annevk: ah, good point, as always :) |
| 12:49 | <annevk> | matjas: if you use a BOM, which is kinda icky, your JS would be fully portable. If you use utf-8 all the time (including in your HTML), it's portable within your set of sites |
| 12:50 | <annevk> | matjas: otherwise you need to involve either <script charset> (eww) or HTTP (annoying) |
| 12:50 | <matjas> | i know |
| 12:50 | <matjas> | my point was that JS library authors have no way of knowing in which environments their code will end up |
| 12:51 | <matjas> | so going ASCII-only is probably the safest bet |
| 12:52 | <annevk> | It seems JavaScript could introduce a "utf-8" directive as the first thing in the file |
| 12:52 | <matjas> | like, d3.js (a popular dataviz library) has `var π = Math.PI` in its source, and every few weeks someone complains about “encoding issues” in their issue tracker |
| 12:52 | <annevk> | for easier editing than the utf-8 BOM |
| 12:52 | <annevk> | it's somewhat backwards compatible in that it doesn't throw and would definitely improve matters going forward |
| 12:53 | <matjas> | directives suck, though |
| 12:53 | <annevk> | what's the problem? |
| 12:53 | <matjas> | 2008: // nothing |
| 12:54 | <matjas> | 2009 (ES5): 'use strict' |
| 12:54 | <matjas> | 2014: 'use strict'; 'use utf-8'; |
| 12:54 | <annevk> | I think "use strict" is regarded as a mistake |
| 12:54 | <annevk> | and "utf-8" is not really like "use strict", it's more how to decode the source, like @charset |
| 12:55 | <matjas> | would’ve been cool if 'use strict' implied utf-8, though |
| 12:55 | <annevk> | anyway, I guess it hasn't really been needed thus far, might not be worth it |
| 12:55 | <matjas> | but too late for that i guess |
| 12:55 | <annevk> | yeah, workers imply utf-8 |
| 12:55 | <annevk> | but not use strict |
| 12:56 | <matjas> | xhr.responseType='json' too |
| 12:56 | <annevk> | we should make modules imply utf-8 |
| 12:56 | <matjas> | which is pretty awesome |
| 12:56 | <annevk> | yes |
| 12:56 | <matjas> | yeah! |
| 12:56 | <annevk> | I force utf-8 everywhere |
| 12:56 | <annevk> | it's like a side mission in Zelda |
| 12:57 | <matjas> | out of curiosity, has no one complained about not supporting other encodings for `responseType='json'` since you specced it in 2011? |
| 12:57 | <matjas> | would be surprising, as there is no reason at all to not use UTF-8 |
| 12:58 | <annevk> | haven't had complaints yet |
| 12:58 | <annevk> | note that XHR send() also forces utf-8 |
| 12:59 | <annevk> | bbl |
| 18:09 | <TabAtkins> | heycam: Yes, I did, but I last bikeshedded it before I had syntax support for token references. |
| 18:35 | <TabAtkins> | heycam|away: All fixed up now, and your bugs in the list fixed as well. |
| 22:52 | <heycam> | TabAtkins, cool |
| 22:53 | <heycam> | (I got my quote style wrong yesterday in irc) |