Jupyter IPython Dev, 23 May 2017

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Jupyter/IPython Dev Meeting, May 23, 2017

Description

Meeting of the Jupyter/IPython development team, May 23, 2017

Meeting Notes: https://paper.dropbox.com/doc/May-2017-Jupyter-Weekly-Meetings-NaidawTLtrI06iFJzIdwZ

A

Held up hello, everyone and welcome to project Jupiter's weekly developers call this is Tuesday May 23rd we cancelled last week's and so big part of the group is out at PyCon and I will go ahead and get started. We have action items from last week and I think we can start to check some of those off and maybe delete some of them. There is still a blanket reminder to register for Jupiter con at Jupiter con-com. We have the discount deadline for that event. Coming up in two days, which is this Thursday?

A

Is the they'll be in the next price jump, so yeah don't forget to register for that I think the PI data Paris happens already right is I. Think Sylvan did feels and go to that was that sylvans event? Do you speak? Is he on the call? Maybe he wants to talk about it? I don't know not on the call. Okay. Maybe you can tell us about that next time, so other action items from previous month, I don't is made on the call.

A

Does he still want people receiving it is on the call, then, do you still yeah.

B

A

People interested you want to talk about. Oh.

B

Yeah we're going to do that, some that was at it to the meeting agenda for this week, so we're gonna go through stuff in person. Okay,.

A

So should I take that off of here. Okay, all right Mike is Mike on a call and Michael's.

C

Not here today, Mike.

A

Is not here today, so if you have thoughts on user stories for configurable exports looks like you should reach out to Mike I pie widgets, who talks about that.

C

Anybody I'm nipa, sylvan and jason. That's a reminder for deployment 7.0 release. Okay! Does that still need to be on here? Wow! That's a feather in the hat reminder week back to 77 supposed to be released chased mostly. You can provide an update on that when we get to that section.

A

Okay, so go ahead, leave it okay got it grant. Do you want to talk about closing out Jupiter on notebook issues, I'm.

D

Sure I've been busy weeks with another project, so I really haven't made a lot of progress, but that is still after I wrap up a couple things. That is what I will be focusing on, and if anybody would like to help me that'd be great but I'm not really expecting it. It.

C

Is so fun you just really bug that out like.

A

It I can hear it in Grant's. Myself is super exciting about it.

A

Okay, so on to the next one Jupiter lab who wants to talk about that.

E

Yesterday, big, the extension mechanism should be fairly settled down and documentation is out there. We're still looking for more folks to write extensions, gives feedback, there's a nuclear hub extension. It's been recently updated and Jason is also recently updated. That I thought were just one. So there has been some feedback on it.

A

Okay, awesome, so the last one on there is looks like Matthias is looking for feedback for his repple implementation and there's a link there for anyone who's interested in doing that. So now we can move on to do the leaders of help me out here guys the leaders of these sections of the weekly news by organization is that kind of just anyone who's on the call, or is that the same people every week? How does that work.

F

I update the documents each week, but we don't typically talk about it. It's there for review for people who are listening at home or following along okay.

A

So those clink those leak, those links lead to kind of individual places where people can look for news on those topics. Correct got it okay, perfect, so any props, I'm moving on to project management, then Jamie. Do you have any updates? I know you sent out those um the diagrams for the different teams, anything else on that or anything else. You want to talk about a.

C

Dummy in gave me a thumbs up on that, so thanks so much for the feedback dummy and it's nice to get any kind of feedback on stuff, so he just gave me a thumbs up, so that was great I'm just wrapping things up. On my end, as most of you know, my last days, Friday and I'll be starting some data science stuff summer, but I I can give a little update on on things that have come up here at at.

