Node.js Core Technical Committee, 5 Oct 2016

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: 2016-10-05 Node.js Foundation Core Technical Committee (CTC) Meeting

Description

https://github.com/nodejs/node/issues/8915

A

Foundation ctc meaningful the fifth of october 2016 and let's get straight into it,.

A

So let's do a stand-up of who's here on. Are you here? Oh.

B

Yeah but I have nothing really to report, so I, pasok, right to Bradley.

C

um I went to tc39 last week you got some leeway to adjust the spec to our needs, but I'm still in discussions for some of it. So I need to send the PRS to tc39 and see how I guess.

A

Colin I think you're. Next, we don't have been no Chris.

D

I also don't have much to report this week, just issues and pull request.

A

That would make heaven yeah.

E

I'll work a little bit on improving the commitment, winter and then smita, PR or yesterday, from proving performance and process next to it. By like ten twenty percent. Next is Jeremiah.

F

Yeah so today I've got the exists: Inc FS that existing and application emerged, which has been signed off by many people and came up here multiple times and then other random PR is an issue. Stuff I, don't think, there's anything else. Notable Tracy's next.

G

Alright, let's see so, we've got the AL Ricci application drive in full swing. If you know folks who are eligible that would be rad to send them our way. The information is up on nodejs, start org, slash, I, think, outreach, e um and I'm sure the mentors will express this later on, but we've got the project.

G

The suggested product projects up and supporting those mentors will be awesome if anyone is interested in helping with that, we still have that issue open under the CTC and, let's see we met late last week to talk about code and learn and collaboration summit and how we can set up the Austin event for success with that. So stay tuned for issues that you all can help contribute to make that better.

A

I, don't believe we have video here, and so that would make Michael dolzhny you here. Yes,.

H

I just joined so I spent some time trying to remember, since the last meeting spent some time doing, miscellaneous peers reviews and lands fixed up an issue with one of her be a test that was running.

H

That's about all that comes to mind right this minute.

A

I

Nothing much except investigating some specific issues, both on the on the bill side and on on on the CI platform.

A

Michael Rogers.

J

Hey I'm, mostly just working on the new budget for 2017 for the foundation. um I was also at East 39 last week, with Bradley getting stuff done um to productive, looking into ways for me to be able to go longer term as well and kind of keep things moving and that's been. It.

A

K

Looks like you filled out to stand up in absentia, I.

A

Don't believe ollie is only heal, no sue, then that's gin, no update just observing myself edge, which I haven't filled out because I haven't I've been typing for other people. Oh, so a little bit of foundation stuff some some build work, Aguerre with I'm hoping well, not really helping I'm Johannes refactoring, all the variants ansible stuff, so I'm I'm watching on his stuff and an testing sort of it out that PR should land soon.

A

Hopefully it's a pretty big refactor and back onto the LTS read me: rework trying to get really would like I get that landed, but there's some more stuff to be added to that. Since the last time it was reviewed, Seth Thompson, are you with us? Yes,.

L

Hi there, let's see the main updates on RN async await, is still on track to ship in VA version 5.5, although we haven't reached the branch date quite yet, so that's still so I'm confirmed official aids and be an inspector work. Ongoing and yeah I think those are the main updates from this last week you.

A

I, don't believe we have she aqui here so Stephen Lewis now he's Ortiz away today, tis me I, see you 58 will be out in a couple of weeks and we'll have an updated PR at that time. Miles.

M

Hello so then, working on the 4x back boarding defining the a treaty project that I'm going to mentor on fixing some regressions engineering, the gold mine and working on new communication plan for LT estates. Next up is a the fourth I.

N

Hi I was sick. The Holy I am I've, been declined on my fever, so nothing no soap it. For me.

A

Next is true: I.

D

Finished work on some performance Russians for facing cookie p, /, PR I'm almost done with it, so that it will be feature complete with what the EP outlines and then we'll be doing tests and eight API documentation.

D

Next is rich I've.

K

Been working to put together the ctc meeting rotation proposal with a lot of help from Nikita and doing that was more tedious than I could have possibly imagined um and we'll talk about that in a few minutes, not the tedium, but the actual proposal I also been. I sent the you know, Oh as everybody's on the CTC list knows: we've been talking about two-factor authentication for collaborators.

