WEBVTT 00:00:00.140 --> 00:00:04.200 Are you using interactive notebooks for your data exploration or your day-to-day programming? 00:00:04.200 --> 00:00:06.080 What environments do you use? 00:00:06.080 --> 00:00:08.440 Was it Jupyter and now you've made your way over to JupyterLab? 00:00:08.440 --> 00:00:13.540 That's a great choice, but did you know there are more environments out there to choose from and to compare? 00:00:13.540 --> 00:00:15.880 Have you heard of Callisto or Iodide? 00:00:15.880 --> 00:00:17.780 How about CoCalc or PolyNote? 00:00:17.780 --> 00:00:19.940 Those are just the tip of the iceberg. 00:00:19.940 --> 00:00:24.260 That's why I'm happy to have Sam Lau and Philip Guo here to share their research, 00:00:24.260 --> 00:00:27.380 comparing and categorizing over 60 notebook environments. 00:00:27.740 --> 00:00:32.400 This is Talk Python To Me, episode 268, recorded June 11th, 2020. 00:00:32.400 --> 00:00:52.020 Welcome to Talk Python To Me, a weekly podcast on Python, the language, the libraries, the ecosystem, and the personalities. 00:00:52.020 --> 00:00:53.960 This is your host, Michael Kennedy. 00:00:53.960 --> 00:00:56.080 Follow me on Twitter where I'm @mkennedy. 00:00:56.280 --> 00:01:02.280 Keep up with the show and listen to past episodes at talkpython.fm and follow the show on Twitter via at Talk Python. 00:01:02.280 --> 00:01:05.520 This episode is brought to you by Linode and Sentry. 00:01:05.520 --> 00:01:07.680 Please check out what they're offering during their segments. 00:01:07.680 --> 00:01:09.320 It really helps support the show. 00:01:09.320 --> 00:01:12.080 Sam and Philip, welcome to Talk Python To Me. 00:01:12.080 --> 00:01:12.960 Pleasure to be here. 00:01:12.960 --> 00:01:13.600 Same here. 00:01:13.600 --> 00:01:15.360 Yeah, nice to have you on the show, Sam. 00:01:15.360 --> 00:01:16.640 Philip, welcome back. 00:01:16.640 --> 00:01:17.980 You've been on the show a few times. 00:01:18.200 --> 00:01:23.880 You were one of the very first guests way back when, diving into the CPython source code internals. 00:01:23.880 --> 00:01:32.420 And then we've talked about a bunch of fun topics, you know, coding in your golden years and maintaining open source projects individually and things like that. 00:01:32.420 --> 00:01:34.660 It's been a lot of fun to have you on the show and nice to have you back. 00:01:34.660 --> 00:01:35.700 Yeah, it's great to be back. 00:01:35.800 --> 00:01:38.360 And yeah, that was my first ever podcast with you. 00:01:38.360 --> 00:01:40.480 That was, I think, almost five years ago. 00:01:40.480 --> 00:01:43.220 It's good to be back as a podcast veteran now. 00:01:43.220 --> 00:01:44.040 That's right. 00:01:44.040 --> 00:01:44.820 You know the whole deal. 00:01:44.820 --> 00:01:46.640 And Sam, welcome to the show. 00:01:46.640 --> 00:01:52.880 Before we get to the main topic, I always like to set the stage for where the guests are coming from. 00:01:52.960 --> 00:02:02.920 So when somebody says, you know, like, I think you should really do this, like, they might say that because they're a biologist using Python, not a person running Facebook or the other way around, right? 00:02:02.920 --> 00:02:06.620 So maybe just tell us really quickly about your background and how you got into Python. 00:02:06.620 --> 00:02:15.840 My background in Python came from, actually, back in high school, I ran into the first wave of, like, massive open online courses. 00:02:15.840 --> 00:02:17.400 Right. 00:02:17.400 --> 00:02:18.440 Right. 00:02:18.440 --> 00:02:22.580 And so, like, at my school, there wasn't much emphasis on programming much at all. 00:02:23.100 --> 00:02:36.900 And so when I ran into, when I found out about those courses, I got really into it and decided to major in computer science at UC Berkeley and then ended up doing, ended up starting this PhD at UCSD with Philip studying cognitive science. 00:02:36.900 --> 00:02:38.300 Cognitive science is really interesting. 00:02:38.300 --> 00:02:40.520 I've talked to a few folks from there. 00:02:40.520 --> 00:02:46.120 There's a lot of interesting computational CS type problems happening in cognitive science. 00:02:46.120 --> 00:02:46.380 Yeah. 00:02:46.380 --> 00:02:47.420 And data science as well. 00:02:47.420 --> 00:02:47.840 Absolutely. 00:02:48.280 --> 00:02:55.960 I worked for a while at a company that was basically a spinoff from a research lab that was a bunch of cognitive science PhDs. 00:02:55.960 --> 00:02:58.440 And we did all sorts of cool stuff with eye tracking and programming. 00:02:58.440 --> 00:02:59.940 And, man, it's just a lot of fun. 00:02:59.940 --> 00:03:01.380 I really enjoyed my time in that space. 00:03:01.380 --> 00:03:01.660 Yeah. 00:03:01.660 --> 00:03:07.880 People here are, cognitive science is so broad that you have people here using programming for all sorts of tasks. 00:03:07.880 --> 00:03:18.960 Like, we have people here training deep neural networks, but also people studying fMRI signals and processing those and working with neuroimaging. 00:03:18.960 --> 00:03:23.580 And so you just get a really diverse, broad range of perspectives and uses for Python. 00:03:23.580 --> 00:03:23.920 Yeah. 00:03:23.920 --> 00:03:26.820 It's interesting, too, because I feel like there's still a lot of open questions. 00:03:26.820 --> 00:03:27.140 Yeah. 00:03:27.140 --> 00:03:28.200 There's a lot we don't know. 00:03:28.260 --> 00:03:31.720 And it feels like the more we know, the more we know about what we don't know. 00:03:31.720 --> 00:03:33.620 Yeah, exactly. 00:03:33.620 --> 00:03:36.340 Every answer opens up to more questions. 00:03:36.340 --> 00:03:36.720 That's right. 00:03:36.720 --> 00:03:37.140 Exactly. 00:03:37.140 --> 00:03:37.600 All right. 00:03:37.600 --> 00:03:48.680 Well, we're going to talk today about the research that two of you have done in the space of notebooks, as in computational notebooks, Jupyter notebooks, that type of thing. 00:03:48.680 --> 00:03:53.240 And it turns out, actually, there's more than just IPython and Jupyter, right? 00:03:53.240 --> 00:03:54.560 Quite a few more. 00:03:54.560 --> 00:03:55.160 Yeah. 00:03:55.160 --> 00:04:03.560 So in our paper, we talk about a total of 60 different notebook systems that we put together and try to draw out some patterns from. 00:04:03.560 --> 00:04:04.400 Yeah, it's really interesting. 00:04:04.400 --> 00:04:12.600 I mean, people, I think Jupyter, IPython, when it was first started, became Jupyter, really changed the way people thought about programming. 00:04:12.600 --> 00:04:13.860 I mean, there were tools like this. 00:04:13.860 --> 00:04:17.560 Maybe you guys can throw out some that felt kind of like that. 00:04:17.560 --> 00:04:24.540 So when I say that, the one that comes most to mind is, I think, Mathematica or Maple, right? 00:04:24.540 --> 00:04:36.380 They're really beautifully visual, but then they also had computation, but they also had like output, you know, MATLAB, but MATLAB's not quite as pretty, for whatever reason, more utilitarian, I guess. 