C

Where are we hike on and sorry I have like Jupiter Khan suck in my head, oh yeah, so yeah pipeline, you know folks, are starting to ask about moving to Jupiter, lab and expectations for, like the rate of the expected transition rate from the classic notebook to Jupiter lab, and one thing that I've just in responding with is you know it's we're, probably looking at like a year year and a half horizon on the transition and brian is just joined, so he can chime in if, if, if there's additional information that you want to add there Brian so folks are asking about the conversion from the classic notebook to Jupiter lab and asking when Jupiter lab will be coming out.

C

So I have told them that we're aiming for sometime this summer is that is that right, Jason scene, I'm, getting a nod from Jason, be pretty good. Oh great I'm.

G

Eternally optimistic, though, so about that nun, yeah.

B

G

Think if we had I mean I really hope that we have a particular con people beating yeah, okay, but that would be this summer. Ish yeah yeah.

C

G

C

Yeah in a buck, favor, okay, so somewhere summer, somewhere great um and let's see I- spoke with an unfocused folks. They have a PI data, Seattle coming up July 5th through 7th right before thigh-high, and they would love a talk from the Jupiter team.

C

I thought, real-time collaboration might be a good good thing to share with the PI data community, but you know I would of course want confirmation from mr. Ian. So if that's something that you're interested in thinking about might be a good test run for site I or jupiter con you.

H

Know I don't know what the status is that I had also offered to them to give a talk on jupiter lab, okay,.

C

H

Don't know I'm not heard back from this one: okay, okay,.

C

I'll circle back up to you, so it sounds. We have two options. She said she has a spot for us, so Kirk.

C

And I think that's about it. I'm really trying to wrap up some things like I said I'm like moving over our note or hackpad notes into Dropbox paper, so I'm hoping to finish that by the end of the week, and then you know here and there I might help grant out with was closing and following up on older issues.

A

Okay, great, can you still hear me: okay, I'm, having some issues with my blue jeans. If we can hear you, okay, great dear all, frozen but I'll, just keep moving on, sounds like the audios working, so we're.

H

Going to skip confuse.

A

Okay, great so we'll skip over and be convert since Mike's, not here today and if there's any updates on the next couple of sections, they don't have notes in them. Does anyone want to talk about services or the dashboard.

A

If not we'll move into Jupiter lab, but just wanted to address those before moving on, sounds like no and.

C

Then we located for a backup for a keeping track of MB viewer is that crossed off in the action items?

C

Last time when you were looking for a backup deploying idea, we were we.

B

Are going to do that next week in person? Okay,.

A

Okay, so let's move on to Jupiter lab, then any updates or things that you guys want to share on that sure.

E

We released version 22 last week that included several enhancements from new contributors. I think Brian is working on a blog post about that as a hackathon a couple weeks ago, also working on cleaning up, the API has stability and some last last remaining beta issues. So we can calm this down and have a stable API for beta as much as possible.

E

Yeah Darien is continuing to work on the Settings system. I've got most of the client-side works, done right now, using local storage as a proof of concept, and we would plan to move that back to using the server-side storage for settings as a follow on PR and then Ian opened APR and enhancing our contents manager to add the ability to map different drives, which he plans to use for the real-time collaboration. So that should be the last major change to enable real-time collaboration.

E

We hope that and again that's part of things we're trying to do is get these major major API changes done before we wish yours to beta and people start building extensions in earnest and that's it.

A

Okay, thank you, so anything on IP, widgets.

G

Tons of nuclear's we release into alpha we're, making good progress towards a beta I think we have 11 issues and that keeps going up and down just slightly as we open an issue and knock down to and open another one. We have a bunch of new contributors, several new contributors, including from the hackathon that was just mentioned, and yeah between 20 and 30, PRS Birds, so making good progress, and hopefully, with the Vegas one we're racing to their lab now. Good kitty.

A

Nice, okay, so Jupiter hubs.

H

And everything we're spending.

B

It I :. First, you grab the kind of thing and are getting stuff going on kubernetes and getting binder, which has been in a bit of trouble rebuilt on top of. Perhaps that's going play well.

