| 08:38 | <ryzokuken> | robpalme: 👋 can I get a tech check please? |
| 08:38 | <robpalme> | of course - logging in now |
| 08:39 | <ryzokuken> | perfect, thanks. |
| 08:42 | <robpalme> | looks like we need to kickstart team - it's in lobby mode |
| 08:43 | <ryzokuken> | okay |
| 08:43 | <ryzokuken> | let me rejoin |
| 08:47 | <ryzokuken> | does it work yet? |
| 08:48 | <robpalme> | i have issues my end |
| 08:48 | <ryzokuken> | ah okay |
| 08:48 | <robpalme> | is anyone else in? |
| 08:48 | <Bakkot> | I'll join in a minute as soon as my tea's up |
| 08:48 | <michaelficarra> | I'm in |
| 08:49 | <robpalme> | i am in! it's working |
| 08:55 | <michaelficarra> | are we gonna delete all these notes from before the meeting starts? |
| 08:57 | <Bakkot> | yes |
| 09:00 | <robpalme> | we're almost ready to begin! |
| 09:00 | <robpalme> | Teams is go! TCQ (including all new temperature checking) is go! Automatic dictation is go! |
| 09:01 | <ryzokuken> | futuristic TC39 meeting |
| 09:03 | <shu> | i see aki talking but hear nothing |
| 09:03 | <michaelficarra> | bakkot: you can turn it off during like voting topics and such, yeah? |
| 09:04 | <shu> | can other people hear or is it just me? |
| 09:04 | <Bakkot> | by killing and restarting the process on my other machine, yes |
| 09:04 | <michaelficarra> | I can't hear Aki either |
| 09:04 | <rbuckton> | We have automatic dictation? |
| 09:04 | <michaelficarra> | rbuckton: yes! |
| 09:05 | <rbuckton> | I was about to ask if any of the chairs or the host had considered using Teams' built-in transcription feature as a way to help with note taking. |
| 09:06 | <ljharb> | i believe the transcription mechanism we're using will work with any VC app, so we're not coupled to any of them |
| 09:10 | <brad4d> | congrats to whoever fixed the tcq bug that used to require me to manually enter the meeting code even when using the link including the code. Thanks! |
| 09:10 | <drousso> | oh wow yeah |
| 09:14 | <rbuckton> | whoever is typing needs to mute? |
| 09:14 | <michaelficarra> | brad42: please mute when typing |
| 09:14 | <michaelficarra> | *brad4d |
| 09:15 | <brad4d> | sorry, i thought I was muted |
| 09:18 | <michaelficarra> | this is an unfortunately low turnout :-( |
| 09:21 | <ystartsev> | 32 people, yeah its low.. |
| 09:22 | <shu> | how ominous |
| 09:22 | <shu> | "chairs are always listening" |
| 09:23 | <ryzokuken> | plot twist: big brother is actually TC39 chair |
| 09:23 | <ljharb> | robpalme: will the table on the 2020/11 agenda be updated with that info? |
| 09:24 | <robpalme> | doing it now! |
| 09:24 | <ljharb> | ty |
| 09:27 | <robpalme> | 35 participants in Teams now |
| 09:28 | <ljharb> | would be interesting to graph teams count over the course of each day :-p |
| 09:29 | <akirose> | does anyone have a 60-minute they would be able to move to today? I'm thinking shu & rbuckton are maybe the only option? |
| 09:30 | <shu> | akirose: when today? still working on slides |
| 09:30 | <michaelficarra> | we're not gonna have the transcription do the slides presentations? would like to have those notes when possible |
| 09:30 | <michaelficarra> | *I would |
| 09:30 | <akirose> | shu: 14:00 CET, last item of the day |
| 09:30 | <Bakkot> | [I'll try to kill the note-taker during slides, as a rule] |
| 09:31 | <littledan> | I would prefer to have notes during slides, if possible |
| 09:31 | <michaelficarra> | Bakkot: why? notes during slides are useful |
| 09:31 | <littledan> | there's often a lot of interesting stuff that people say during presentations |
| 09:31 | <Bakkot> | hm, ok |
| 09:31 | <littledan> | I felt like we killed the notes before mostly to give humans a break |
| 09:31 | <ljharb> | hm, we don't usually take notes during slides |
| 09:31 | <shu> | akirose: i should be ready by then if you need to move me |
| 09:31 | <akirose> | HERO |
| 09:31 | <michaelficarra> | littledan: same |
| 09:31 | <Bakkot> | humans still need a break, is the main thing |
| 09:31 | <ljharb> | seems fine to do so tho, ofc |
| 09:31 | <littledan> | but, you can update it afterwards |
| 09:32 | <michaelficarra> | Bakkot: then just take a break, we stil have the transcription |