00:04:36.380 --> 00:04:46.380 And when you thought about programming, that was, you know, maybe if you're trying to emulate that, it would be a script, and then it would like output some stuff, or it would pop up a window, it'd save like a JPEG or something. 00:04:46.380 --> 00:04:49.700 But the notebooks really brought those two worlds together. 00:04:49.700 --> 00:04:57.380 Say, look, you can write code in a, like a pro programming language in a way that you really would use like libraries and stuff that are not just baked into these tools. 00:04:57.380 --> 00:05:00.240 But then they have this kind of interactive world. 00:05:00.240 --> 00:05:04.420 I think it's really interesting to see all these derivative concepts, right? 00:05:04.460 --> 00:05:10.040 We've got Jupyter that had its way of doing things, but there's challenges there, and there's a lot of variation. 00:05:10.040 --> 00:05:13.340 Like, were you surprised to find this many different systems around? 00:05:13.340 --> 00:05:14.460 I certainly was. 00:05:14.460 --> 00:05:21.260 I think I came into this project thinking, oh, it's probably just Jupyter plus like maybe three or four other systems. 00:05:21.920 --> 00:05:25.900 And I think the deeper we went, the more we were like, whoa. 00:05:25.900 --> 00:05:35.900 People are really taking Jupyter and changing it, adding new things to it, making it fit their use case more effectively. 00:05:36.380 --> 00:05:45.600 And there's a lot of, it turns out that there's a lot of different groups studying Jupyter and notebook systems in general and trying to figure out how they can make it better for themselves. 00:05:45.840 --> 00:05:50.860 Yeah, I think I'd just add to that, to what Sam said, was that, you know, we started this project. 00:05:50.860 --> 00:05:59.480 So this is, this paper is co-authored by Sam, one of my other PhD students, Androsos, one of my undergrad students, Julia Markle, and myself. 00:05:59.480 --> 00:06:04.440 And we basically just kind of split up the work and just went to find a lot of these notebook systems. 00:06:04.440 --> 00:06:15.900 And the motivation for this was that, you know, in the academic literature, in the kind of academic world of studying programming tools, in the past few years, there's been a lot of people studying how people use notebooks. 00:06:15.900 --> 00:06:19.240 I believe Adam Rule was on your podcast, Michael, was Adam? 00:06:19.240 --> 00:06:21.000 It sounds familiar. 00:06:21.000 --> 00:06:25.720 Yeah, he was about the Jupyter, the, you know, analyzing the million Jupyter notebooks. 00:06:25.720 --> 00:06:26.620 Yes, that's right. 00:06:26.620 --> 00:06:27.080 Yeah, that's right. 00:06:27.080 --> 00:06:27.280 Yeah. 00:06:27.280 --> 00:06:33.600 So there's a bunch of studies around Jupyter notebooks and how people use them for science and stuff, and then people developing new prototypes. 00:06:34.240 --> 00:06:40.380 And at the same time, there's a bunch of industry people, you know, every startup company is like, we're going to do a notebook, we're going to do this other notebook, this other notebook. 00:06:40.380 --> 00:06:41.880 So it became like this wild west. 00:06:41.880 --> 00:06:44.740 That's right. Google CoLab, DataLore, Azure notebooks, all those, yeah. 00:06:44.740 --> 00:06:51.660 Yeah, so it became like this wild west, and we felt it was around time for this sort of what we call a survey paper or a meta-analysis, right? 00:06:51.660 --> 00:07:03.640 It's like, this paper is not doing original research in that we're not making a new notebook, but we are collecting together dozens of these notebooks from both academic prototypes and also industry products and trying to, like, 00:07:04.040 --> 00:07:06.860 analyze them and see, you know, how does this space fit together? 00:07:06.860 --> 00:07:12.660 So I got a few of my students together, and we kind of brainstormed and looked for a lot of notebooks and stuff. 00:07:12.660 --> 00:07:17.680 And then Sam, who is the first author on this, kind of, you know, did a lot of the leadership on this work. 00:07:17.680 --> 00:07:22.760 So I'll let him talk more about the details, because I forgot all the details, so I just delegate them. 00:07:22.760 --> 00:07:22.860 Yeah, sure. 00:07:22.860 --> 00:07:26.400 Yeah, and your primary research is in human-computer interaction? 00:07:26.580 --> 00:07:27.080 Right, right. 00:07:27.080 --> 00:07:31.760 So my primary research area that my students are in is called HCI, or human-computer interaction. 00:07:31.760 --> 00:07:36.800 And the kind of industry term for this is, you know, people know it as UX design, right? 00:07:36.800 --> 00:07:39.160 User experience design, user interface design. 00:07:39.160 --> 00:07:49.740 And, you know, HCI, or human-computer interaction, is sort of the academic term for the research that goes into how do you make better user experiences for, you know, different kinds of users. 00:07:49.800 --> 00:07:52.900 In this case, it's for scientific programmers and data scientists. 00:07:52.900 --> 00:07:53.260 Yeah. 00:07:53.260 --> 00:07:54.020 Okay. 00:07:54.020 --> 00:08:05.780 Now, I think probably we should start this conversation with a definition, because I kind of opened it saying, well, Jupyter is the prototypical thing of what a notebook is. 00:08:05.780 --> 00:08:09.980 And it's kind of the natural progression of something like Mathematica. 00:08:09.980 --> 00:08:14.940 But what would you all define a notebook system to be? 00:08:14.940 --> 00:08:19.180 So, like, how do other things that are, you know, not exactly that fit into this world? 00:08:19.420 --> 00:08:32.080 In this paper, we define a computational notebook as a system that supports literate programming, where you can blend a text-based programming language with program outputs in a single document. 00:08:32.080 --> 00:08:32.340 Right. 00:08:32.340 --> 00:08:34.140 And this is not a super new idea, right? 00:08:34.140 --> 00:08:38.520 This, I mean, Donald Newt is like in the 80s or something talked about this idea, right? 00:08:38.520 --> 00:08:38.880 Yeah. 00:08:38.880 --> 00:08:45.560 And actually, I think the entire tech program, like LaTEC, the tech that Donald Newt made, was a literate program to begin with. 00:08:45.860 --> 00:08:50.140 He was like the, I would say, like most prominent proponent of literate programming. 00:08:50.140 --> 00:08:50.460 Okay. 00:08:50.460 --> 00:08:54.660 So, notebooks are these embodiments of this literate programming idea. 00:08:54.660 --> 00:09:07.340 Basically, I guess I'm hearing that it's the ability to blend together the presentation of, like, prose almost, and computation and visualization around that. 00:09:07.440 --> 00:09:08.160 Yeah, yeah. 00:09:08.160 --> 00:09:12.260 And it kind of shifted the dynamic of how we view a program. 