K

I began a lot of help from Nikita on that and reached out to collaborators who don't have two factor authentication and talk to ask them about the possibility or any obstacles to turning it on and other than that lots of miscellaneous issue tracker commit activity, helping some know to do type folks get their first commits and fixing a test here, they're finding it finding that Michael Dawson was not on the CTC email list and adding him and oh okay,.

H

Well, thanks I didn't hear some messages. It was my fault. That's.

K

Right anyway, uh let's see here next up is Josh: hey Josh. How are you hey.

O

I'm good, how are you guys um so not so much for me, I was actually often most of the week, but we did just hold the diagnostics working group meeting, which was nice and productive, to look for the notes on that yeah and that's it second, a highlights.

F

O

Some pretty cool tracing stuff that the VA team is is working on.

F

Alright sounds cool I'll check it out. Yeah.

J

A

Previous meeting review, we talked about the scheduling meetings thing which were talked about again this week. We had a bit of a talk about node module version of for ABI number and we moved that back to get hard, but it was I think the the reprisal the proposal stood and wasn't any major objections there. So I was just a discussion. Then we did a little bit of version, 7 release and version 6, LTS planning discussion, stuff, not hugely eventful meeting.

A

So this week we're going to jump straight in and look at the first of all supported platforms list which is pull request. Number 892 to this came. This went across to the bill working group. Tom johan actually did most of the work here. To put this together, collaborating with the.

A

A bit ly bv came and also looking at some of what a v8 supported coming up with this list. It's had a bit of discussion and yeah.

A

So, as has everybody had a chance to look at this, that wants to look at it, don't believe, there's anything controversial in their exact. A couple of nits of typos well grandma.

H

It's pretty much I'm, just looking now I hadn't seen that is pretty much what was in the working? Yes, yes,.

A

It the table has not changed. It's the I just edged some of the text, bringing it across just to just some copy and editing.

K

Can you clarify what there's that bit that I don't understand about since we rely on some dependencies, we need to adopt their supported platforms and potentially add to their list. So so that says that if v8 and the boobie both support it, we support it or what does that mean exactly hang.

A

On a minute, that's.

A

Is that recording directly from there oh yeah, we.

K

Mic and it sits under input, is a headless. We rely on a few dependencies that make us who we are most importantly v8 and lib UV. We therefore need to adopt our supported platforms and potentially add to their lists based on tests and our release coverage I.

F

Think it could be worried better. I'm pretty sure it means that if they don't support on platform like we kind of adopt that by default and if we want support for that like we need to add it ourselves.

A

Yeah, I think that needs clarification but yeah, basically that and we're also constrained by them. So it goes two ways. Oh.

K

Cool uh yeah I'm other than that I'm really really happy with doc. The.

A

Fact that the fact is that, most of that, our our support platforms list is I, think it's narrower than both Libby v and v eight. So it's we're not like. We can more constrained by them than we are the other way around. Yeah.

H

You know to say that we're a subset because their constraint.

H

A

H

Add to their list yeah.

K

Yeah, it's the confusing part, so I mean I would. I would recommend, clarify that or even consider just removing it I don't know up to you really but yeah. It's something I found confusing.

A

Yeah, removing it wouldn't be the wake up way to go.

A

mmm Okay, no I'll I'll I'll do something about that. So there any objections to this list as it stands. This can be edited along along the way and, as I said in the PR its it is targeting version 6, which is probably going to be a same as a version 7. We needs to be changed for version 4, so we'll have to backport it so I'll ping. The ctc again when we do the back porting.

A

Okay, well, doesn't select any objections in here. But if there are concerns, objections and you've gives you got time to go into that issue and and say something before lands, but as it stands, there's no major objections, except for the wording of the text around the list.

A

Okay, easy next one is the this. This is James's issue, but I think we can still talk about int'l, consider deprecating, int'l, dot, v8 break iterator, and this comes from the VA team.

A

The sonra here is that they want to deprecate it because they're moving towards a more formal tc39 prefer a language model for in int'l dot segment, ER api, which would make this thing redundant. Little dan from the VA team has said that he'd like to he hopes to deprecate it in a future version. So.