| 09:32 | <Bakkot> | transcription is... not great without live correction |
| 09:32 | <michaelficarra> | seems like it depends on the accent and speaking style quiet a bit |
| 09:32 | <Bakkot> | (though it varies by speaker; it does worse with non-American/British accents, unfortunately) |
| 09:32 | <marja_> | +1 to michaelficarra, would totally like to have notes from slides too. |
| 09:33 | <michaelficarra> | it seems like speaking slower is better, too |
| 09:33 | <marja_> | often the slides don't have speaker's notes and it's not easy to decipher what was said after the preso |
| 09:33 | <michaelficarra> | plus, you'll probably avoid umms and ahhs that way |
| 09:33 | <littledan> | shu, akirose: ryzokuken and I have already written our Temporal slides and could go, if that's easier for you |
| 09:33 | <michaelficarra> | marja_: also, some people put very little detail on the slides themselves |
| 09:33 | <littledan> | if it's in the afternoon |
| 09:34 | <wsdferdksl> | The transcription is comically incoherent |
| 09:34 | <marja_> | yeah, sometimes a slide is just a code snippet, and it's practically impossible to know what was said about that code snippet |
| 09:34 | <Bakkot> | wsdferdksl for this particular speaker or previous ones? |
| 09:34 | <akirose> | littledan: I'm all settled but it's v helpful to know y'all's slides are ready to go |
| 09:34 | <wsdferdksl> | It depends on the speaker. About half of them… |
| 09:34 | <Bakkot> | (this one, yes, it's very poor; I'm not fixing it live and it's also doing worse than with previous speakers) |
| 09:35 | <michaelficarra> | to be fair, I have a hard enough time understanding Istvan myself |
| 09:35 | <gibson042> | robpalme: what is the correct URL for 2021 schedule/logistics slides? |
| 09:35 | <littledan> | I think if we get a block of incoherent notes and the speaker isn't available to correct it, it's fine to delete it |
| 09:36 | <robpalme> | https://docs.google.com/presentation/d/1vsOSn9alFLnmtYStNt71_KJU6uDW1nm-GsqRRIfajTw/edit |
| 09:36 | <gibson042> | thanks |
| 09:36 | <wsdferdksl> | Istvan seems clear to me. I have a lot less trouble understanding his speech than some other speakers here. But the transcription doesn't like him. |
| 09:36 | <robpalme> | I have PR'd the dates to the agenda - they are already in the calendar |
| 09:41 | <michaelficarra> | Aki: you are echoing? |
| 09:41 | <michaelficarra> | maybe it's coming from someone else though, sorry |
| 09:45 | <akirose> | lmk if you think it's me next time i talk |
| 09:59 | <haxjs> | It seems MM mention a issue of speccing iterator/generator ? Where is the issue we can follow ? |
| 10:00 | <ljharb> | haxjs: not an issue, just a question, which was answered |
| 10:01 | <ljharb> | haxjs: whether the Yield macro could also send/receive a value - which it can |
| 10:05 | <wsdferdksl> | Aki responded to a question by asking for questions to be put on the queue and then ignored the queue. |
| 10:05 | <haxjs> | ljharb Thank u, hope it can also work with `function.sent` proposal (receive the first value) which may be affected by the changes. |
| 10:06 | <shu> | wsdferdksl: what's the lookahead bug? could you please file an issue if there isn't one already? |
| 10:06 | <ljharb> | haxjs: i would expect it either has no effect, or makes it slightly easier to specify |
| 10:07 | <akirose> | wsdferdksl: I asked for questions to be directed to the Issues queue |
| 10:08 | <mathiasbynens> | (updated "meeting info" link in topic) |
| 10:08 | <wsdferdksl> | So we're not using tcq any more? |
| 10:08 | <akirose> | of course we are |
| 10:09 | <akirose> | those were status updates |
| 10:09 | <akirose> | not discussion topics |
| 10:09 | <wsdferdksl> | It's still not cool to just ignore questions. |
| 10:10 | <wsdferdksl> | I don't want chairs silently doing that. |
| 10:11 | <akirose> | i didn't ignore. I'm sorry you we seemed to miss each other there. |
| 10:12 | <wsdferdksl> | Aki, you did. You moved on to the next item without draining the tcq. |
| 10:14 | <ljharb> | akirose: i can talk about gus's item if there's time after the other normatives are discussed |