00:09:12.260 --> 00:09:22.640 Instead of having a program be instructions for a computer, it allowed us to see programs as kind of like the outputs themselves, as a thing that you would show to other people in the end. 00:09:22.720 --> 00:09:25.720 Yeah, because traditionally, it would be, here's the graph. 00:09:25.720 --> 00:09:26.280 Exactly. 00:09:26.280 --> 00:09:28.780 We ran the analysis, and this is the output. 00:09:28.780 --> 00:09:33.340 Here's the description of the output, maybe a description of the algorithm, and here's the output. 00:09:33.340 --> 00:09:37.780 And, you know, there's been a big push in the scientific space for reproducibility. 00:09:37.780 --> 00:09:42.300 And I think also just in the data science space for explorability, right? 00:09:42.300 --> 00:09:43.200 Like, I've got data. 00:09:43.200 --> 00:09:44.300 I don't really know what it is. 00:09:44.300 --> 00:09:48.180 I need to just get in there and play with it before I turn the algorithms loose on it. 00:09:48.280 --> 00:09:50.380 And that's kind of what's happening here, right? 00:09:50.380 --> 00:09:51.160 Yeah, definitely. 00:09:51.160 --> 00:10:04.620 I would say many of the projects that we looked at in this paper were birthed out of this, directly out of, like, the reproducibility issue, where some person might write a script to generate a graph somewhere, but then lose it. 00:10:04.620 --> 00:10:06.600 And then all you have now is, like, a graph in the end. 00:10:06.600 --> 00:10:14.780 If you had a notebook, the pitches, if you had a notebook, you wouldn't need to have a separate graph, like JPEG, and then a separate program to run it. 00:10:14.780 --> 00:10:16.000 You could just have it in one place. 00:10:16.000 --> 00:10:17.120 Right, right, right. 00:10:17.540 --> 00:10:30.960 How often does version control in the science space or the sort of beginner developer space look like a bunch of zip files where they're named the thing and then the date or the thing of the date and then, like, two and then three? 00:10:30.960 --> 00:10:31.660 Oh, yeah, exactly. 00:10:31.660 --> 00:10:32.100 Something like that. 00:10:32.100 --> 00:10:35.700 It's like zero five underscore final final V3, you know? 00:10:35.700 --> 00:10:37.260 Exactly. 00:10:37.260 --> 00:10:38.200 Final edited. 00:10:38.200 --> 00:10:39.560 That's not amazing. 00:10:39.560 --> 00:10:52.640 So having something in here that sort of brings a little more formality and brings, I guess, the paper and the presentation and the script or code that ran it, like, into one thing that's versioned continuously seems a little bit better. 00:10:52.640 --> 00:11:00.380 Yeah, and actually, there are the basic way to version control a Jupyter notebook is to put it in a version control system like Git. 00:11:00.380 --> 00:11:11.020 But oftentimes, it's not good enough for actual scientists who have, like, the notebook files themselves, but also their data and also the libraries that the notebook use to generate the outputs. 00:11:11.020 --> 00:11:17.820 And so there's a group of projects that we surveyed that not only version a notebook, but also in version a notebook and the dependencies. 00:11:17.820 --> 00:11:18.360 Okay. 00:11:18.360 --> 00:11:19.440 Yeah, very nice. 00:11:19.440 --> 00:11:28.080 One of the challenges, however, I think it's ironic, like, it brings stuff together here to make the code and the analysis and the presentation all as one. 00:11:28.080 --> 00:11:30.940 But it actually, Jupyter notebooks are not very easy to version. 00:11:30.940 --> 00:11:32.240 Quite hard, actually. 00:11:32.240 --> 00:11:32.580 Yeah. 00:11:32.580 --> 00:11:40.340 Yeah, like, if they store the last run of the output, and if that is dynamic in the sense that it's, like, not every time you run it, you get the same output. 00:11:40.340 --> 00:11:43.500 That can be a, I mean, even if it changes, right? 00:11:43.500 --> 00:11:44.800 Even if it is that case, it's still hard. 00:11:44.800 --> 00:11:51.540 But if it's, like, changes every time you run it because it pulls something from the internet that varies or whatever, it's basically always a merge conflict. 00:11:51.540 --> 00:11:57.200 So there's tools like NB strip out, which are nice, that are, like, pre-commit hooks that'll let you do that and other ones. 00:11:57.300 --> 00:12:05.220 But still, it's, I think it's, these notebooks, they come both with, like, a new power and accessibility, but also their own challenges. 00:12:05.220 --> 00:12:06.000 Yeah, exactly. 00:12:06.000 --> 00:12:06.420 Yeah. 00:12:06.420 --> 00:12:12.620 So in your paper, one of the things you all talked about is you said around notebooks, they have these several themes. 00:12:12.620 --> 00:12:19.280 They have end-user programming, exploratory programming, live programming, and literate computing. 00:12:19.280 --> 00:12:23.060 So what do you mean by, like, live programming? 00:12:23.060 --> 00:12:25.580 I understand end-user programming, which is pretty cool. 00:12:25.940 --> 00:12:27.240 Observatory makes sense. 00:12:27.240 --> 00:12:28.340 But what's the live here? 00:12:28.340 --> 00:12:31.980 So, yeah, so there are a few kinds of definitions of live programming. 00:12:31.980 --> 00:12:34.920 One definition comes from kind of the creative space. 00:12:34.920 --> 00:12:42.860 So there are these people who are graphic artists or musicians who actually literally, you know, they're streaming on Twitch or they're doing these, you know, pre-COVID, I guess. 00:12:42.860 --> 00:12:52.340 They're doing these live performances on stage where they're using some kind of graphical software to, you know, Max MSP or whatever to put together live music performances or graphical art. 00:12:52.960 --> 00:12:59.080 Another form of live programming that I think what we mean in this paper is more like a programming environment that updates live, right? 00:12:59.080 --> 00:13:03.420 So as you're typing the code, it's actually just constantly running it and generating output. 00:13:03.420 --> 00:13:10.900 So some of the notebook systems we've looked at allows you to kind of do this live programming where as you start typing, it might autocomplete. 00:13:10.900 --> 00:13:16.640 It might start generating graphs for you just to kind of cut down on this edit run loop and such. 00:13:16.880 --> 00:13:17.180 Right, right. 00:13:17.180 --> 00:13:20.220 It's not this I'm going to edit when it's time to hit run, I run. 00:13:20.220 --> 00:13:25.180 Or, you know, maybe in some languages there's a compile, then a link, then a run, and there's a big delay. 00:13:25.180 --> 00:13:35.720 Would something like some of the features in the editors, like PyCharm and probably VS Code, I don't know what plugin you need for it, but where it's like continuously running unit tests just as you type? 00:13:35.720 --> 00:13:38.400 Yeah, I mean, that's a form of liveness, right? 00:13:38.400 --> 00:13:39.620 The kind of, you know, whatever. 00:13:39.620 --> 00:13:42.180 And even like, you know, even the squiggly lines, right? 00:13:42.180 --> 00:13:45.220 The IntelliSense and the, you know, linting and those things. 