| 00:03 | <croraf> | astearns, aha, like "thanks/respect to" |
| 01:55 | <Krinkle> | croraf: thx, nice little historical quirk |
| 01:56 | <Krinkle> | (Hat tip indeed) |
| 11:53 | <andreubotella> | the navigate algorithm doesn't require browsers to treat a navigation as a download if the resource has "binary" bytes, regardless of the mime type, right? |
| 11:55 | <andreubotella> | all browsers seem to do just that, and it's biting me in the ass when I'm trying to test control characters on multipart/form-data |
| 11:56 | <annevk> | andreubotella: the interaction of navigation and MIME sniffing is not that great I think, so it might well be wrong |
| 11:56 | <annevk> | andreubotella: you wouldn't navigate to a multipart/form-data response though, right? |
| 11:57 | <andreubotella> | the test infrastructure from https://github.com/web-platform-tests/wpt/pull/8618 has echo-response.py as the form's action, with an iframe as the target, and then reads from the iframe |
| 11:59 | <andreubotella> | s/echo-response/echo-content/ |
| 12:00 | <annevk> | andreubotella: I guess I don't really understand how that helps, but multipart/form-data alone should be enough for a download |
| 12:01 | <andreubotella> | /fetch/api/resources/echo-content.py sets Content-Type: text/plain |
| 12:01 | <andreubotella> | but anyway, I can find some other way to test those cases |
| 12:02 | <annevk> | andreubotella: oh I see, it tries to serialize what it sees on the server |
| 12:02 | <annevk> | andreubotella: perhaps the server should just do \x notation for control bytes and bytes >0x7F |
| 12:04 | <andreubotella> | annevk: yeah, I think I'll do that |
| 12:05 | <andreubotella> | btw, any clue why wpt.fyi has safari crashing on some tests that pass on a webkit nightly I built on my (linux) machine? |
| 12:05 | <andreubotella> | https://wpt.fyi/results/FileAPI/file/send-file-form-iso-2022-jp.tentative.html?label=experimental&label=master&aligned |
| 12:08 | <annevk> | andreubotella: no, but they have touched their encoding impl recently so I'll alert someone |
| 12:16 | <annevk> | andreubotella: via gsnedders: https://bugs.webkit.org/show_bug.cgi?id=218112 |
| 12:20 | <gsnedders> | andreubotella: specifically, it's a macOS only crash, related to CFNetwork, and doesn't occur in WebKitTestRunner but only in Safari |
| 12:21 | <andreubotella> | I see |
| 19:44 | <Domenic> | annevk: it seems folks ignored my preference to use a non-versioned reference for WCAG in https://github.com/whatwg/html/issues/5761#issuecomment-664499216 ... not sure if we just want to merge https://github.com/whatwg/html/pull/6149 or suggest the contributor change. |
| 20:09 | <annevk> | Domenic: oh that’s my bad I think |
| 20:09 | <annevk> | Domenic: I would ask |
| 20:09 | <annevk> | Happy to do so myself tomorrow |
| 20:32 | <Domenic> | Sounds good, will let you do it :) |