| 10:15 | <akirose> | ooh neat thanks |
| 10:16 | <leobalter> | Bakkot: funny enough, there is a big spec editorial inconsistency for using else, otherwise and using ; or . before inline else/otherwise. |
| 10:18 | <michaelficarra> | leobalter: still today? |
| 10:18 | <michaelficarra> | I thought we cleared that up recently |
| 10:21 | <michaelficarra> | brad4d and others: please make sure you've added yourself to the attendees list in the notes! |
| 10:21 | <robpalme> | @wsdferdksl I think Aki directed your item on lookaheads to the reflector? the status updates section of the meeting is not really meant for discussion - though we do not intentionally ignore tcq |
| 10:21 | <michaelficarra> | it helps note takers quickly grab initialisms |
| 10:22 | <michaelficarra> | (and I'm sure is also good for posterity purposes) |
| 10:24 | <brad4d> | +michaelficarra done, thx |
| 10:24 | <leobalter> | and I lost my connection |
| 10:25 | <robpalme> | we load DanielR from Teams |
| 10:25 | <robpalme> | we lost |
| 10:28 | <msaboff> | shu: So were talking about searching the web for "await x ** y" or internal google code? |
| 10:28 | <shu> | msaboff: brad4d said he searched internal google code |
| 10:29 | <shu> | msaboff: i was talking about searching https://httparchive.org/, which is public code |
| 10:29 | <msaboff> | Got it. |
| 10:29 | <shu> | it's too big to download and search locally, so you need to craft some bigtable query for it |
| 10:30 | <msaboff> | Yeah, that's why I was wondering. |
| 10:31 | <leobalter> | littledan wsdferdksl: the grammar limits is something I'd really enjoy to discuss but 2:30AM is not a time I can really do it with my brain properly in order |
| 10:32 | <littledan> | yes, I didn't want to get into the discussion today either, but just register my disagreemeent |
| 10:32 | <Bakkot> | the transcription really, really wants to transcribe "annex B" as "NXB" |
| 10:33 | <leobalter> | it's unfortunate my connection is dropping and the conclusion was to follow up with additions to Test262. I hope this comes as a discussion at Test262 |
| 10:36 | <shu> | Bakkot: sounds pretty similar |
| 10:36 | <leobalter> | and then I have a strong agreement with littledan in this different topic. |
| 10:36 | <littledan> | :) |
| 10:36 | <leobalter> | littledan: (echo chamber) who benefits of the optionality? |
| 10:41 | <ljharb> | chairs: i think the queue has the wrong topic |
| 10:41 | <akirose> | it does my b |
| 10:41 | <akirose> | gibson042: can you re-add your piece in a sec? |
| 10:41 | <akirose> | oh wait |
| 10:41 | <akirose> | that's |
| 10:42 | <gibson042> | akirose: yeah, we're good |
| 10:44 | <akirose> | ty |
| 10:54 | <michaelficarra> | ljharb: __proto__ getter/setter are bundled too, right? |
| 10:54 | <michaelficarra> | and separate from __proto__ syntax? |
| 10:54 | <ljharb> | michaelficarra: oh, yes, true. akirose can we have 15 seconds after this to confirm that? |
| 10:54 | <ljharb> | or wait |
| 10:54 | <ljharb> | michaelficarra: the getter is fine, it's only the setter that presents security issues, no? |
| 10:55 | <littledan> | ystartsev: I think our current process provides much of the appropriate safeguards |
| 10:55 | <ystartsev> | ok |
| 10:56 | <littledan> | what we might do is require an implementation or two for normative PRs before landing |
| 10:57 | <littledan> | though this particular thing wasn't done as a small change, it was part of a big feature |
| 10:57 | <ystartsev> | That may be a good idea, we've started tracking normative prs more closely |
| 10:57 | <leobalter> | Bakkot: I heard a mention about Test262, does it need any update to match 2210? |
| 10:58 | <Bakkot> | leobalter I'm not sure, but rkirsling has been pretty on top of those things |
| 10:58 | <leobalter> | I'd suggest to follow up with an issue there |
| 10:58 | <leobalter> | to avoid tests that would now become wrong |
| 10:59 | <shu> | brad4d: it's the same worry as something that has never thrown suddenly throws, even if it's often a logical error |
| 10:59 | <michaelficarra> | Bakkot: kill the bot until after lunch? |