F

And then just a little feedback from icons. There was lots of science and data science stuff going on here. So it was great to have jibber folks out in force and many many many many many attendees wanted to pass along their thanks to the entire team for all the work that everybody does to make Jupiter to stop so really successful two weeks. So we're looking forward to seeing everybody and catching up in person, nice.

A

It's good to get that feedback in person and connect with the community. That's all it's.

G

Rent work is being done, is their sprint work on obvious on Jupiter? How bizarre others might work that strictly related and.

F

Not any official sprints, it's pretty much men and you be and I are here- we're kind of very centrally visible within the Convention Center and outside all the sprit rooms and people have been coming up to us all week and asking us informal things. So I will be here through Tuesday early afternoon and probably integrating more with the Panda spoke week goes on all something.

A

Nice, so any anything on ipython doesn't look like there's any notes, but just.

C

Calls to you Oh kind of lose anyone else on the call has s they want to add on ipython this.

H

This is Brian and one of our designers this summer, at Cal, Poly, started to lock up some redesign work for the ipython website and I. Think that the main question we'll want to answer is what what content do we want to continue to host at the ipython website?

H

That would love input from the folks who are spending most of their time on, like I, think Colonel and yeah, basically, two girls with what what do we want that separate website to look like this point from the content perspective and then working with the deployment visual, because that's something that's just getting spun up at least wanted folks to know that. Well,.

C

When are you when are you meeting on that? Is that, like in the next week or so or there's.

H

No no planned meeting yeah, it's just students, basically of their own initiatives, start to just work on redesign stuff, so I see some early, mock-ups or gonna try and try to meet with the students. But there's no particular time in great.

C

If everything goes well, we should have a new technical writer on board and a couple weeks that were in the very, very last stages. So for those who gave up hope you actually here.

I

Nobody gave up hope this.

A

Is always help degree.

C

It would be great to have happen. You writer fold it into this conversation, so your sponsors I can get integrated into this associations early on. So when it comes to work on the content, the foundation set is either writing the content of the letter. That's.

H

That would be really helpful.

A

And it's definitely looking forward to a redesign of that website. So yeah sounds like a like a good project, so anything else on.

A

Let's see, our next point is another.

C

A

Talk about that? Yes, the other.

J

Yes, so I'm proposing some changes to the way that I am finding and choosing colonel specs what what Colonel's works. So the two key things are a more easily extensible system for finding Colonels. So we can have our colonel spec based colonel finding machinery, and you know, for example, a continuum wants to add I, think that finds Colonels based on Kandra environments that have by PI Colonel installed.

J

Then they can add that without it sort of replacing our machinery- and so you have a list of different colonel finder objects and it queries all of them to see what kernels are available and then there's a namespacing system for the kernels that it can find and the that that's the bit, that sort of at least partially implemented and in a PR the other bit Dutch.

J

That I want to kind of build on that with is a system for picking the kernel to start when you open a notebook that doesn't hinge on putting the kernel name in the notebook match data, because often if that name is tied to an environment and that environment only exists on your own computer, then that name is it's not really something that belongs in the document that you ark, that you're sharing with other people and see I've links to a discussion issue and the PR for the first part.

J

The kernel discovery part in the in meeting notes. If people want to have a look at those.

A

Okay, awesome before I move on I want to flag for you, Jamie and Brian that my blue jeans is frozen. The U is frozen, so I may want to log in with your moderator code in case I need you guys to enter the recording when we get to that point. I'm.

C

In is a project Jupiter, so shouldn't all right, you're seamless, so.

A

Okay, awesome we'll just continue that and I can't interact with the UI, though so I'm not seeing the chat or anything else that's going on. So, if you can help me all thought, that'd be awesome. Okay, let.

C

Me just make a note, then, in a chat, I dropped a promo code for Jupiter Khan that Deb can share with their colleagues and then also the link to the PyCon videos. Jake stander classes. Talk was great. On the first day, k. T-Half talk was great as well lots of mentioned the notebooks and Jupiter in general, so they're worth a watch. Awesome.