00:13:45.220 --> 00:13:50.200 You know, anything that kind of cut down on the, you know, the friction of having to write a bunch of code. 00:13:50.200 --> 00:13:51.680 Because especially for beginners, right? 00:13:51.740 --> 00:14:00.420 So for beginners and for people who might be scientists and people who may not be as programming experienced, you know, one thing you often see beginners do is they write a lot of code, right? 00:14:00.420 --> 00:14:02.320 They write a lot of code and they sit and they hit run. 00:14:02.320 --> 00:14:03.400 They're like, why doesn't it work, right? 00:14:03.400 --> 00:14:11.540 Whereas more experienced programmers kind of know to write a line, check it, print something out, make sure it works, and then write the next line, write the next section and such. 00:14:11.540 --> 00:14:16.700 Yeah, it's easy to do a bunch of work and then have it, especially when you're new, and have it not come out the way you like. 00:14:16.700 --> 00:14:18.280 And you're like, well, now it's broken. 00:14:18.280 --> 00:14:18.980 Oh, no. 00:14:18.980 --> 00:14:20.300 What am I going to do, right? 00:14:20.500 --> 00:14:22.760 It's small bites, of course, small steps. 00:14:22.760 --> 00:14:28.340 This portion of Talk Python To Me is brought to you by Linode. 00:14:28.340 --> 00:14:37.280 Whether you're working on a personal project or managing your enterprise's infrastructure, Linode has the pricing, support, and scale that you need to take your project to the next level. 00:14:37.280 --> 00:14:47.780 With 11 data centers worldwide, including their newest data center in Sydney, Australia, enterprise-grade hardware, S3-compatible storage, and the next-generation network, 00:14:48.260 --> 00:14:51.940 Linode delivers the performance that you expect at a price that you don't. 00:14:51.940 --> 00:15:07.980 Get started on Linode today with a $20 credit, and you get access to native SSD storage, a 40-gigabit network, industry-leading processors, their revamped cloud manager at cloud.linode.com, root access to your server, along with their newest API, and a Python CLI. 00:15:08.400 --> 00:15:16.020 Just visit talkpython.fm when creating a new Linode account, and you'll automatically get $20 credit for your next project. 00:15:16.020 --> 00:15:17.300 Oh, and one last thing. 00:15:17.300 --> 00:15:18.080 They're hiring. 00:15:18.080 --> 00:15:21.180 Go to linode.com slash careers to find out more. 00:15:21.180 --> 00:15:22.500 Let them know that we sent you. 00:15:25.620 --> 00:15:36.800 When you had live here, and you talked about the live, sort of interactive, almost onstage type of stuff, I was thinking of FoxDot and the live, like, musical composition with Python. 00:15:36.800 --> 00:15:38.000 Have either of you guys seen that? 00:15:38.000 --> 00:15:38.920 No, I haven't seen this. 00:15:38.920 --> 00:15:39.360 I have not. 00:15:39.360 --> 00:15:40.500 I think it's called FoxDot. 00:15:40.500 --> 00:15:42.440 I'll put a link of a presentation. 00:15:42.440 --> 00:15:45.580 So there's this thing called FoxDot, which is Python-based. 00:15:45.580 --> 00:15:48.540 People can go up there, and they basically create little, like, symphonies. 00:15:48.540 --> 00:15:49.720 Oh, that is awesome. 00:15:50.040 --> 00:15:52.840 And with Python, and it's super interesting. 00:15:52.840 --> 00:15:56.440 Anyway, I was going to ask your thoughts on that as a sort of HCI person. 00:15:56.440 --> 00:15:57.700 Yeah, yeah. 00:15:57.700 --> 00:15:58.560 I'll get it next time. 00:15:58.560 --> 00:16:01.380 Yeah, I grabbed a, yeah, I'm just looking at the page now. 00:16:01.380 --> 00:16:02.520 But yeah, that's a whole other world. 00:16:02.520 --> 00:16:03.480 I mean, it's an awesome world. 00:16:03.480 --> 00:16:08.140 I mean, this whole world of just live performance and, you know, live streaming and stuff is fascinating. 00:16:08.140 --> 00:16:14.500 You know, maybe the next Twitch stars are going to be data scientists, right, who are, you know, live streaming their exploratory analysis. 00:16:14.500 --> 00:16:19.380 There are some data scientists I've seen who do these live streams or do these, you know, recorded sessions. 00:16:19.380 --> 00:16:24.960 Like, let me just play with this data or do this Kaggle thing and just watch me play with it, which is really cool. 00:16:24.960 --> 00:16:26.820 That's really, I think you're right. 00:16:26.820 --> 00:16:27.780 I think it's super interesting. 00:16:27.780 --> 00:16:34.700 I feel like a lot of algorithms and libraries and just software products, I guess you'd call them. 00:16:34.700 --> 00:16:38.040 It could be very stale, you know, much like math or science. 00:16:38.040 --> 00:16:39.280 Like, here's the final result. 00:16:39.280 --> 00:16:40.920 Here's the algorithm or the formula. 00:16:41.420 --> 00:16:51.020 But that doesn't give you an appreciation of, like, seeing the, how do I, like, bumble around and, like, discover and think and try and explore and then come up with an answer. 00:16:51.020 --> 00:16:55.220 I think that would be really interesting for data science and, like, Kaggle competitions and stuff. 00:16:55.340 --> 00:16:55.500 Yeah. 00:16:55.500 --> 00:17:11.440 And I think that segues to our notebook work, too, because I feel like a lot of these notebooks, maybe Sam, I'm going to talk more on this later, is that, like, a lot of these sorts of interfaces could support this sort of liveness and this sort of iteration and keeping different versions around and seeing how you, you know, bumble your way toward a final result. 00:17:11.640 --> 00:17:11.980 Yeah, yeah. 00:17:11.980 --> 00:17:16.640 Maybe the next big particle physics discovery Nobel Prize will be live streamed. 00:17:16.640 --> 00:17:16.980 All right. 00:17:16.980 --> 00:17:19.580 So, Sam, let's get to the analysis. 00:17:19.580 --> 00:17:31.060 So, you said there were a bunch of different environments that you studied and you broke them into three categories, the academic world, the product world, and the R&D world. 00:17:31.060 --> 00:17:34.440 Maybe give us a sense of what's in each one. 00:17:34.440 --> 00:17:42.680 I don't know if we want to go through all of them, but read off maybe, like, five from each of those that you feel like is representative of that area, just to give people a sense of what's all out there. 00:17:42.680 --> 00:17:43.060 Yeah. 00:17:43.060 --> 00:17:51.900 So, probably start with the industry world because that world is probably the most familiar to us and maybe to your listeners as well. 00:17:51.900 --> 00:18:15.280 So, in the product world, we kind of have Jupyter Notebook and we also have, like, Google Colab, which is Google's version of Jupyter Notebook, as well as Azure Notebooks, which is kind of like this up-and-coming notebook system for JavaScript. 00:18:15.480 --> 00:18:28.600 Okay. Oh, right, right. Yeah. There is this move to try to make a JavaScript equivalent, basically, of what Python has, right? I mean, it's pretty natural to run JavaScript, but the trick is to find the libraries, right? 