| 06:45 | <hsivonen> | annevk: I've tried to look away from the HTML5 article on Wikipedia to avoid conflict-of-interest editing. Back in 2007, when I tried to do edits in this space, I got a note on my user talk page: https://en.wikipedia.org/wiki/User_talk:Hsivonen (It says, among other things: "Initially I and others felt WHATWG wasn't particularly notable, so we put the smack down on excessive mention of it in the |
| 06:45 | <hsivonen> | article.") |
| 06:50 | <hsivonen> | annevk: trying to follow the protocol for COI editing avoidance and pointing out issues on the talk pages in the hope that an editor without COI comes in, sees the correctness of the suggestions, and edits them in is futile. I've tried it with the EME, Big5, and Firefox articles and failed. |
| 06:52 | <hsivonen> | (Entirely unsurprisingly, the only other note on my user talk page mentions notability.) |
| 06:56 | <hsivonen> | (I wanted to write an article about ripgrep, as it's pretty ridiculous that it doesn't have one, but I figured that since I'm the author of one of ripgrep's direct dependencies, a broad interpretation of COI could count is as COI editing, so I instead wrote a trial-balloon about exa, whose notability is supported by the same sources as ripgrep's. I'm surprised in a positive way that the deletionists |
| 06:56 | <hsivonen> | haven't deleted the exa article yet.) |
| 07:49 | <annevk> | Yeah, those policies make contributing back to Wikipedia rather hard |
| 09:41 | <yhirano> | annevk: are you going to add coep tests for service workers soon? If not I'm willing to add some. |
| 09:42 | <annevk> | yhirano: please do |
| 09:42 | <annevk> | yhirano: I'm looking at ImageBitmap atm |
| 09:43 | <annevk> | yhirano: seems that Fx might be lacking there somewhat, e.g., no support with BroadcastChannel |
| 09:43 | <yhirano> | annevk: thanks, will do |
| 11:53 | <hsivonen> | Do we have WPTs that check that a reused TextDecoder retains its BOM handling mode across reuses? |
| 11:57 | <hsivonen> | annevk: Also, we're already late: https://github.com/rustwasm/wasm-bindgen/blob/487289cf9b88913124e8aa22f4f21aa145c47733/crates/cli-support/src/js/mod.rs#L1103 |
| 12:02 | <annevk> | hsivonen: I’m not sure if we have for all possible reuses |
| 12:03 | <annevk> | hsivonen: streaming a BOM is covered I think |
| 12:03 | <hsivonen> | annevk: OK. Firefox BOM ignoring code looks bad for reuse purposes. |
| 12:04 | <hsivonen> | annevk: my hypothesis from code inspection is that in Firefox, a reused TextDecoder always removes the BOM |
| 12:06 | <hsivonen> | the spec is a bit hard to read to figure out what the expected survival of these flags is |
| 12:09 | <hsivonen> | annevk: indeed, a reused TextDecoder removes the BOM in Firefox |
| 12:11 | hsivonen | goes file a bug |
| 12:14 | <hsivonen> | filed as https://bugzilla.mozilla.org/show_bug.cgi?id=1574892 ; I'll add WPT assertions while at it |
| 12:33 | <hsivonen> | too late now, but the terminology "ignore" the BOM is super-confusing in this case |
| 12:34 | <hsivonen> | when ignore means don't remove |
| 12:34 | <annevk> | Thanks hsivonen |
| 12:35 | <annevk> | hsivonen: I guess in a way that is confusing, but you could also see it as a way to tell the decoder to ignore (in terms of looking at) the BOM |
| 12:35 | <hsivonen> | annevk: yes, I know that's what it means, but I still find it confusing |
| 12:42 | <annevk> | hsivonen: yeah, we could add more prose to the specification, but not sure if that helps |
| 16:40 | <Mattumd> | hi |
| 16:42 | <Mattumd> | I write a tutorial about HTML5 custom data and JS API. How can I submit it? |
| 17:24 | <annevk> | Mattumd: submit it where? Please give some context |
| 17:25 | <Mattumd> | https://blog.whatwg.org/submit-article |
| 17:27 | <annevk> | Mattumd: ah, I’ll DM |