| 11:00 | <Bakkot> | michaelficarra done |
| 11:00 | <shu> | brad4d: historically we've also had issues with changing runtime errors to early errors, like with `f() = e` |
| 11:00 | <ljharb> | akirose: heads up that i can do the default constructor topic any other day, but i'm going to bed, unless someone else wants to handle it after the break |
| 11:02 | <robpalme> | @ljharb default constructors is first thing tomorrow |
| 11:02 | <ljharb> | great |
| 11:04 | <robpalme> | Hallway track is in Hubs (I won't post the URL here) |
| 11:04 | <leobalter> | akirose robpalme not a super constraint, but I'd like to "lock" the Realms presentation as currently set in the draft schedule or for the end of the day (same time slot) in another day |
| 11:05 | <akirose> | leobalter: i'll make a note of it as a schedule constraint |
| 11:07 | <leobalter> | thanks! My constraint is more like a "please lock the time after its set in the draft" so I can get my coworkers to join. |
| 11:07 | <leobalter> | and I appreciate the draft schedule so much! |
| 11:08 | <leobalter> | the draft schedule is one of the most brilliant tools I've seen for the management of the meetings |
| 11:08 | <akirose> | thanks, that means a lot to me. we put a ton of work into it, i'm so glad it's helpful ♥️ |
| 11:09 | <akirose> | i was so nervous to share it when i first started doing it because I was afraid people would get mad when things inevitably moved around, but it looks like the actual upshot is people understand how hard it is to tetris that shit together 😄 |
| 11:10 | <Bakkot> | akirose also it means those changes get telegraphed better, which is great |
| 12:01 | <robpalme> | we are starting the second session now |
| 12:20 | <akirose> | do we have enough note-takers? |
| 12:20 | <michaelficarra> | seems so |
| 12:21 | <akirose> | ty |
| 12:40 | <haxjs> | transferable sharedrealm is great! Hope we can have that soon! |
| 12:44 | <littledan> | as someone who works for a company making a significant investment in JavaScript-on-MIPS, I appreciate the callout to "maybe MIPS" here |
| 12:45 | <littledan> | FWIW here's a sketch I made for user-definable postMessage that I wrote https://github.com/littledan/serializable-objects/ (this is different from binary/JSON serialization) |
| 12:51 | <michaelficarra> | not taking is going REALLY well right now :-) |
| 12:51 | <michaelficarra> | *note |
| 12:54 | <ystartsev> | I am enjoying this a lot more than usual note taking |
| 13:03 | <michaelficarra> | I think the results are mostly better, but I still find it draining and hard to simultaneously follow the actual content |
| 13:04 | <ystartsev> | when there are a lot of errors its hard to focus, but when michael was talking , the transcription was so good that i caught more of it than usual |
| 13:26 | <michaelficarra> | I'd prefer a match-with-indices method or something |
| 13:27 | <michaelficarra> | whether you want indices is a property of the current match, not of the regexp itself |
| 13:27 | <michaelficarra> | putting it in the flags is inappropriate IMO |
| 13:27 | <haxjs> | could exec have new param ? |
| 13:27 | <rbuckton> | A flag doesn't necessarily mean "instead of", but could be in addition to |
| 13:27 | <rbuckton> | we've looked at all of these |
| 13:29 | <rbuckton> | The problem with a param is `@@match` |
| 13:30 | <haxjs> | oh, forgot `@@match`... then i have to support new flag :P |
| 13:34 | <michaelficarra> | btw the sticky flag has already made this category error that we are talking about |
| 13:34 | <haxjs> | maybe we could use a uppercase for those flags affect the result :) |
| 13:34 | <michaelficarra> | all of the other flags affect what the regexp actually means |
| 13:34 | <michaelficarra> | sticky affects how it is used, and this flag would affect how the results are presented |
| 13:35 | <michaelficarra> | I don't think we should make this error again |
| 13:45 | <michaelficarra> | I realise there's difficulties with modifying the RegExp APIs to put this information in the appropriate place, but I think it's worth looking into it |