00:18:28.600 --> 00:18:29.580 Exactly. Yeah. 00:18:29.580 --> 00:18:30.480 And maybe integers. 00:18:30.480 --> 00:18:31.080 Yeah. 00:18:31.080 --> 00:18:36.900 No integers in JavaScript can make that a little hard, but still, very cool. 00:18:36.900 --> 00:18:52.180 All right. And then, next category is academic notebook projects. And these projects typically come in the form of, like, papers. So, people will prototype some interaction and then submit a paper in a human-computer interaction conference. 00:18:53.140 --> 00:19:12.260 And these papers typically have, like, they take a base notebook system and extend it in one particular way. For example, the Callisto system takes Jupyter Notebook and adds in, like, some live collaboration tools, like a chat window, as well as, I think, being able to see other people's cursors and where they are on the screen. 00:19:12.260 --> 00:19:15.560 A little bit like a Google Doc-ification of Jupyter. 00:19:15.760 --> 00:19:41.200 Yeah, yeah. That's what they're going for. We have tools like Rex, which is a tool specifically for, like, data cleaning. So, the idea there is that you would, you could display a data frame in your notebook and then, by example, show Python, in a sense, like, what you want the data to look like after data cleaning. And it can try to infer the Python code that would generate that, that would change the data in the way that you want. 00:19:41.200 --> 00:19:45.080 Oh, that's pretty cool. And so, it uses a bit of machine learning or something. 00:19:45.680 --> 00:20:15.280 Yeah, it uses what's called program synthesis. Yeah, that actually was a project led by Ian, who was my other PhD student, who was a co-author on this paper. He did this at a Microsoft research internship. So, they have this program synthesis technology. They have this, you know, engine that basically synthesizes pieces of code. And they applied this technology to notebooks. So, what you can actually do is you can, you know, in a data table, you can say, here are some examples of the way I want to clean the data. And then it'll infer. I don't think it actually is machine learning because it's not database. It's kind of rules-based. 00:20:15.600 --> 00:20:45.580 Right, right, right, right. 00:20:45.980 --> 00:20:53.500 So, you kind of have, like, these experimental, but also, like, you look at it and you're like, well, that could be useful in a real system. 00:20:53.700 --> 00:21:02.640 Yeah, yeah, yeah, I could see that for sure. You maybe just drop a marker, like, hey, I'm going to go explore this. I might want to, like, a save point type of thing. Maybe you want to get back here. 00:21:02.640 --> 00:21:03.140 Exactly. 00:21:03.140 --> 00:21:14.440 And then, right, it'll just sort of, that's, like, behind the scenes. We're going to create a Git branch and start doing stuff for you, right? Like, that'd be really nice for people who are, they don't know how to, like, deal with Git branches or want to deal with it and stuff like that. 00:21:14.500 --> 00:21:21.380 Yeah, that's exactly the use case that they were going for. It's for this very exploratory, early stage, like, data exploration. Yeah. 00:21:21.380 --> 00:21:34.140 Yeah. Yeah. Okay, cool. The one in the product space that seems to be closest to that is Gigantum. I had those guys on the show and there's a little bit of, like, auto-versioning collaboration stuff in that world. 00:21:34.280 --> 00:21:40.000 Yeah, yeah. We looked at Gigantum in this paper and I think we did mark it down as being relatively similar to that. 00:21:40.000 --> 00:21:43.160 Yeah, yeah, cool. All right, the third category was experimental and R&D. 00:21:43.160 --> 00:21:59.660 Yeah. So experimental R&D is, in a sense, very much like the academic projects in that they are, they're trying to test out new interactions or new features that might not, like, don't have immediate, like, industry use cases or maybe, like, business propositions. 00:22:00.200 --> 00:22:12.640 And so we have tools like Carbide is an interesting one. Carbide actually allows you to, it creates widgets for your Python variables in line with your Python code. 00:22:12.640 --> 00:22:17.360 And then it allows you to, like, use sliders to change your variables on the spot. 00:22:17.360 --> 00:22:18.280 Okay. 00:22:18.280 --> 00:22:25.620 And it's, like, way out there. It's really interesting. But you can, it actually infers, like, what values your Python variables should take on to produce an output. 00:22:25.620 --> 00:22:35.440 So if you use a widget to say, like, oh, I want Y to be, like, 32, then if Y depends on X, then it'll also update X at the same time. 00:22:35.440 --> 00:22:44.060 So it does, like, it's weird. It does look like it's really, like, intriguing, like, program synthesis or, like, program, like, inference, like, behind the scenes to get your program to look right. 00:22:44.060 --> 00:22:46.600 Right, to look at the dependencies and figure out what it has to rerun. 00:22:46.600 --> 00:22:47.420 Yeah, yeah. 00:22:47.620 --> 00:22:55.620 Yeah, that's really cool. It sounds a little bit like the interactive widgets in Jupyter that you can have, you know, you can sort of put some sliders and adjust some things. 00:22:55.620 --> 00:23:03.760 But I don't believe the Jupyter one has a lot of pendency type of stuff on down the line, right? It's just like this one cell as you slide the widget recomputes. 00:23:03.760 --> 00:23:09.660 Yeah, yeah. Like, Carbide is like those Jupyter widgets, but super next level. The widgets can change the code itself. 00:23:09.760 --> 00:23:10.760 Okay, that's pretty wild. 00:23:10.760 --> 00:23:29.760 Yeah. We have a few industry projects here as well. We have projects from, like, Netflix and Stripe, where they take notebooks, but use notebooks as, like, kind of, like, recurring jobs instead of just one-off, like, exploratory scripts, where they use notebooks as these jobs. 00:23:29.760 --> 00:23:35.420 And so they can re- the notebook might automatically rerun, like, once a day to perform some data operation. 00:23:36.080 --> 00:23:40.620 They're kind of, like, broadening the scope of what notebooks can be used for in a large-scale setting. 00:23:40.620 --> 00:23:45.740 Yeah, you know, the stuff that Netflix was doing with, like, Paper Mill was super interesting. 00:23:45.740 --> 00:23:56.660 How they're using it to automate a bunch of stuff and sort of schedule a bunch of things to work, but also the fact that if it fails, the notebook has the output of the failure. 00:23:56.660 --> 00:24:02.360 So basically, you can, you know, save the notebook in that state and just go look at it and see what went wrong. 00:24:02.420 --> 00:24:09.140 Rather than there's a log message that says something went wrong, like, there's the whole sequence with the history of the notebook output, which is pretty awesome. 00:24:09.140 --> 00:24:13.440 Yeah. I think the Paper Mill stuff, I mean, they have a blog post about it, but we link to all of this in the paper. 