J

Yeah, I think the big thing here is just that um as soon as possible. We want to get any native developers not using this API, so we don't break anything and basically trying not to expose it to our native community native at our community, because we're moving towards the standard that can actually be supported long-term. It's.

F

Exposed in JavaScript it's.

A

A JavaScript API yeah.

J

Yeah we do, we do expose it in JavaScript, but we probably shouldn't.

A

um We expose it because VA does like we're not actually doing anything here. It's Viator's doing it I believe yeah.

J

But we tended to six pose that the it exposes and I think what they're saying is that they don't want to support this API going forward and I mean we need to provide our users an alternative to that API. So I think the proposal is that we actually just like implement the spectrum now and don't expose the v8 course.

C

And the API focused by VA is configurable, so it can be removed.

A

So I had to comment in there. One is that it's before tc39 than the API, but that suggests it could take quite a while to get in through that whole process and the other thing is the uncertainty around when it will be removed. So he says I hope to remove it entirely from a future version of v8.

A

But you know when might that be because that the impact here is we've got version 7 coming up? Is it possible? This thing will be removed at you know in v8 5.5 5.6, which are versions that we may land in version, 7 note version 7 and if so, then that would be a breaking change in the API and we'd have to backport we'd have to polyfill the old API, or is it something that might be removed the where we can play in for version 8 because it'll be removed in v8 version 60?

A

So it's the uncertainty that I have concerned about here, because it's yeah.

J

I don't know if we have a great answer on that. I know that so Daniel from VA team is pushing both the standard and I think basically has to maintain this API, and so he wants to get rid of it as soon as possible. But I don't think that pervy policy they can actually put in the new replacement api until it makes it into a certain stage in the spec process, so depending on how quickly that listed the spec process that this will get removed before after that, right.

C

Yes, and they also discussed, they wanted to remove it immediately, but something like 0.02 percent of websites use it, so they can't remove it until that drops some.

J

So I don't know about that. Then we be able to remove before that. I think that the main thing is that they have to have an alternative vertically as when they do remove it. So Oracle.

C

E they haven't removed, left that's greater than triple digit precision.

K

Seth, here's something to say here on this: yes,.

L

I'm trying to think daniel right now to get to get his most recent. Take on this I think that generally, we will wait until a feature at t. 39 is a stage three before we implement it.

L

As for a sort of removing in Tovia break iterator, we rely on a couple things, including usage, but we don't have an absolute number. I would say in this case yeah. I guess I'm not sure how much in Dan's mind he's dating the removal on the success of Intel segments ER, but but I can follow up and I mean yeah, I, guess I, don't have anybody else, but it up, but certainly I, don't know how this interacts with the house. How support on the node side interacts with the requirement to use?

L

Oh I see you because I believe it's only exposed if some full seu is just as a simply is not the default.

A

I'm not suggesting I have trouble with us, removing it. It's just that the timing. Uncertainly is a bit strange if.

B

A

Had a signal that said yeah this is going to be removed in 55 or 56 or whatever um you know some some soon version only future version then great. We could go ahead and do this and it would fit our timeline better. But if it's, if it's like we'd like to remove it, and but that may not happen until you know, 65, then I'm not sure why.

L

A

L

Question is that it's a bit more aspirational than a concrete plan right now, but let me let me confirm it down and see if I can get smart for fishermen.

A

Ok am sir james is driving. This will leave it in his hands to push forward more and move back to that issue. Discussion that issue 8865.

A

So, let's move on for this meeting and thanks for the input, dear sister, that's really helpful. The next issue is just.

L

Mention one one: quick note that I think part of the reason that they would have been activated by this Danis dental office service and next week. I believed.

A

Okay, cool. Thank you um we'll make sure James knows about that.

A

Next, one is pull request. Number 8 for 19, which is multiple little events fail silently this one is.

A

Take this and James change it amana. Would you like to tell us about this one because you're involved in the conversation.

B

um I, don't know it, James isn't here, Ryan, no.

A

B

B

O

Thing is good, I think.

E

M

B

Think he was the one who put it on the agenda and.

M

G

B

Wait for the last couple of days so I'm.

H

Pretty frustrated because it's a potentially breaking change, but there was a request to get it into seven. Yes,.