A

And the promo code gets people what? What is that it.

C

Says, though, the promo code is jib, core 20, jup, Co re 20, and it gives people 20.

C

A

Awesome and then do you have any updates on governance that you want to talk about.

C

That's usually Brian Brian, any updates on the governance side, but.

H

A

Okay, so it looks like in for conferences and outreach, there's a link to the project calendar the team-wide project calendar within that section, so just a reminder to continue adding your conferences to that calendar, as you confirm your participation, especially if you're speaking or just to kind of draw people in the community to the events that that you're participating in please continue to add your events.

A

There I already talked about Jupiter Khan and the team meeting, no new updates there other than reminder to everyone in the community that the discount deadline is this Thursday for the first tier of pricing. Of course, people can continue registering after that, if they haven't decided yet, but that initial discount deadline is coming up. Pretty quick.

A

Jamie, there's that section that says new but I don't know is that should I take that out of there? No.

C

Those are new events that are coming at o new events got it.

A

We're good and the list of already announced events I'm not going to go through that that's listed out within the agenda and you guys normally talk through the action items for the week or people supposed to just add those in there. I.

C

J

Agree with that yeah.

A

Okay, so actually you're kind of both we're talking at the same time. So if there's nothing listed there, you don't talk through them. You just kind of move on.

H

Yeah yeah, so what one question I missed the first part of the meeting. We have a Google document for everyone to record agenda topics for the upcoming tea meeting the person team meeting. We talked about that Google Documents.

A

Yes, yeah I kind of I reminded folks about that, and let them know that you and Fernando had started taking a look at that. So yeah.

H

And I think the only thing I would add is obviously oh there's a lot of topics that we end up talking. Nothing in person which were ad-hoc, I, think the things that are most important record. There would be things that we all want to talk about together, mainly because we we only have so much time to discuss big project spanning topics. One of the cure, deprioritize forms were actually interested in and feel like urgent.

H

So if you have ideas on projects spanning times, please let us know in some capacity either on that document, just email for another night or g-meter like for being able to okay.

A

Awesome so the last section looks like releases for last week for this week to be really soon it's kind of quiet in there. So I don't know if that's because the notes haven't been updated or if there isn't much happening with that this week. So any updates from anyone on that.

H

Chain is there an iPod, widgets relief? You want to.

G

B

G

Too and I'm not going to promise beta this week back, but I sure hope that we can get very close to it before the definitey.

H

Put under soon hi widget show yeah yeah.

G

Yeah since well, theta.

H

Okay, got it okay and.

C

Which Jupiter lab releases beta.

H

But the way that we're thinking about this is where we start to make a grid all of the packages inside Jupiter lot and to basically identify three different dimensions. One is a Fiat stability, the other is usability. The other is feature parity and we are you beginning to develop that grid. We want to use that to communicate to the community from where things are at our our beta and our one dotted line are sort of releases are not going to be sort of a particular set of things on that on that grid.

H

So, for example, we're not going to have 100% feature parity or even 100%. Api stability with amazing we're trying to shift towards is communicated clearly where we are at today.

H

I think the main thing that we're aiming for with the beta is that most of them boxes in that grid under the usability call are, are green, so to speak, and so we're we're working right now. I'm building that grid out and adding all the information to it. I don't know we're doing fairly, often releases right now and so I don't know.

H

If that's going to be a you know, 29 point 32, it's hard to say from that perspective, where we're trying to do these regularly and shift our attention to communicating where access is crystal and.

G

Our target for bida is essentially, people could use it day to day little friction. Okay,.

F

Thank you. Okay,.

A

So anything else on releases or anything else for that matter, kind of last chance catch-all looks like we're at the end of the agenda, thanks to all of you for helping me through it anything else before we go ahead and close, the closeout.

A

G

A

Amy, we hope to see you soon: music, okay,.

C

Thanks everyone.