Jupyter Community, 30 Nov 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Jupyter Community Call - November 30, 2021

Description

Recording from the Jupyter Community Call in November 2021.

The notes from this call can be found here: https://jupyter.readthedocs.io/en/latest/community/community-call-notes/2021-november.html
Read more about these calls on Discourse: https://discourse.jupyter.org/t/jupyter-community-calls/668

A

Yes, recording in progress, hello, everyone and welcome to the november 2021 jupiter community call. This will be our last one for the year and we have officially had a year of returned community calls. So that is awesome. I'm excited at least it's been great to spend the last year with y'all and see what you're working on sharing. What things you remember from the past, we've had a lot of good looks back on that, but there's a few things I need to say before we get started. First of all, this is a jupiter community event.

A

um We are held to the jupiter code of conduct. Here you can find that anytime at jupiter.org conduct. We also are recording this meeting. You already heard that, but just so anyone knows listening to this in the future that these calls are always recorded.

A

um Do with that, what you may, I believe we haven't had anyone add since so I won't link the agenda notes again, but I'll be keeping an eye on that. If you haven't been here before welcome, you are always welcome new or returning or haven't been here in a while, because I've been super busy and yay a quick rundown of what we're going to be doing in case you haven't been here before we have an agenda notes.

A

Document people have signed up ahead of time, we'll just be following that, in order of items I'll be trying to watch the time to make sure that we get through to everyone, so people share what they have and we'll discuss it's kind of just a big show and tell but the first things first actually is. We have short reports celebrations and shout outs, so these are quick things that don't require discussion, but people want to celebrate or let people know we do have one of these. Oh luik. Are you ready to celebrate?

A

This looks like a celebration.

B

Yeah, it is yeah, I mean we have this extension just lab h5 web to open hdf5 file in jupyter lab. So I can show you just a small glimpse of it.

A

Awesome we're seeing the agenda.

B

A

It's just checking.

B

Okay and so basically hdfi, if you don't know it's like a file system in a in a single file, so you have groups that act as well as folders and you store your data in data sets that you can see as the files in your file system.

B

So you can imagine that this is when you have a big file like this, it's kind of difficult to explore and know exactly what kind of data you have inside. So that's why we have this jupiter h5 web extension, where basically, you can open a file and then have access to the data sets that are stored in the file and then um so this one. You can see this metric visualization, so you can see the data as a metric, but there are more interesting ones.

B

Perhaps let's see this one yeah where we have, we can have some complex visualizations and some line visualization hit my visualization, basically all kind of visualizations that you would expect for n-dimensional data sets and this extension it has been in development for quite a while now and we had a version for shipito lab2, and lately um we have a version. 1.0 that brings support for jupiter lab 3 as a private extension.

B

So that means that you only need to install it with pip pip instructor h5 web to have it in your jupiter lab, which is super handy, and we have still a lot of features that are added into this. I'm not going to go into the detail of all the features that we had is still in active development. We are still supporting also the jupiter 2 extension, because this is uh this is what we have here at esrf and we want to continue supporting the jupiter 2 extension and yeah. Well, that's it!

B

That's all that I had to say about it and if you are interested just so, you can contact me get in touch or just try. The extension tell us how it went. We are quite interested to have feedback from outside esrf as well, so yeah. That's all.

A

Fantastic congrats and that's been really awesome. Seeing updates on that every month, a couple of months, yeah.

B

I must say that uh pre-built extension is really really really great, and I want to congratulate all the people that are involved in this, because now that I've moved to the private extension, it's really hard to go back. I mean it's really great to have just keep installing that it just works. You know. So it's really awesome.

A

I think we've got some people who worked on that on this call, but regardless I can bring that over to jupiter lab meetings too. I think they'll be very happy to hear that. So thank you. um Next on the list we have nick hi.

C

Guys, how's everybody uh just quick note: we uh we busted through the last kind of some of the last two pieces that uh kept jupiter light from being fully self uh self-contained, so you can put custom wheels in. You can uh put your math jacks in there, so there are no assets that it draws from external cdns.

C

uh This now makes it archival grade um so you'd be able to you know, put your dissertation jupiter lab environment that runs in the browser on a cd and it'll open up as long as that, cd is readable media and you can start a web browser. So um there's one last piece: that's not automated and that's getting 178 megabytes of pi died and maybe there's some way we can figure out to not need. All of that like do. You always need psychic learn.

