| 00:03 | <sirisian> | Jack Works: Is this in new APIs/proposals already? |
| 00:44 | <Jack Works> | Jack Works: Is this in new APIs/proposals already? |
| 19:57 | <guybedford> | Our usual call tomorrow overlaps with the Involvement in Standards discussion at the Node.js Collab Summit... I have a preference towards keeping our meeting especially with the upcoming progress in the Wasm CG and TC39 but wanted to see if many others here plan to go to the collab summit? |
| 20:27 | <nicolo-ribaudo> | I plan to prioritise our modules meeting |
| 21:09 | <guybedford> | ok, I will be attending the modules session just before but skipping the standards discussion, so will see you in the modules meeting, possibly a couple of minutes late if things run over a bit |
| 21:45 | <littledan> | I'd really love it if you all joined the Node Collab Summit session about modules tomorrow |
| 21:45 | <littledan> | The goal is to have more of a cross-over here |
| 21:46 | <littledan> | You can register for remote participation with this link: https://zoom.us/webinar/register/WN_3FJCCjgiRseMunHFhb8NoA#/registration |
| 21:48 | <guybedford> | littledan: the timing is just a bit unfortunate because of TC39 next week for Deferred evaluation and the ESM Integration going to the Wasm CG next week we do have a lot to discuss |
| 21:49 | <littledan> | Sorry I'm fine with skipping the general standards part, I'm talking more about the modules session |
| 21:49 | <littledan> | should we consider moving the modules session to enable participation? |
| 21:49 | <guybedford> | ohhh yes, I plan to be there for the modules session! |
| 21:49 | <guybedford> | it is right before our usual modules meeting |
| 21:50 | <littledan> | Good, so that time works out for you? |
| 21:50 | <guybedford> | yes, sorry I missed you were referring to the modules session, I will be there and I hope others will too! |
| 21:51 | <littledan> | also, if you have thoughts on what I should include in that session, it'd be great... honestly you'd be better to host it than me, guybedford |
| 21:53 | <guybedford> | I always struggle to know how much to share longer term modules directions within the Node.js project, when there are very real immediate standards questions that are more pressing. I did previously present our modules work at last years collab summit, but it wasn't very well attended. |
| 21:53 | <littledan> | well, now people won't have a choice! it's single-track |
| 21:53 | <littledan> | I think many people were inspired by Joyee's recent work, and we can build off of that momentum |
| 21:54 | <littledan> | I was thinking of focusing on reviewing what we have, and what are the immediate future plans in both Node and TC39 (Stage 2+ TC39 proposals maybe), and then gather input as to what people want to happen in each place |
| 21:58 | <guybedford> | I think it sounds like you have a clear way to present it, I'm happy to share some thoughts on the current harmony efforts as I did previously. Also very interested to hear what people are interested in discussing. |
| 22:01 | <kriskowal> | The recent CJS imports ESM work has interesting parallels with deferred export. They both apply only to synchronous subgraphs, and imply that not-yet-evaluated transitive dependencies will execute in the same microtask. They both run ESM on the caller’s stack. They’re both limited to exposing the namespace object and not destructured imports. |
| 22:09 | <littledan> | The recent CJS imports ESM work has interesting parallels with deferred export. They both apply only to synchronous subgraphs, and imply that not-yet-evaluated transitive dependencies will execute in the same microtask. They both run ESM on the caller’s stack. They’re both limited to exposing the namespace object and not destructured imports. |
| 22:09 | <littledan> | they both relate to the theme "actually ESM has never been all that async, so we can do this important stuff" |
| 22:12 | <guybedford> | There's an interesting edge case with cycles - the ESM Evaluate function was actually never designed to be reentrant. With CJS importing ESM that effectively introduces reentrancy. And import defer is actually now fixing that reentrancy invariant to more clearly define what reentrancy is allowed (non cyclical). |
| 22:13 | <guybedford> | But it does unfortunately mean telling Joyee she can't have ESM - CJS cycles at all :P |