A

Okay, well, I can, I can give a rundown then, if need be, and I.

B

Think they've good.

A

Ok, so it's simply that you can call listen twice on a a socket so on the server netserver. If you run, if you're listening, you can call it again and it will be a silent failure, but in doing so it'll change the connection key and pipe name. So it's it's. It's obvious. It's an obvious bug, so the discussion here was around solution, whether it should emit an error or something else.

A

So the the pull request here, which so far been last one I, think it's bits. Fine, like there's no objections to. It is simply to track whether it's been listening lot and if you called it, then you get a idiot meets an error event with a an error that says: listen method has been called more than once, and the code is manually added to be e, addressing use since the manual code that so that is the proposal.

A

B

Ya one thing I'm not exactly sure about and why I think I don't have expressed a really strong opinion on that PR is James. Had the pr open a while ago, I am gonna post a link here, it's a 294 um where he documented that I calling listen multiple times would actually reopen the server, and you know I read it as something like. If you call listen again, it will work just with a different part or sub, and that didn't seem to be the case. So I'm a bit confused about that.

A

Alright, that's so that would be listening on something that's already closed by the reading. Reading that change.

A

But but but you're right, the conflict might be that this PR I think tracks I, don't think it actually tracks we're not listening anymore and you listen can be called again because it's about this variable called hat. Listen has been cold, there's just a boolean and it's only set true. It's never set to false again, so that does sound like a conflict.

F

From what I read, it should be set to false if, if it's closed.

A

Sorry, you know enclose it set to false.

B

Yeah right I wonder.

A

Is that the only way that sockets can be closed with the close method called.

F

I doubt it yeah.

A

That's what I'm wondering here, I.

B

Would actually think it is the only way, at least far for listening, sockets.

A

I'm gonna yeah see that's tricky one. This might have to come back on the agenda next week, we'll let drink, James drive it through, but I'll mention that it's gone back for discussion and I'll drop in that note about them.

A

The other issue that you mentioned that.

F

A

That, before we move on does.

F

Anyone have any particular objections against this.

H

B

H

We accept December major in 27 at this point.

K

That's the reason it's in this meeting yeah well.

E

Has there has anything been rolling against the ecosystem? I guess that one's gonna be a bit more difficult to check for.

A

It very difficult and and the fact that you can trigger bad behavior by doing this, suggests that if somebody is actually using this, then you know this is like it's like that: XKCD one where the guy uses his cpu fan to do something in emacs.

K

Pulls down the spacebar that makes the CP overheat and the fan kick Sonia and Sonia.

K

So yeah- let's, let's, let's, let's punt this to next week when James can be here, for it is my vote. Yeah.

A

Okay: let's do that and be easier. A crook.

D

Fyi, if you, if you run, listen twice on the same on this same server object. What happens is that it overrides the underscore handle property. So basically, the the user will lose all access to the handle and can never again close it, but it will still sit open.

F

That's pretty bad I guess.

H

The quick can they can they can in the new one that they've got. Can they continue to accept connections on that like is it that they get a new one? They can connect, get exceptions on or they then broken from that perspective as well. You.

A

Mean often stuff to close has been told so.

H

After they've done the second listen, you know you basically lost the first handle, but if you have a new one in it, you can continue. You can get.

I

H

Up connections or not this.

A

Poor request is about rejecting the second listen if it's already listening, yeah.

F

I I think like they might. You might actually get an error from trying to listen to the same socket. I'm. Sorry, it's a mattress, I, don't know anyway. We should take that back to the issue. Yeah.

K

I think we do ourselves a disservice by not putting this in the issue right and how did the conversation here because we're just going to have to have the same conversation next week? Yep.

A

That's right, and so, let's for the sake of productivity, let's move on so next one is this scheduling meetings it's an issue in the CTC repo issue, number 14 and and rich. You have done an amazing job with this time zone to the worst so tell us where we're at here.

K

ah So I'm just going to cut right to the latest greatest proposal, which I will paste a link into the sidebar. It should be there I hope. Yes, this is the least terrible proposal. I could come up with all all possibilities are bad. It's thinking the least bad out of the options so talked about this in private email. We've talked about it on in the issue, tracker, so I, guess I.

