| 08:38 | <annevk> | JakeA: you around? What's the simplest way to create a ReadableStream that's read from and has released its lock? |
| 08:38 | <annevk> | JakeA: my main problem is that read() is forever pending |
| 08:39 | <annevk> | JakeA: do I basically have to pass something to the ReadableStream constructor? |
| 08:42 | <annevk> | https://irccloud.mozilla.com/pastebin/VQZupKkf/ |
| 08:42 | <annevk> | I guess I'll go with that for now, it seems to do the trick |
| 09:37 | <yoav> | annevk: Anything else blocking https://github.com/whatwg/fetch/pull/773? Do you want to related HTML changes to land at the same time? |
| 09:40 | <annevk> | yoav: I'd like to figure out what remains to be done |
| 09:47 | <annevk> | hsivonen: it's rather mindboggling how bad C screwed up with locale-dependent functions |
| 09:48 | <annevk> | hsivonen: I guess it says something about the programming community's diversity that this still isn't addressed in some way |
| 10:23 | <hsivonen> | annevk: it's addressed in two ways: ICU if you are processing natural language and rolling your own ASCII-only functions when dealing with protocol text. |
| 10:24 | <hsivonen> | annevk: also, it's sad when committees do things like *look* like i18n is addressed, but the solution is so unreliable that the facility didn't gain suitablity for i18n and also lost suitability for what it worked for before i18n was attempted |
| 10:25 | <hsivonen> | annevk: and they aren't just locale-dependent but an UB trap, too. fractally wrong. |
| 10:26 | <hsivonen> | annevk: I don't expect the committee to address it. See how passive-aggressive the _s functions in C11 are when they supposedly addressed security. |
| 10:31 | <hsivonen> | (instead of introducing new header files with the _s functions, they introduced a preprocessor dance to make the _s functions visible in the old headers) |
| 10:32 | <annevk> | UB = undefined behavior, I guess? |
| 10:33 | <hsivonen> | annevk: yes |
| 11:02 | <JakeA> | annevk: that looks right to me |
| 11:14 | <nox> | annevk: Re: Diversity: |
| 11:14 | <nox> | annevk: the RFC for non-ASCII identifiers in Rust currently has 593 comments. |
| 11:15 | <zcorpan> | TabAtkins: I tried installing bikeshed (macOS). I get "TypeError: unsupported operand type(s) for +: 'dict_values' and 'list'" |
| 11:17 | <zcorpan> | hmm, seems like i needed to use pip2 |
| 11:17 | <zcorpan> | the python situation is terrible |
| 11:18 | <nox> | https://twitter.com/nokusu/status/1031252970342674440 |
| 11:28 | <annevk> | nox: I guess I don't really know the tradeoffs involved for low-level languages, but sounds bad |
| 11:28 | <nox> | annevk: Most of them is BUT I DON'T UNDERSTAND CHINESE |
| 11:32 | <annevk> | o_O |
| 11:37 | <hsivonen> | part of it is "I know Unicode edge cases and want my edge cases to be banned (as opposed to just letting people not do things that don't make sense)" |
| 11:37 | <hsivonen> | (contrast with the CSS definition of identifiers) |
| 11:39 | <nox> | hsivonen: Oh god CSS identifiers. :D |
| 11:40 | <annevk> | I guess you could do what JavaScript does, but seems nicer for parsers to not have to validate too much |
| 11:40 | <annevk> | Anyway, with close to 600 comments that thread doesn't need more opinions |
| 12:11 | <annevk> | yoav: so I don't know what happened with Client Hints since the last time someone would fix the situation in Fetch was in March |
| 12:11 | <annevk> | yoav: it's nearly half a year later now |
| 12:12 | <annevk> | yoav: given the state of things I'd rather just add some issue markers and wait for someone to fix it up completely |
| 12:12 | <annevk> | yoav: since the updated spec is still not something we could recommend non-Chromium to implement |
| 12:12 | <annevk> | yoav: "updated spec" meaning your PR being applied |
| 12:18 | <yoav> | annevk: Regarding timelines, https://github.com/whatwg/fetch/issues/726 was filed mid-May with a plan for a fix. https://github.com/whatwg/fetch/pull/773 started at June. |
| 12:20 | <annevk> | yoav: my bad |
| 14:52 | <nnyby> | Hello, I have a question about <input type=range>. Can I make a range input's value step up and down on click? I couldn't find where the spec defines this before. More details for my question are here on stackoverflow: https://stackoverflow.com/q/51988195/173630 |
| 15:01 | <zcorpan> | nnyby: added an answer |
| 15:02 | <nnyby> | zcorpan: sweet! looks promising.. i'll give it a try. |
| 15:41 | <TimothyGu> | annevk: Would you wanna take a look at https://github.com/whatwg/html/pull/3946? If not Domenic and I are planning on merging it later today |
| 15:55 | <annevk> | TimothyGu: ah sorry, I have already and sg |
| 16:11 | <TimothyGu> | annevk: great, thanks! |
| 22:23 | <domfarolino> | Is it me or are most of the links in https://github.com/whatwg/meta/blob/master/LABELS.md mostly broken due to missing end quotes? |
| 22:29 | <TabAtkins> | domfarolino: Nope, not just you, that's definitely the problem. |
| 22:30 | <domfarolino> | Ah. Will fix |
| 23:19 | <TabAtkins> | Domenic: Geez, you're fast. ^_^ |