C

I don't know, um but with that you know, it'll be that'll, be a thing like that will be a real actual feature, a real system that is not going to fall over tomorrow, because the internet hates you, and so um you know we thank all the folks that have used it along the way. uh That's about it.

A

That's about it! That's fantastic, nick congrats.

A

I don't know- maybe I'm the only one really excited about that. That's amazing.

C

um Well, so, to put it in context, there was going to be a demo and I couldn't get the demo work because uh we didn't get the jupyter server proxy release out. Oh yeah jupiter server proxy 3.2.0 dropped. Yesterday.

C

um You can now rewrite the requests from your upstream web servers to make them do whatever you want, which is hilarious um so uh but uh yeah. The idea is, you know now that you can make an archival grade: offline reproducible, interactive computing environment, with your data sets and all that stuff you can put it onto these newfangled distributed web technologies so that you can. You know, post that stuff in the swarm. uh It doesn't quite network on binder and we're trying to figure out how to do that right.

C

You can pull stuff down from, say ipfs on binder, which can't host stuff from ipfs, so um getting getting very close to something. I don't know. We've been trying to do this for like five years now, and it almost works.

A

Yeah that, oh man, that is amazing, um okay, yeah, I don't I don't know what I'm so gobsmacked. I suppose I don't know what else to say, but I'm excited to see more work with that.

A

That is the last uh short item. We have on the agenda right now, but if anyone has been inspired by two really really really cool celebrations, uh I will have about 10 seconds of silence for you to decide, you want to speak up so here we go.

D

A

Thank you, darian, that was fantastic. Okay, then we're moving on to the main agenda items. Oh.

E

That was the least words I've ever heard. Darion say it once.

A

E

Usually, I need a dictionary to listen to talk to darian, but that was good. So nice only.

A

To darien, I feel like I could say that for a lot of people here I always learn a lot from everyone in the community. So yeah, that's impressive, only darian is dictionary worthy.

A

um We have several agenda items which is awesome. Some of these have popped up recently, so I would love to get that started. Make sure we have time for each um the first one we have is lilac tyrell. I think. Since apparently we had two on the call and I didn't make that clear. Are you ready to share.

A

I can't hear you I can see your screen share, though just so you know, can other people confirm? Is it just me yeah? I can't.

D

Hear it's just like kind of staticky.

A

And we can always switch if you want time to um to clear that up, we can start with another share and get back to you. If that's okay,.

A

I'm hearing some staticky stuff now.

A

Okay, if it's okay with you, don't panic, I would like to give you time to work on that and we can switch to start with other. Can you confirm a chat if you're? Okay, with that.

A

Thank you, um and let me know if I can do anything to help you then next on the agenda we have, I don't actually know this username at all.

A

Do we have? Is it I think it was jeremy? But jeremy was here on the next on the agenda. Is that correct.

A

I don't know if I'm seeing them here. Okay, so maybe we're moving another one down. Wonderful! Look at this. I was looking at the long agenda and getting worried. I I think frederick you're, the next one on the list. If that or.

F

A

Get your sound back.

F

Is it okay now.

F

Can you hear me, oh.

A

Yes, I can hear you yes, wonderful.

F

Okay, so, like I have a, I have like a button here and it's always pushed, and I I forgot about it. Sorry so sorry, sorry fast, so yes, I'm luik tetrel, so I was working on on this project called neural library for quite a while now sonori neuro libre. Basically it's it's a new preprint service for the neuroscience community, where we want to give access to the user, interactive environment. So, for example, you can see like basically something like that.

F

So you have a submission, a paper and you want to have access to like to some html content and you also provide like a binder environment.

F

So for this presentation I just focus on one specific submission here: you have access to also to the dock. So if you want to check it exactly how what are the different steps to submit, but we are still in alpha stage but but if you want to look at it, just just go ahead. Basically, when someone submits a paper, there will be a github issue that will be created for him using our robonaur. It's a bot, a robot.

F

Basically, there is a reviewer who interact with this robot and this will make build.

F

Build submission on our test cluster, so this is where, like you, can see, like many people interacted with interacting with this little robot like trying to compile notebooks, and this is what you get at the end, uh so this robot interacts with a backend and api stay developed. So basically, here we have like a build command to our test cluster. So this this api is really internal for us. This is how we interact with the binderhub cluster and we have also access to all the submission or the editor books here.

