| 06:34 | <annevk> | Ah, Domenic left a comment already |
| 09:04 | <Katy Romero> | hi |
| 09:12 | <annevk> | Noam Rosenthal: so for navigation timing, where do we actually set navigation start currently? In your presentation you mentioned it might be before unload runs, but in the specification it seems all it gets is timing information from fetch, which would be a later point in time |
| 09:18 | <Noam Rosenthal> | Noam Rosenthal: so for navigation timing, where do we actually set navigation start currently? In your presentation you mentioned it might be before unload runs, but in the specification it seems all it gets is timing information from fetch, which would be a later point in time It’s specified in the HRtime spec |
| 09:19 | <Noam Rosenthal> | The other parts are relative to that time origin It would be better to move it into HTML to make it explicit |
| 09:20 | <annevk> | Noam Rosenthal: yeah, I would have expected HTML to create a time entry, similar to what we do for the other things |
| 09:21 | <Noam Rosenthal> | Noam Rosenthal: yeah, I would have expected HTML to create a time entry, similar to what we do for the other things |
| 09:31 | <annevk> | Jake Archibald: hey, did you see the review request in https://github.com/whatwg/fetch/pull/1329? |
| 09:33 | <Jake Archibald> | Thanks for the ping, I'll do this today |
| 11:00 | <Jake Archibald> | Done! |
| 11:36 | <Noam Rosenthal> | Done! |
| 13:40 | <annevk> | Jake Archibald: https://github.com/whatwg/fetch/pull/1343 maybe too if you have another 10min or so? |
| 13:48 | <Jake Archibald> | Tomorrow hopefully. Picked up a regular non-exciting illness so I've bailed on work today |
| 13:49 | <annevk> | Jake Archibald: take care! |
| 13:57 | <Jake Archibald> | Ta! It's just a cold, but one of those ones that stops you sleeping so brain no workie |
| 14:30 | <annevk> | Oh wow, Chromium has multiple JSON parsers: https://bugs.chromium.org/p/chromium/issues/detail?id=1263590 |
| 14:33 | <annevk> | fetch("data:,1%0A//test").then(res => res.json()) does seem to reject thankfully |
| 14:47 | <annevk> | Domenic: did you think about moving the remainder of serialize/deserialize to Web IDL once JS adopts all the JS bits? |
| 15:27 | <Domenic> | annevk: I hadn't considered that; I don't know what would make Web IDL or HTML better in cases like this or the incumbent tracking stuff etc. |
| 15:28 | <annevk> | Domenic: I was mainly thinking IDL because it defines platform objects so it might as well define infra for them as well |
| 15:36 | <Domenic> | omg https://source.chromium.org/chromium/chromium/src/+/main:base/json/json_reader.h;l=27-36;drc=75cd3c0e2bc80ef1e93a51285cf7892c098e3a08 |
| 15:38 | <Domenic> | Thankfully that version (our third JSON parser) appears to be used only for browser stuff like settings files, Chrome sync stuff |
| 15:38 | <Sam Sneddon [:gsnedders]> | I still want an optional arg for JSON.parse that makes it parse JSON5 ☹️ |
| 15:39 | <Andreu Botella (he/they)> | Not extension manifests? I thought those allowed // escapes |
| 15:39 | <Domenic> | The second JSON parser (which has less-documented departures from the JSON spec, one of which is comments) is used for manifests |
| 15:39 | <Andreu Botella (he/they)> | I see |
| 15:39 | <Domenic> | Oh extension manifests |
| 15:39 | <Domenic> | Not sure which one for that |
| 15:39 | <Domenic> | Web app manifests was what I was talking about, which is getting fixed... https://chromium-review.googlesource.com/c/chromium/src/+/3247155 |
| 15:40 | <Domenic> | But we still gotta fix import maps, trust tokens, web bundles, speculation rules, ... |
| 15:41 | <annevk> | Ouch |