K

You know, there's more to say: I guess we can say it here now, but ultimately, I think I was hoping that we would be able to either agree to give the schedule the for the next four meetings that I put in that link that I just posted, give that a give that a shot for for for you know for weeks and then evaluate how it is or isn't working and what might be better or worse or if we're just going to shoot this down and and move on and try to figure something else out.

K

So that's where it's at.

A

So that that's for the next week it would start October the 12th.

A

A

A

Ok, I'll unlikely be did for that one, but that's fun. That's.

K

Ok, we'll do everything. Asynchronously will be amazed.

A

Okay, before we jump into next week, shifting time with, are there any objections to doing that because it was really about next week.

A

So it so next week is going to be UTC four o'clock.

A

Which is a bore, clock p.m. or got PMS aria, which is u.s. west coast 9am US, East, Coast, midday.

A

No okay, that will be that should be much nicer for people in Europe and also India is 930 p.m. so that's much more manageable. I think.

K

I went down with the way more way less discussion than I thought. I was going to go ahead. Good psycho, Priya.

N

Yeah Mantha to be in the bar rather than one. Thank you thanks for the confirmation.

A

Okay, so that meetings yours next week, rich so cool, okay, make it so um yeah.

B

A

And so general version 7 release, planning and LTS and discussion. The only thing I have to add in here before anyone else does is that the LTS working group agreed last week that we'd put out version 6 LTS a week before, which makes it then sometime next week, not next weekend or the week after the 18th.

F

Pretty sure it's the 18, it.

A

Tends we just choose J, so that's not next week the week after so that's when version 6 would go LTS.

A

And then, the week after that, we would I aim for version version soon. If.

M

We're aiming for the LTS release on the 18th. We should try, probably try to get in a release on the 11th that has all of the Denver minor stuff that we're looking to get in before hand.

A

M

A cool I was thinking.

A

About that yeah, so who wants to do that? Who wants to do a week release next next Tuesday.

E

You're, my do you want it since that's it, the last one did.

F

You think you took the security release right, yeah.

E

F

Take this one: okay,.

E

And so either why.

A

Okay, so Jeremih you'll do a release next week and if you, if you need to plan that off to Evan, make sure you do that sooner rather than later, and then one of us will do sig version. 6 LTS I'm happy to put up my hand for that. Unless somebody else is really keen to jump in.

A

Seven is that is there any on the v8 side? Are we looking to be on track to four v8 5.4 going gold before the 18th, maybe further when they take the what the week after that, that's.

H

A

Everything looking good there, theft.

L

um I I think Ali is the one working on actually merging into the node tree, so I'm not certain on the way to status there, but I haven't heard anything negative.

A

Okay, so on track anything else to discuss about the release planning here, I, don't think I still think we haven't, got somebody or somebodies putting up hands for helping with or even driving version, 6 LTS back ports and releases. We can't just say hey miles: you do that as well and I think miles wants to do that and that we also just need a team anyway. So we're still looking for people to put their hands up for that.

F

Alternatively, um now this would be up to my hounds: we could, we could do a switch, so maybe miles does v6 then, and someone else can do before before from now on, we'll probably get a little less commits go into it, so it might be a little bit easier. It might be more difficult, I, don't really know.

F

You have any thoughts, a lot miles.

F

Looks like you're unmuted, but we can't hear you.

F

K

You waiting for miles to reconnect.

D

M

Hey thanks for being patient. Can you hear me now? Yep, okay, excellent, so yeah like I, mean I kind of have some mixed feelings here and I think that this will be a broader question for LTS I. Think as long as v4 is still in active that we, we have a you know a commitment to actively be making sure that we're going through everything now I do think some of the stuff that we talked about in this week's LTS working group meeting about tooling that can automate.

M

Some of this process will make the whole process better, but I kind of feel like we need to be giving for still the same attention that it's been getting as long as it's in you know this window there will be a lot more stuff that needs to be backported 26, but a lot of that stuff will be a lot clearer as far as you know what it is and when it should be back ported, you know, I think the biggest thing thing that we could probably do, and maybe I'll mull over this before next week's meeting is just maybe coming up with a a better process as far as how things end up in staging- and this is something that I think we need to be doing for current as well, rather than just kind of like getting everything in there at the last minute and just kind of coordinating how we're working with that staging repo.