F

So if we look to one specific author, which is the one I I showed you, and so we have at the end this and when you click on this little link, this will launch this binder instance.

F

So now, we'll take a little time to explain you how I did that so I'm in charge of all the back-end generally like the maintenance of the kubernetes cluster, the virtualization layer, so the api and all the binary customization stuff.

F

So in the configuration file before what I did was to inject some code, I think it was in the jupiter hub somewhere at this place, and this is where I I was building the jupiter book for the user using the user image user environment.

F

But now I found actually a better way to do that using jupiterhub extra files. So basically we have some batch scripts that can be spawned just before the user environment is spawned.

F

If we want to look quickly at this little script, I think I have some here.

F

So this is the egyptian book build process uh like I will not go into these scripts, but basically, what the most important part is is uh like this one here, jupiter book build. So this is uh this process is called every time, a user spawn an environment, and we can also look a little like how the cluster looks like yeah. So here you can see like we have one binder instance that is running it's this one that I spawned before and here this is where you can interact with the code.

F

And we have also another process, so it's not like part of this presentation. It's like getting this data because we also provide the data to to the user, so they have access basically to a mounted folder when they have all data set available to for them to run uh anything. So we could have like one terabyte of data and the notebooks could could use it.

F

So we have this mechanism report to data to get the data into our cluster into our backend, so user can use it after so, yes, I was quite happy to find this extra files things because it was not easy to to find that but yeah. Now it's it's much better. We still have issues because, basically, when we build the jupiter book, we save it into our back end and for that we need to have right access to our backend in the user environment.

F

So that's why? That's not sorry, that's why that's not ideal, and ideally I would like to have init containers instead, but that's not possible now, so I have a.

F

I think I posted in the uh discourse in the discourse. The egyptian hub deals course. So I posted here all the rationales and what we would need to have like a minute containers running.

F

So yes, that's that's it I'm thinking there is something else specific to this workflow.

F

So I was assuming that you know what a jupiter book is so just really quickly. Basically, it's a way to convert notebooks into html until to give a really fancy an interface really suited for publications. Basically, as we can see here, we have like the chapters we can have, also like reference, etc.

F

Okay, that's it so. Do you have any questions, maybe or suggestions.

A

I love seeing demos, I don't know well, no, I do. I don't always have as much familiarity with the fields that everyone's working in, but that that's awesome to see, and I have a little bit of background on the discourse issue. I was actually looking to link it for people. I think you're talking.

C

A

This one maybe, um but I I'm really enjoying seeing you put all these pieces of technology together to all these separate things and kind of making a unit with a process for submitting to that's really cool.

F

Yeah, it's this one. Yes,.

E

So who's this who's, the primarily the consumer of this in mind. Is it the author or is it the readers? Are you guys? Are you all trying to give more trust to the reader? I'm sure it's a little bit of both, but who does this primarily benefit readers, reviewers.

F

uh So definitely like there will be more readers than authors like it's internet is working like that. Now there are more people that read things that instead of creating things, so we expect to have more readers in terms of like. I think it benefits a lot authors because it allows them to so it gives to it, gives them like the trust of okay.

F

I have a publications, I have publication and I'm confident enough in it to share with you my code, my reproducible code, so you can test it and see that what I am showing is not is actually good. You know, so it's a way for the author to to gain trust.

F

So I think it benefits a lot of authors uh for the readers. Also, it is it's beneficial because it's it can uh like in one click, they can have access to all the noisy work of the author, so they can have the figures interactivity with the jupiter book and also the binder instance for developers who want to check and maybe change some parameters, some settings and re-execute the environment, but yeah it's complicated to say which, uh which one does it benefit? Most but uh yeah it's my thought, thumbs up.

E

Thank you. Can I ask one more question about the author's experience with this? Yes, um a lot of these systems. I feel, like you, upload a paper when you're done. Is this system similar or is it there to provide like intermediate support in some way to the author, as they are doing their writing and building.

F

Yeah good questions, actually it's yeah I mentioned in the beginning: it's like a print service, but it's actually a little more because we have all these technicalities. You know to make these things work so so when, when we have this issue, we interact a lot with the author and the- and this is where, like we'll say to the author, okay, maybe clarify some like there is a little technical uh review, but there are a lot of back and forth with the author and reviewer uh and also like some technical things like if the.