| 13:45 | <rbuckton> | I'm open to another suggestion as well, but of the four options the flag has the least issues. Callbacks and option bags have subclassing concerns, new methods overcomplicate the api surface of RegExp and limit retrieving indices to those methods (i.e., they wouldn't be available from `String.prototype.match`, etc.) |
| 13:50 | <rbuckton> | Other languages don't need a flag for this because they already provide access to both the strings and indices (usually by doing the opposite of what v8 and JSC are doing and only storing the indices and deriving the strings). Trying to change an engine to do the same ended up with hard-to-optimize internals due to the fact we represent a `RegExp` exec result as an `Array` with a few extra properties added on. |
| 13:56 | <rbuckton> | gibson042: I've been considering proposing the `n` and `x` flags at a future meeting, along with a few other RegExp nice-to-haves. |
| 14:02 | <gibson042> | yep, I've considered the same from time to time. And it's also just generally important to me that JS does not define a RegExp flag used for a different purpose by other languages. |
| 14:05 | <robpalme> | that was amazing - we sustained 38 attendees to the end. |
| 14:06 | <akirose> | 🥳 |
| 14:06 | <michaelficarra> | make sure you subtract one for the note taker bot |
| 14:07 | <michaelficarra> | 47 pages of notes for 4 hours of plenary, that's gotta be a new record |
| 14:07 | <gibson042> | are you kidding, they were definitely the workhorse of the group! |
| 14:07 | <shu> | oh yeah what was i supposed to clarify in the notes |
| 14:07 | <shu> | running on fumes i still got like 2 more mtgs |
| 14:07 | <michaelficarra> | gibson042: still not people |
| 14:07 | <shu> | but do they satisfy a legal definition of personhood |
| 14:07 | <michaelficarra> | shu: search for TODO |
| 14:07 | <robpalme> | i believe bladerunner covered this |
| 14:08 | <robpalme> | and maybe that film with will smith and the moving teddy bears |
| 16:51 | <ljharb> | littledan: is the decorators meeting still happening tomorrow, given it’s tc39 week? |
| 17:59 | <Bakkot> | rbuckton re: callbacks have issues with subclassing, didn't we officially decide we no longer care? |
| 18:00 | <Bakkot> | as long as it doesn't break existing code, anyway |
| 18:29 | <rkirsling> | Bakkot: thanks again for presenting #2210 |
| 18:39 | <Bakkot> | yup yup |
| 18:39 | <Bakkot> | hope I did it justice |
| 18:41 | <rkirsling> | I'm certain you did |
| 18:41 | <rkirsling> | > (sighs) I do not object, but very reluctantly. |
| 18:41 | <rkirsling> | is pretty funny though |
| 18:47 | <rkirsling> | omg I can't decide which of Shu's decks is more amazing |
| 18:49 | <rkirsling> | Bakkot: also re "all hands" -- I just meant that I'm in the midst of two weeks of basically being on-call during waking hours haha |
| 18:50 | <rkirsling> | for PS5 launch (since we've yet to launch in Europe) |
| 18:51 | <rkirsling> | also webkit contributors' meeting is today and tomorrow on pacific time |
| 18:54 | <Bakkot> | ah, well, close enough |
| 18:55 | <rkirsling> | :) |
| 18:56 | <Bakkot> | reviewing the notes - man do we produce a lot of text |
| 18:57 | <rkirsling> | mhm |
| 18:58 | <Bakkot> | capturing everything that's said instead of just the things the note takers catch appears to have ~tripled the wordcount (25k words vs 8k words) |
| 18:58 | <Bakkot> | a lot of that is just going to be that we have notes for the presentation too, I guess |
| 18:59 | <rkirsling> | yeah I was surprised that the presentations got transcribed |
| 18:59 | <Bakkot> | I was going to turn it off for those but a couple people said they'd prefer they be captured |
| 18:59 | <rkirsling> | ahh alrighty |
| 19:01 | <drousso> | i think it depends on the presentation (e.g. if the linked presentation has speaker notes or lots of in-slide content then it's not as needed) |
| 19:24 | <Bakkot> | yeah. but in practice I think all or almost all the speakers say a lot of things that aren't precisely captured by the slides and speaker notes. |
| 20:02 | <jridgewell> | There are soooo many notes this time |
| 20:02 | <jridgewell> | We should continue doing this. |