2024-04-11 [12:07:12.0894] Which room is the plenary convo? [12:07:24.0790] Yehuda Katz #tc39-delegates:matrix.org [12:08:00.0455] I'm a delegate. How can I get permissions? [12:10:30.0217] > <@wycats:matrix.org> I'm a delegate. How can I get permissions? have caridy file a new delegate issue on the admin and business repo, and i'll get you set back up [12:17:50.0437] ljharb: he was already moved [12:18:03.0609] oh ok then np [12:18:10.0108] I op'ed on delegates already 2024-04-15 [22:28:39.0311] The more I work on my Parser Augmentation proposal the more it seems like a _really_ good idea. It doesn't require browser buyin, and it has a really slow and gentle adoption strategy, which just feels nice 🙂 https://es.discourse.group/t/proposal-parser-augmentation-mechanism/2008 [22:32:14.0139] but the best part of all is, it's just describing stuff people are already doing [22:42:08.0630] this is basically just a spec of "this is what it means to be a transpiler", but it has legs 2024-04-17 [17:37:28.0531] > <@dmchurch:matrix.org> The more I work on my Parser Augmentation proposal the more it seems like a _really_ good idea. It doesn't require browser buyin, and it has a really slow and gentle adoption strategy, which just feels nice 🙂 https://es.discourse.group/t/proposal-parser-augmentation-mechanism/2008 First draft of the proposal is up! 😄 I'll appreciate any feedback of course, but don't feel obligated, it's kind of a dense read! https://github.com/dmchurch/babel-rfcs/blob/parser-augmentation/rfcs/0000-parser-augmentation-rfp.md 2024-04-21 [15:45:11.0357] I wonder if Parser Augmentation would make more sense to people if I described it more as "asm.js for the parser" than as "transpilers for the runtime" 🤔 [15:46:30.0539] because it's the same kind of concept - ES code that _can_ be executed at runtime without penalty, but is _intended_ to be statically internalized [15:47:58.0120] * because it's the same kind of concept - ES code that _can_ be executed at runtime without (significant) penalty, but is _intended_ to be statically internalized [16:02:31.0019] and it doesn't need TC39 buy-in any more than asm.js did, but the process of writing my own prototype implementation into V8 is reconfirming to me that having expert input on the design from the start will make implementation and integration much easier if and when it _does_ get to runtime engines 2024-04-22 [09:07:38.0945] if it doesn' [09:07:57.0580] * – [12:01:11.0851] I forget - if I want to make a new proposal, should I start it in my own repo (and then transfer it to `tc39` at stage 1), or can I go straight to opening a `tc39/proposal-*` repo? [14:01:37.0489] TabAtkins: ideally you'd start from the template - most people don't have permissions to make repos in tc39, but you could start it right in tc39-transfer if you wantd [14:01:38.0454] * TabAtkins: ideally you'd start from the template - most people don't have permissions to make repos in tc39, but you could start it right in tc39-transfer if you wanted [14:02:13.0178] okay, cool. 2024-04-23 [17:19:23.0542] But yeah you make a personal repo [17:20:08.0743] TC39 org kicks in at stage 1 [17:57:03.0980] it's not that a proposal repo is disallowed in the org pre-stage-1, just that it's not a requirement until then, and if the proposal isn't going to go anywhere, it may not want to live there [07:12:19.0619] Do we have any pre-stage-1 proposals in the TC39 org? How do we determine which ones can enter? [08:43:01.0789] continuing discussion on this in #tc39-delegates:matrix.org 2024-04-24 [20:36:46.0551] Ooh, I think I've just come up with a better way to describe PA! Parser Augmentation is two things: 1. an addition to the ECMAScript syntactic grammar for making assertions about the source text that follows, such as whether it's a script or a module, strict or not, whether there are any `var` declarations after first point of use, whether it has type assertions and what kind, what experimental operators it uses, etc. 2. an optional runtime API that allows ECMAScript code to instruct the parser on how to polyfill any unknown assertions (or which ones can be safely ignored), by presenting an interface designed to allow engines to optimize their parser performance even with unexpected or unusual syntaxes.