A

F

Is not working on our brain, so it's not as we want to to make it as more simple as possible, but still it's more complicated than just a pdf, because we have all the binder environment or the book build. So it's it's crazy, like the amount of of work but yeah, we were in a a way to make things really simpler, but yeah user from the author will expect to be more.

F

We need to be more involved than just a standard print pdf.

B

I have a question: how do you see the interaction of your service with other print services and the peer review publication? Do you want to link them together, like if you have a publication that shows some result saying that in this publication there is a link to run the notebook or things like this.

F

So um so I'm not like I'm really like the usually I'm this guy who developed things in the in the dark. Let's see so I don't have like the big overview of this project, but um yeah we. So we are in contact with uh journal of open science. This is where we actually got our front end running so the front end is basically a big fork of jaws.

F

We are also in in communication with some publishing uh in neuroscience. We have this uh the hbm organization, which is like a big community gathering on like in the yeah like a community thing, and they want to create a new journal, and this is where we will also help them to have the infrastructure.

F

So I think the idea is mostly to give to publishing platforms to give them some help to to have their platform ready basically, but the branding will still still be on them, but now we focus on normal library, but yeah later it's the plan is to to share and to to actually uh give this opportunity to other platform as well.

F

So yeah I mentioned aperture just may be interested in that also in the future, and there are others, but I don't remember.

B

Okay, let's see thanks.

A

Great, I'm gonna be quiet for a few more seconds. Let people have room for any more questions. That's your warning for income.

A

Okay, oh go ahead, though,.

F

Yeah just last thing, so uh I think you you you, you send like the link to the isabella. You see you send the discourse post. So if there are some uh binaural developer whatever that could locate at it, it would be really nice because now it's basically like the workflow, I show you is like a big security bridge because we give access to our backend storage. Basically, we don't have choice uh like the init container solution is really important for us.

F

I think we had the pr for like quite a while now, but if things can move on that side, it would be amazing. Also.

A

Do you have a link to that pr and you.

F

Can post it whenever.

A

I just think it would be okay.

F

I will find it and post it.

A

Okay, thanks and also are you able to make it to any of the jupiter hub meetings? In that case, I know sometimes they um review prs there. We have some hub people, so let me know if I'm overstuffing that but yeah that may also be a good place to get their their thoughts on this.

F

Okay, okay, I think I was I did like last time. It was a jupiter hub call, actually yeah. So.

A

Yeah awesome. Thank you so much for a great demo and great fielding of questions. That was wonderful thanks.

F

A

Next on the list we have jupiter, nas, which I was looking at the the discourse post, because I remember this one showing up on jupiter discourse, and I I don't see the person who originally spoke about it was anyone here planning to present on this work say it once more: jupiter nas, I think, and aas yeah going once.

G

So I can try to pick them if they would like to join. Maybe in a few minutes, I don't if there is anyone else in the schedule. I haven't checked the schedule.

A

Yeah we do have some others in the schedule, so we can loop back. I just wanted to okay. Thank you, uh frederick. You are the next person. Are you ready.

H

Yeah, hey guys, so um that's like going ahead a bit with with the point of ashlin, so we've for the next major version of jupiter notebook.

H

There is a some challenge of porting, uh some extension to jupiter lab infrastructure, and one of them is rice so for people that don't know about that extension, it transformed a notebook in a slideshow.

H

So I started the pr to do that and I will demo the current status. So it's still a work in progress and the idea of showing it today is also for people. I think the pr there is a binder link directly in it. So if you want to try or try it online, please do so and provide feedback because yeah, I know to develop the extension, but I'm not a strong user of the extension myself. So so that would be cool if people that are more like common user, can give a try and provide feedback.

H

So the idea is to render the notebook in another format. So it's using a javascript library, that's called reveal, but whatever, and so it it creates a slideshow. A slideshow is defined by metadata on every cells. So the metadata can be shown here, so this is the slide type, so some can be slide.

H

I think there is some that are sub, slides.

H

Forget which one okay, some are fragments, so it's it will uh impact the flow of the the slideshow. So if I go through like here, you can see that stuff appears like like you expect, for a slideshow. um There are one of their power of rice. Is that it's not a static view of the notebook, so you can edit the cell.

H

H

I'm I'm hit by a shortcut, that's a black. I cannot type t it's a shortcut so and you can run you can run a cell if you want it supports of uh also then widget. So this is just a matpot clip one, and if I want this one, it will uh take the matplotlib put it in uh interact stuff, and so you will be able to interact with me like this, so you will be able to interact.

