| 03:05 | <Domenic> | Bikeshedding help requested for an error type that includes properties with how much X is available, versus how much X you requested: https://github.com/whatwg/webidl/issues/1463 |
| 03:18 | <sideshowbarker> | I don’t think that to developers using “Quota” will seem to storage-specific. I think developers are familiar with the context of “quota” being used in the web platform to mean a browser/UA-imposed maximum. I don’t think usage of “Limit” or other imaginable alternatives would have the same familiarity to developers. |
| 03:21 | <Domenic> | That is helpful, thank you! |
| 03:23 | <sideshowbarker> | New/separate unrelated question for the room: At https://w3c.github.io/uievents/#event-type-keypress I find this requirement for the
In other words, per-spec: Browsers must not fire the |
| 03:24 | <sideshowbarker> | But… that’s not what existing engines do. Instead, they do fire keypress for the Enter key, and for modifier keys, etc. — everything |
| 03:25 | <sideshowbarker> | Well, all engines except Ladybird (which implements that requirement per-spec) |
| 03:27 | <sideshowbarker> | So… that requirement doesn’t seem to be web-compatible. And that makes me wonder why it’s in the spec, and how it made its way into that spec to begin with — I mean, if implementors don’t support it. And given that if everybody did implement it per-spec, it seems like it would break all kind of existing sites. |
| 03:28 | <sideshowbarker> | I’ll file an issue for it, but in the meantime I’m wondering if anybody here might know that background on where it came from |
| 04:15 | <Domenic> | I vaguely remember there being stuff around keyCode vs. other properties, and only one of them makes sense as producing a character value. Maybe the intent was to have an event that only fires when that character value property is set. (But nobody implemented it that way.) |
| 04:19 | <sideshowbarker> | I misspoke a bit, actually: UAs do all seem to conform to that requirement for all non-character keys — except for the Enter key |
| 04:20 | <sideshowbarker> | That is, all UAs do still fire keypress for the Enter key |
| 04:20 | <sideshowbarker> | So I guess I’ll open a spec PR with a patch which states that exception |
| 04:50 | <sideshowbarker> | Well, it seems there are existing already-reported issues: |
| 04:55 | <sideshowbarker> | Anyway, I really just want to know what I should implement in Ladybird in order to be interoperable with existing behavior in other engines |
| 04:56 | <sideshowbarker> | And per Nakano-san’s analysis in https://github.com/w3c/uievents/issues/183#issuecomment-448091687, that seems to maybe come down to being Enter, Shift + Enter and Ctrl + Enter as the exceptions |
| 10:20 | <annevk> | Hype: https://bsky.app/profile/tabatkins.com/post/3lggxdwp4h22w |
| 10:26 | <sideshowbarker> | oh wow, finally being able to get the HTML spec build ported over to Bikeshed would be a big win |
| 10:27 | <sideshowbarker> | hope Tab gets the perf stuff worked out |
| 10:27 | <sideshowbarker> | zcorpan: FYI https://github.com/w3c/uievents/pull/392 |
| 13:47 | <Colin Alworth> | sideshowbarker: I usually just lurk here, but doesn’t enter emit a newline “character value”? |
| 14:48 | <sideshowbarker> | Colin Alworth: per-spec at https://w3c.github.io/uievents/#character-value, it seems to mean the same thing as https://w3c.github.io/uievents/#unicode-character-categories, which is:
|
| 14:50 | <Colin Alworth> | Hmm, surely space and tab too? Or are you saying those should be excluded as well? |
| 14:50 | <Colin Alworth> | Newline is certainly a Unicode character |
| 14:51 | <sideshowbarker> | I’m saying what the spec says it means is “all the Letter (Ll, Lm, Lo, Lt, Lu), Number (Nd, Nl, No), Punctuation (Pc, Pd, Pe, Pf, Pi, Po, Ps) and Symbol (Sc, Sk, Sm, So) category values” |
| 20:02 | <nsITobin> | I think it is time to bring an XML web back. |
| 20:02 | <nsITobin> | along the 1.0 lines not the modular nonsense. |
| 20:10 | <nsITobin> | Here is my vague semblance of a plan.. HTML5 living standard does as the living standard does.. But codify versioned strict predictable serializations as XHTML 5.x or Year or whatever. Anyone wanting that is more likely to accept or prefer the strict markup requirements. "Unversioned" XHTML 5 == current standard. Anyone think there is anything remotely good about my vague list of desires presented as a plan? |
| 20:18 | <TabAtkins> | lol i hadn't asked y'all about it yet, just been doing some exploration first to make sure it is indeed possible. i'll poke y'all in q2 or q3; gotta get multi-page output working first |
| 20:26 | <TabAtkins> | But yes, getting rid of the accidental quadratics was easy, and I'm continuing to poke around to see what else I can shave off now. Currently the "15 copies of DOM" spec is 1m50s on my laptop, which indeed right about 15x the time of DOM itself, meaning I'm just tightening up some surprisingly linear bits now. |