M

This staging branch- but I think you know if all of us, at least on the CTC, are making an active effort to label things and backport things. In this case, the load of this can be distributed in a much nicer way.

F

um I think I like kind of disagree about the fact that like well v4 is an active. It needs to have the same amount of like activity the entire way through from the start.

F

There was always at least in like I, don't know how written down it is, but there was always some amount of explicit understanding that over time, that would be become harder and harder to backport as much as possible, and so activity would drop, and that was an okay thing, but there would still be more activity than maintenance where maintenance is pretty much okay. If there's a security issue or something's really critically broken in this branch, but specifically, then we go fix it. Otherwise, we don't really back part anything to maintenance, yeah.

M

Well, I mean like most of the work is literally just going through and auditing the backlog that hasn't had labels put on it, and that was the kind of thing that we were talking about, could maybe be audited in a nice way. Right.

A

M

But I found a non-trivial number of things that would have just gotten left. If we didn't do that. ?.

D

M

A

Thanks for that, Rick.

D

I'm not upset no meaning or.

A

Is that Trevor I.

F

Think they're ever ever.

M

A

Think you need to meet.

M

That's okay um to do so so yeah so I mean I. Think we should talk about this. A little bit more offline, intimate ways to prepare this bit um I'm a little bit uncomfortable with not auditing everything, or at least like not having some automated process to audit, because the things that would get missed, sometimes even if they seem trivial, aren't super trivial but again, I think this is something that a little bit more work on our end to combined with some tooling, can make the workload a lot later.

E

Miles I'd be interested in helping.

A

E

That too, luhley.

M

Excellent slick, let's chat tomorrow, but that.

E

Works for me, but man.

F

If you would please.

A

Right, ok, so we're hurtling towards version 7 things that I'm getting ticked off. My only concern about version 7 now is actually this v8 procure the right a thing. If so, that's gonna probably be there's barely a bit more pressure to get that that resolved within the next week.

A

Maybe so, if you can, it might be good to turn into the discussion and see how it results, because if we should, if we need to deprecated it or remove it, then we probably should be doing it with version 7.

A

L

A

L

Bit more context on the you were talking up: the Intel v8 iterator, yes, yeah I, just a just got so as mentioned and out and he's the one driving driving this whole process, but the language suffer from some. Some decimal language came they're, pretty sure that the plan is to not deprecated anything before Intel segment. Ders is reaches more.

L

You know, conclusiveness within piece of 39, but everybody is epic fail, not that we have a say, but everywhere everybody would be fine with node deprecating it earlier, and you know it may just be a forward forward. Looking thing to do, given that it's already only usable for a subset of node users, anyways yeah I, don't know if that helps make anything more country.

A

That does help actually I'm pay some money to the issue new for.

F

Subset is that? Because it's only available with full I see you yeah, oh okay, good.

B

A

Is there anything else about v6v 7, there's also v V 0 10, that's in the mix you if anyone has anything to say.

F

Does anyone have anything to say of 0-10 I have a couple things on seven point: oh.

F

Okay, then I'm. Looking at the seven point, 0 point 0 milestone on noches issue tracker. There are currently six open issues still um and none of them I've been discussed today, so I just want to give people an idea what's they're so that we can get this thing. I cleaned out a bit. There's a discussion on openssl one point: 0 point: 0 planning that was opened by rod um close to a year ago, wow, there's a potential thing to update inspector ready message to be vendor agnostic, I'm.

F

Currently, a printer Chrome I'm, not sure for not sure. If there's anything else, that's up to speed with the inspector actually.

O

We because we did discuss that briefly in our meeting today, we had a couple possibilities: I think we're I think we included to go with the message, but add a line for now pointing to wiki page, but there's a few up options.

F

Okay, so we can move forward to that yeah.

O

F

There's some planning for v 7 and the rollover for v6 Delta s by James, there's some noches event listener bug!

F

That's for some reason: tagged as some point: oh I'm, okay, workin progress, steps, upgrade Libby, v2, 1, point 10 point: Oh, bicycle I'll follow up on the Lots fall upon all of these in the issue, but just to keep people in the loop and also there's another thing back on the opus ssl one point: 0 1 point 1 point 0 thing: there's a poll request for that, but it doesn't seem to be either working or something right. Now: okay,.