H

So it's really like the classical notebook you expect in terms of features, and it's just presented presenting it as a as a slideshow.

H

We have already ported to some features like being able to have a whiteboard being able to annotate things. So that starts to be useful. Let's say the shortcut I was mentioning. You cannot see it, but actually is popping up another window in which you can have the notes for the presenter and okay. Of course, there are some some stuff that need to be improved like this one. So one of the trouble for now is that the notebook is actually open twice and there is no synchronization except on with the file.

H

So if you change on the notebook on one side, you you may end up with that kind of warning. So that's one of the points we are working on, but yeah mainly it's really for people to to see what it look like. If you don't know the extension and for people that already know them, please please go to the pr and launch binder and try it yourself and provide feedback of what school was not so cool and stuff like that, and, of course what I haven't demonstrated, but that I don't think sharing will work.

H

In that case, you can uh transform it into a full screen presentation and that's it for me.

A

That that's it's really tricky speaking, that.

E

A

Yeah, it's tricky to develop something. Oh is my audio being weird: oh, okay! Well, it's tricky to work on something that you don't you aren't as familiar with so collapsed. You because I've heard a lot of demand for this and yeah. That's awesome.

A

You've got some comments in the chat about.

E

The collaborative flag- I do have one one suggestion frederick, when we give when you give that demo, I think you give it in full screen mode. Then it looks more like the classic notebook like I've. Never I haven't seen rise like in lab before which was kind of a cool experience, but yeah it's a really great demo. Now to the collaborative part.

H

Yeah thanks yeah nick the the collaborative true can at least saw some trouble of the synchronization on the file, and we are just desperately waiting for having the model on the back end. It will solve everything.

C

So you had mentioned is this: this is a fully custom application, all the way down to the bottom of the dom, or is it built on top of retro or how's that getting there.

H

Yeah for for now, it's like a simplified retro app to just bring the notebook editors in which we are messing up for for having reveal happy.

C

Yeah you can't fight reveal it we'll ruin your day.

H

Yeah, that's the trouble thanks.

A

Great any last discussion on that places. We should share the pr so that more people um who use ryze get to try it out thoughts.

B

Every question, in fact: oh easy! Oh, oh, oh hard! Is it to install it locally to test it.

H

Should be pretty straightforward, like uh I can update the note, it's like classical thing, you pip install editable mode, and then you have to install the the extension. That's like the structure is a bit weird, but yeah. Normally everything runs quite easily because I've already have some challenge helping uh damien for installing it. So it should work. I will copy the the how to in the notes.

B

Cool thanks. I'm excited to try it out, because we are giving some trainings here that and we are used to using the notebook interface because we use ryze due to the interactivity and the fact that it's yeah, it's a slideshow. So it's easier to present and I would be excited to try it out in jupiter, live instead.

A

I'm also seeing some comments for you actually in the notes document saying that this is. This is very cool, probably.

A

Okay, any any more last thoughts.

A

Play some kind of uh elevator music: okay, um awesome! I is anyone here now ready to present on jupiter nas. I don't think I've seen the number of participants change, but I want to give another check.

A

Okay, then oh savon did you have something to say. Oh.

G

I was just gonna say they did not respond to anything. So, okay,.

A

Well, thank you so much for reaching out to them. We we will have other community calls so not the end, but awesome then the next person we have is actually darian. If I'm remembering yeah correctly, are you ready to share sure yeah.

I

um So I I think this probably doesn't come as a surprise, because there's been so many public conversations, but I wanted to bring it up here as well. There is linked in the agenda a draft jupiter enhancement proposal that we are pretty close to undrafting, I think um within 24 hours.

I

At this point, um it's gone through a lot of iterations from conversations that came out of the big thread in the notebook repo to things on discourse to things in the jupiter lab call to things in the jupiter notebook calls and to a few governance calls, and it is a pitch for notebook version, 7 effectively being what we today call retro lab, because um the user experience of the notebook is a thing that people have come to really value and love, and we want to be able to offer that to them with a secure extensible code base that is compatible with jupyter lab compatible with security needs and compatible with the expertise that we have and the large developer community and code base that we have built in the past few years with lab.