00:24:13.440 --> 00:24:15.660 So if you, you know, link to our paper, we have all the URLs. 00:24:15.660 --> 00:24:17.180 I mean, I thought that was interesting. 00:24:17.180 --> 00:24:22.620 I mean, we didn't really capture this in the- in our space because, you know, most of it is for data scientists. 00:24:22.620 --> 00:24:28.900 But the Paper Mill project is interesting because it's using notebooks for, like, DevOps and for, you know, production deployment work, right? 00:24:28.900 --> 00:24:34.560 It's like really using the notebooks as a kind of a DevOps-y sort of tool rather than a direct data science tool. 00:24:34.560 --> 00:24:37.800 So I thought that was a really interesting and out-of-the-box use case. 00:24:37.800 --> 00:24:44.400 It's like instead of running cron jobs and seeing, like, gigantic log files on your terminal or grepping through text files, 00:24:44.400 --> 00:24:50.980 you can just, you know, reconstruct it in a notebook and actually use a notebook to debug your production software, which is cool. 00:24:50.980 --> 00:24:56.420 Yeah. And the essence of that, I think, was one of the things you do with Paper Mill, if I have it remembered correctly, 00:24:56.420 --> 00:24:59.740 is the notebooks can take inputs and they can generate outputs. 00:24:59.740 --> 00:25:03.680 They're, like, sort of become functions in the traditional black box sense of it. 00:25:03.680 --> 00:25:04.000 Yeah. 00:25:04.000 --> 00:25:05.180 Yeah. That's pretty neat. 00:25:05.180 --> 00:25:08.000 Another one I see that you have in there is NB Dev. 00:25:08.000 --> 00:25:10.360 NB Dev is a really interesting project. 00:25:10.360 --> 00:25:15.820 I just, we talked about that, Brian Nock and I did a couple episodes ago on Python Bytes. 00:25:15.820 --> 00:25:18.140 And I was like, oh, well, this might be kind of interesting. 00:25:18.140 --> 00:25:21.800 And I looked and I'm like, wow, this is something really quite special. 00:25:21.800 --> 00:25:24.120 It does a lot of neat things over there. 00:25:24.120 --> 00:25:32.460 Right. So if I'm remembering correctly, NB Dev allows you to take a notebook and, like, deploy it as a Python library. 00:25:32.460 --> 00:25:35.640 Yes, exactly. Yeah. You can basically turn it into a package. 00:25:35.640 --> 00:25:38.020 Right. And it can have the test, like, built into it. 00:25:38.020 --> 00:25:47.660 And, yeah, it's, again, a really interesting example of, like, notebooks are kind of being reused in all sorts of ways that we would see traditional Python scripts being reused. 00:25:47.660 --> 00:25:54.440 So in this case, instead of writing a set of Python files to become a Python package, you could write a notebook instead. 00:25:54.440 --> 00:25:56.040 And that could be, like, your package. 00:25:56.040 --> 00:26:00.780 Yeah. I'll give you the quick rundown for people who are listening and might be interested just off their little highlights. 00:26:01.320 --> 00:26:04.200 So you can export the functionality of a notebook to a library. 00:26:04.200 --> 00:26:08.560 They have a CLI commands you can use to interact with NB Dev. 00:26:08.560 --> 00:26:11.760 Export to HTML builds documentation for your library. 00:26:11.760 --> 00:26:13.800 Sync brings the Python. 00:26:13.800 --> 00:26:17.200 So you can edit the Python code and then bring it back into your library. 00:26:17.200 --> 00:26:20.060 You can put the tests in there and run them in CI. 00:26:20.060 --> 00:26:24.660 There's just a button that also has this, like, NB strip out concept as well for checking stuff in. 00:26:24.760 --> 00:26:29.280 So, yeah, that's, I mean, there's a bunch of exploration and innovation happening here. 00:26:29.280 --> 00:26:30.240 Yeah, it's really interesting. 00:26:30.240 --> 00:26:34.480 One other interesting project in the experimental section is, like, Iodide. 00:26:34.480 --> 00:26:37.440 Iodide is, Iodide is, like, super experimental. 00:26:37.440 --> 00:26:44.120 But they compiled Python and some of its data science libraries into WebAssembly. 00:26:44.120 --> 00:26:49.000 And then because of that, you can run, you can write Python in the browser. 00:26:49.000 --> 00:26:53.020 And it doesn't require a separate Python process to run that. 00:26:53.020 --> 00:27:00.780 Instead, it runs completely within the browser because your browser is interpreting the WebAssembly version of, like, the Python interpreter. 00:27:00.780 --> 00:27:05.180 And, like, I think they have, like, NumPy and, like, Matplotlib support in there, too. 00:27:05.180 --> 00:27:11.480 Yeah, they've recompiled many of the core, what would be C-based Python libraries as well, right? 00:27:11.480 --> 00:27:13.660 Like you said, like, NumPy and maybe Matplotlib. 00:27:13.660 --> 00:27:14.200 I can't remember. 00:27:14.200 --> 00:27:15.420 But some of those types of libraries. 00:27:15.420 --> 00:27:15.760 Right. 00:27:15.760 --> 00:27:19.720 It's really, it took a while for me to wrap my mind around it. 00:27:19.720 --> 00:27:21.460 But I was really amazed when I first saw that. 00:27:21.460 --> 00:27:25.980 I think WebAssembly has some real interesting possibilities to bring a lot of this stuff together. 00:27:25.980 --> 00:27:34.300 You know, it's one of the things that's interesting about some of the stuff you were covering in these different projects here is just what it means to run the notebook, right? 00:27:34.300 --> 00:27:42.880 Sometimes that means a Docker image is spun up and configured exactly for you in some environment or even, like, with Gigantum on your local machine. 00:27:42.880 --> 00:27:47.260 With Google CoLab, there's, like, a way to say, run this on a GPU, by the way. 00:27:47.260 --> 00:27:47.880 Yeah. 00:27:47.880 --> 00:27:49.980 Just to get a GPU to run it on. 00:27:49.980 --> 00:27:53.080 And Iodide runs in the browser through WebAssembly. 00:27:53.680 --> 00:27:56.620 There's just, there's a bunch of options here that are pretty cool. 00:27:56.620 --> 00:27:56.940 Yeah. 00:27:56.940 --> 00:28:01.120 Kind of to Michael's point, it's like, you know, we've just informally kind of starting this project. 00:28:01.120 --> 00:28:02.480 We've kind of heard about all these. 00:28:02.480 --> 00:28:04.080 It's felt like the Wild West, right? 00:28:04.080 --> 00:28:05.420 It's like, oh, you can run the notebooks locally. 00:28:05.420 --> 00:28:06.600 You can run some JavaScript. 00:28:07.100 --> 00:28:08.240 Observable ones run the browser. 00:28:08.240 --> 00:28:10.940 And, like, you know, Iodide, Pyodide compiles. 00:28:10.940 --> 00:28:12.240 The WebAssembly runs the browser. 00:28:12.240 --> 00:28:14.240 And then you need some other ones in the cloud. 00:28:14.240 --> 00:28:16.060 And then some of them have this live editing. 00:28:16.060 --> 00:28:16.640 Some of them don't. 00:28:16.640 --> 00:28:25.060 So that's why kind of motivation for putting together this paper was, you know, how do we just categorize all this in the most kind of concise way we can? 00:28:25.060 --> 00:28:29.840 And, I mean, we're going to talk about more of this detail, but we have one category about runtime environment, right? 