A

F

Does anything anyone have anything else? I will say that 1.

A

1 point 1 point: 0 support is not going to happen. It's not gonna, so we're not going to ever will in the near future. We're not going to have 1 point 1 point 0 for openssl support in in node, but what we may have is the ability to compile against it, and that would be ideal, particularly for disciplining statistics that are upgrading to 110.

A

There's a pull request for that, but that's that I don't believe that in any way a block for version 7, it was just put on that milestone just to make it fit some way, but that pull request could land at any time. It doesn't have to wait for a simple major.

A

A

I have a note, then, about V 0 10, which is that there's a new version of NPM version, 2 out that finally fixes the license document, we're still shipping the old license, I believe in node version, 10 and 12, with the version of 0 a version of NPM version 2, and that we should being there. So this is a question about whether we want to ship another version of at least v10 0-10 with an updated NPM that has the fixed license. Is it important enough.

D

D

F

Not hugely opposed to it, but there's.

A

An issue in the LTS replay finalis, if anyone be shown what's the way on that, then go to the LTS repo. This is an issue there. We can discuss it that otherwise this can be deferred to the LTS working group in stock. We'd have to waste time here on.

A

Anything else about 30, 10, v6 or v7 before we close up.

A

No okay: let's move on to a QA time, Q&A we're for people listening in via the live stream. If you are listening in, have anything you'd like the CTC to discuss answer or it's just some input for us I'm feel free to drop a comment in YouTube there or you can do it in the issue for this meeting in the DOJ's repo or ping, one of us on IRC home. So we're gonna have a will just wait a minute for this stream to get to catch up.

F

Me until I'm, where the next meeting dates.

A

So the next next meeting got four other meetings. That's a good idea! I'll read that upcoming meetings, the ctc next week on the 12th, but that's shifting to UTC 4pm tsc, is tomorrow same time. Slot. As this there's a build working group next week, burn witches am the 11th. Whatever day those Diagnostics is first week first week of November, so that's still got a another month to go. Lts meeting I'm not going to be on the third but that'll be I.

A

Don't win that one will be I, think where we get to weekly for that one, so it would be. It would be next week, it'll be the week after.

F

Alex asks hey: could we get a comment on the stability of v6 at this point? Just want to get a sense of that for the purposes of upgrades, etc.

A

Well, I think, as discussed. I think the answer to that is simply that we're going to have a standard v6 release next tuesday that that's just going to be like, as in v6 as current and then the tuesday afterwards we're going to be dropping into LTS and that's when the stability push will start and you should expect it to be as stable as the v4 releases.

A

So usually, if you're planning on upgrading to v6 LTS, then you probably should be testing it now, and you know, unless it's not urgent for you, but it's it's worth testing it now, there's nothing huge going in there. That's good! That's going to upset the applecart.

F

I'm as I know, next week's v6 release will probably almost certainly be a minor, but there shouldn't be much too much in the way of new features that land on that. So.

A

Yeah, we don't have any outstanding regressions. Do we I.

F

Think the last couple of these extremely cesare been pretty good. In that fact, we had some earlier on in the v6 life cycle that we had to get releases out for afterwards due to progressions introduced, but we've been looking pretty good lately and I. Think I think.

E

There is one Russian in the VA inspector.

E

Camera where it is III.

F

Think I know what you're talking about that's been around since six point: four point: oh ok, I think I, don't know if we've been able to reproduce it or something I, don't.

A

Know whether a v8 inspector I think I think it's we're still marking it as experimental that we rock the arts, marked as experimental that may change during the v6 life cycle. But but it's going into this, it's not going to be removed like we put on the agenda to discuss, but um it still gets going in this experimental for now, which means that it's not not something that you should treat as stable as the other features.

A

Any other questions.

A

A

F

We say thanks but other than that I think we're good.

A

Great well, we've got four minutes to spare, so thanks everyone for joining us. Thank you. Everyone for listening in on the live stream right, see you next week at a different time. Slot.

E

F

As a reminder, the next week's meeting will be at collar.

H