I

And so, um if you are curious about the specifics of this feel free to have a glance at the draft, if you have some comments, you'd like to make, please do that and um pretty soon I think we will be opening that up to the vote of the jet process. This will probably be one of the final jupiter enhancement proposals that will be voted on by the original jupiter steering council, because the software steering council, which will be the new body that adjudicates jeps, has been passed into governance.

I

It just doesn't yet exist because it takes a little while to ramp that up so um yeah, that's pretty much. It.

A

Thanks for for the continued updates on this to people with questions, people background on what's happening here,.

C

And just as a point of point of information there, you know jeps are big and scary and all that stuff, and they are because there are a lot of people. There are a lot of eyes and a lot of people that are like making their living off of the stuff.

C

That's going into that if you have opinions about the notebook experience and what is important to be kept, put those opinions in there right, um because it's just a lot of us like complaining about how hard this is gonna, be to do it long term, which is true right. But um you know that's not the only perspective that matters.

I

Yeah, that's a good point nick and I think it really is. This is our best faith effort to respond to what it is that we've heard from jupiter users about what they want when they use jupiter when they author notebooks and when they interact with this ecosystem that has grown up in the last. You know, 15 years and.

I

There's a huge community whose voices will likely never get to here because that's just how a large community works, but um thanks for inviting people to communicate more, that's a good point.

A

Is there any corresponding place just uh I personally like to keep things all together, but I know not everyone whose opinions may have a github account. I know we have a lot of people with that, but is there any kind of corresponding place for feedback right now, or is this the only place just so? We know.

I

This enhancement proposal has been linked also on this course, um and I guess the only other official channels are public meetings. So it's discourse public meetings and the actual jeb repo.

I

There are only so many places where we'll actually be able to consider. What's being said, so I'm sure there's a lot more conversation we may have missed, but those are the ones we're actively monitoring and trying to account for.

A

That sounds good. I'm gonna look for the jupiter discourse link, then too. Just so. People have that thanks.

A

This is your seconds of silence to think of more things to add everyone.

A

Okay, then yeah thanks again darian for the continued updates on some of the really high level stuff happening in the community.

D

Tony, were you saying something or.

A

Okay, um we actually still have a few minutes which surprised me, because I thought we had a really full agenda, but sometimes that's how it goes, and I get surprised um with all those awesome shares. We do have a little time if someone else has kind of a spontaneous demo.

C

A

Other kind of short announcements: we can loot back to that for the next few minutes, exciting things.

I

I

It's not a jupiter pie, though it's just a pie.

A

I I mean, I think, that's worth celebration.

I

Here I'll link to it.

A

Food.Darien.Link, you have your own place for posting food. That is something else. Wow wonderful.

A

Anyone else have top secret recipe blogs, apparently that they'd like to share.

A

Okay, I will give a few more seconds of science. Oh yeah,.

C

Has anyone ever seen the cooking for engineers site?

C

It has probably the best uh the best recipe visualization I've ever seen. It's like a big grid that has merged cells, and so it's like mix this and sift this and then put them together in a bowl and sort of got a visual representation for it. It is it's brilliant.

A

That's cool: let's check it out.

A

Okay and no tony, it sound like you're going to say something again: I'm not trying to cut you off or.

E

It's all right. I just wanted to do like my recent announcement about shift r in the notebook and jupiter lab with the side-by-side cell view. I was actually wondering if that was going to work in a rise now that I mention it, but there's a new feature where you can, if you're outside of cell editing mode and you hit shift r, you get side by side, input and output. I really like it um and I figured some other folks might so I'd. Let you all figure that's worth sharing.

A

Do you have a quick demo, you can say no.

A

That sounds like a no, though I think I did find the issue for y'all and we do have someone on this call. I think that worked on this here y'all go.

A

Great well, I am not here to make you all uh sit in silence, so I am happy to wrap it up. Then, unless we have any final thoughts, I that was a definitely a call packed full of really cool demos, though so. Thank you all so much for a great november call.

A

I will also be linking you can always give feedback on this call. If I can get the link to you successfully, which is always a thing yeah sorry I thought we'd have another moment. You can always give feedback on the call here and, like I said at the beginning, we will not be doing the december call, because we'd have to do it early in december and you all just great great shares now.

A

So I thought it might be a good time to give everyone a rest, as we have various winter holidays, but we will be starting up again in january, so get excited for that rest well and get ready for the new year y'all. Thank you so much for your time and a year of great calls see ya.

E

A

Hi everybody. Thank you.