00:28:29.840 --> 00:28:34.700 So we categorize, here's the four types of main types of runtime environments people have and other things. 00:28:34.700 --> 00:28:38.540 So the kind of main contribution of this paper was really people have all these intuitions. 00:28:38.540 --> 00:28:40.560 They've heard about these notebooks doing all these different things. 00:28:40.560 --> 00:28:44.300 Let's just try to plop them all together in a map, in a sense, right? 00:28:44.300 --> 00:28:47.960 Like, where does everything lie on this wild, this high-dimensional space? 00:28:47.960 --> 00:28:49.340 Yeah, it's highly dimensional. 00:28:49.340 --> 00:28:51.160 We're going to talk about the dimensions in a minute. 00:28:51.160 --> 00:28:52.500 It's more than three dimensions. 00:28:52.500 --> 00:28:58.060 This portion of Talk Python To Me is brought to you by Sentry. 00:28:58.060 --> 00:29:00.700 How would you like to remove a little stress from your life? 00:29:00.700 --> 00:29:06.440 Do you worry that users may be having difficulties or are encountering errors with your app right now? 00:29:06.440 --> 00:29:09.340 Would you even know it until they send that support email? 00:29:09.340 --> 00:29:19.720 How much better would it be to have the error details immediately sent to you, including the call stack and values of local variables, as well as the active user stored in the report? 00:29:20.180 --> 00:29:23.500 With Sentry, this is not only possible, it's simple and free. 00:29:23.500 --> 00:29:26.900 In fact, we use Sentry on all the Talk Python web properties. 00:29:26.900 --> 00:29:33.440 We've actually fixed a bug triggered by our user and had the upgrade ready to roll out as we got the support email. 00:29:33.440 --> 00:29:35.240 That was a great email to write back. 00:29:35.240 --> 00:29:37.700 We saw your error and have already rolled out the fix. 00:29:37.700 --> 00:29:38.880 Imagine their surprise. 00:29:38.880 --> 00:29:41.560 Surprise and delight your users today. 00:29:41.560 --> 00:29:49.460 Create your free account at talkpython.fm/sentry and track up to 5,000 errors a month across multiple projects for free. 00:29:49.460 --> 00:30:00.720 And if you use the code Talk Python, all one word, it's good for two free months of Sentry's team plan, which will give you up to 20 times as many monthly events in some other features. 00:30:00.880 --> 00:30:02.460 So create that free account today. 00:30:06.200 --> 00:30:12.520 One thing I do want to talk about maybe first is just like some of the challenges of notebooks and then the paradox of choice. 00:30:12.520 --> 00:30:17.760 So first of all, I think this comes from the paper, but also some of the feelings I had. 00:30:17.760 --> 00:30:29.200 So you talked about some of the challenges, Sam, being stale data, out of order execution, abundance of code, and then the inability for composition, things that Paper Mill, for example, is trying to solve. 00:30:29.940 --> 00:30:32.520 Do you want to speak to some of the challenges you saw out in this space? 00:30:32.520 --> 00:30:51.000 Well, one of the first challenges that a user of Drupal notebooks encounters is typically the out-of-order execution problem, where because you can run cells in any order you want, what might happen is you run some cells, and then you delete some cells, and then you write more cells, and you run those in some out-of-order way. 00:30:51.000 --> 00:30:59.160 And what happens is, well, one problem is that sometimes your variables just get changed, and you don't know where a change came from because your code might be gone. 00:30:59.480 --> 00:31:01.540 Right. You might even delete the cell that defined them. 00:31:01.540 --> 00:31:02.600 Yeah, yeah. 00:31:02.600 --> 00:31:06.340 But the kernel's still running, so they're still in memory. For now, it's going to be okay. 00:31:06.340 --> 00:31:13.960 Exactly, exactly. It happens really, really often, in my experience, when we teach students how to use Drupal notebooks for the first time. 00:31:13.960 --> 00:31:22.660 They hit some random keyboard shortcut, and their cell's gone. And it looks like the code's running fine, but then when they try to turn in their code, everything kind of goes up in flames. 00:31:22.660 --> 00:31:40.180 Yeah. Well, but it's so much the power of Jupyter as well, right? Like, there's some reports and stuff I have for my business behind the scenes, where there's like a 30-second bunch of reading data, correlating data, and then you want to graph it and slice it and analyze it. 00:31:40.340 --> 00:31:56.540 And the ability to just go, rerun the cell that sorts it this way and shows that aspect, it runs instantly, right? It just goes and goes and goes. And it's so powerful to not have to rerun that code to be able to change your analysis or how you're exploring it, right? 00:31:56.660 --> 00:32:00.740 Yeah, exactly. That was the selling point of Jupyter. That's why I've got everyone hooked on Jupyter to begin with. 00:32:00.740 --> 00:32:00.980 Yeah. 00:32:00.980 --> 00:32:01.860 It was like that ability to really quickly. 00:32:01.860 --> 00:32:04.900 Yeah, you do the expensive part. Exactly. Exactly. 00:32:04.900 --> 00:32:09.080 But it's like the world's worst go-to and then some, you know? 00:32:09.080 --> 00:32:16.140 Sure. It's like the human-powered go-to, right? Like, you're self-like running the cells and go-to statements yourself. 00:32:16.140 --> 00:32:23.080 Exactly. Like, with a real go-to, it might be really hard to understand, but you can literally go through the code and see it. 00:32:23.080 --> 00:32:24.340 Yeah, it's at least deterministic. 00:32:24.560 --> 00:32:29.020 Yeah, it's at least deterministic, right? But yeah, it's not even with Notebook. So that's a challenge. 00:32:29.020 --> 00:32:29.960 Yeah, exactly. 00:32:29.960 --> 00:32:37.780 Because to me, that's the biggest challenge. Yeah, it's this sort of out-of-order stuff. But it's so hard to say, like, well, let's just not do that because it's also the superpower. 00:32:37.780 --> 00:32:53.560 Right, right. One of the dimensions in this space, I don't know if we're getting there quite yet, but one of the dimensions in this space specifically addresses this because a number of projects that we see from academia and from industry try to, like, address this specific in-order, out-of-order execution issue. 00:32:53.960 --> 00:33:06.380 Projects like Papermode, which we briefly talked about before, address it in a different way by trying to just, like, make sure that whenever a Notebook is run, it only runs from top to bottom like a computer would, rather than how a human might execute it. 00:33:06.520 --> 00:33:16.420 So there's different angles on how to get around an out-of-order execution thing. But as you mentioned, it's like the main benefit and also the main weakness of these Notebook systems. 00:33:16.640 --> 00:33:19.100 Yeah. Did you guys study Streamlit? 00:33:19.100 --> 00:33:20.400 Streamlit? 00:33:20.400 --> 00:33:20.500 Yeah. 00:33:21.640 --> 00:33:27.760 So Streamlit has a really interesting way of solving that in that they basically use Funktool's LRUCache. 00:33:27.760 --> 00:33:31.260 They don't exactly use that, but they more or less, the concept is the same. 00:33:31.260 --> 00:33:35.120 And they've got these different functions that more or less act like cells. 00:33:35.120 --> 00:33:38.960 And when you rerun it, if you give it the same inputs, it uses the cached version. 00:33:38.960 --> 00:33:42.120 But if the inputs happen to change, then it'll give you a different output. 00:33:42.280 --> 00:33:47.520 So it sort of keeps that, you know, don't recompute the stuff I already know aspect. 00:33:47.520 --> 00:33:48.420 Right, right. 00:33:48.420 --> 00:34:01.780 So Streamlit has an interesting execution model that I think we also noticed in some other systems where it looks to you as a user that whenever you make a change, your Notebook gets rerun from top to bottom every time. 00:34:01.780 --> 00:34:06.740 But then behind the scenes, it does some smart caching and computing to avoid the work. 00:34:07.280 --> 00:34:16.500 So you kind of get the benefit of rerunning a script every time without the drawback of having to like wait a long time for some expensive operation to recompute. 00:34:16.500 --> 00:34:16.800 Yeah. 00:34:16.800 --> 00:34:17.460 Yeah. 00:34:17.460 --> 00:34:17.920 Interesting. 00:34:17.920 --> 00:34:21.200 This is a big challenge to put all these things together. 00:34:21.200 --> 00:34:25.680 I mentioned The Paradox of Choice, which is a fantastic book if people haven't read it. 00:34:25.680 --> 00:34:33.280 But I feel like one of the challenges that things you all may experience is if you look at all these different things, you're like, oh, Rex does this amazing thing. 00:34:33.280 --> 00:34:34.640 Callisto does that. 00:34:34.640 --> 00:34:36.140 And Streamlit does this other thing. 00:34:36.240 --> 00:34:38.540 And then Jupyter has this amazing aspect. 00:34:38.540 --> 00:34:42.540 Not any single one of these has all the good things you want. 00:34:42.540 --> 00:34:50.180 But any choice down one path will make you feel like, well, I don't get the machine learning like auto cleanup in this one. 00:34:50.180 --> 00:34:52.040 And I don't get the, you know, the cursors. 00:34:52.040 --> 00:34:59.540 Like, do you feel like there's a bit of a, I don't know, fear of missing or feeling of missing out as you like have to go down? 00:34:59.540 --> 00:35:00.300 You got to use one. 00:35:00.300 --> 00:35:01.100 You can't use them all. 00:35:01.100 --> 00:35:02.260 Oh, yeah, for sure. 00:35:02.500 --> 00:35:06.360 Nowadays, when I use Jupyter, I mean, Jupyter is a system I still use. 00:35:06.360 --> 00:35:13.380 But now that I know about all these other systems, oftentimes when I use Jupyter now, I'm like, dang, if only I had, you know, Next Journal. 00:35:13.380 --> 00:35:15.420 If only I had like Databricks right now. 00:35:15.420 --> 00:35:15.780 Yeah. 00:35:16.040 --> 00:35:18.520 But then you'd want Jupyter to lab back for something else, right? 00:35:18.520 --> 00:35:21.340 I mean, it's like, there's a lot of different things going on here. 00:35:21.340 --> 00:35:23.980 It's a lot of flowers blooming, I would say, I guess. 00:35:23.980 --> 00:35:24.840 Right, right. 00:35:24.960 --> 00:35:30.880 So we talked a little bit at the end of the paper in the kind of discussion part about, you know, what if you did this kitchen sink, right? 00:35:30.880 --> 00:35:32.520 What if you just threw everything together? 00:35:32.520 --> 00:35:34.580 What if someone made a notebook that has everything? 00:35:34.580 --> 00:35:38.120 And, you know, one, it's an enormous software engineering challenge, right? 00:35:38.120 --> 00:35:42.900 Nobody would want to take on that challenge of maintaining 500 different sorts of features. 00:35:43.100 --> 00:35:46.740 And the other one is that it just may make the interface really complicated, right? 00:35:46.740 --> 00:35:58.840 It's not like, if you can choose, imagine, like, you can choose different kinds of execution orders or choose a reactive versus an in-order one and you choose all the stuff, like, it would just be really, really hard for people to set up. 00:35:58.840 --> 00:36:08.860 So, you know, some of these notebook systems, you know, say Jupyter or some of the other ones, I think they've been successful because they've, I would say, I don't want to say straightforward, but their feature set is limited, right? 00:36:08.860 --> 00:36:16.100 Like that, you know, it's, even though Jupyter has this out-of-order issue, it's like, okay, what you see is what you get as long as you execute in that order. 00:36:16.100 --> 00:36:16.440 Yeah. 00:36:16.440 --> 00:36:33.020 And some of the real power is you can have semi-structured, reusable tiny bits in, say, cells when you're just a biologist or an economist and you don't have to become a computer scientist person to, like, learn Haskell to take advantage of this thing. 00:36:33.020 --> 00:36:42.840 Like, it's incredibly, yeah, it's incredibly accessible to just jump in and write a few lines of imperative code and then get some really awesome output. 00:36:42.840 --> 00:36:50.920 And so, yeah, I feel like if it became too advanced, right, you're like, well, do you want the reactive model, the asynchronous model? 00:36:50.920 --> 00:36:52.020 You're like, I don't know what this means. 00:36:52.020 --> 00:36:53.440 I just want a graph, right? 00:36:53.640 --> 00:36:56.340 Yeah, because, I mean, one of the challenges with, say, Observable, right? 00:36:56.340 --> 00:37:03.140 I mean, Observable has this really elegant reactive thing where this stuff auto-updates, and they have a lot of examples, right, on their website. 00:37:03.140 --> 00:37:09.540 But I still think it's going to be hard for regular people, regular programmers to pick up because it looks kind of like JavaScript. 00:37:09.540 --> 00:37:15.960 But then when you write it, it's like it has all these, like, little extensions to JavaScript that you need to wrap your head around. 00:37:15.960 --> 00:37:22.700 So, you know, it's the age-old thing of, you know, more powerful programming languages environments are just going to be higher barrier to learn, right? 00:37:22.700 --> 00:37:24.160 So it's very hard in practice. 00:37:24.160 --> 00:37:24.700 Yeah, yeah. 00:37:24.700 --> 00:37:26.360 So let's talk about the dimensions. 00:37:26.360 --> 00:37:34.460 Sam, you put together 10 different dimensions on how you sort of evaluate these things and, you know, a spectrum along them. 00:37:34.460 --> 00:37:39.580 For example, like data sources, you have local files, cloud storage, large data, streaming data. 00:37:39.580 --> 00:37:41.640 Want to run us through this? 00:37:41.860 --> 00:37:42.300 Sure. 00:37:42.300 --> 00:37:45.200 We can start with, like, the highest level breakdown. 00:37:45.200 --> 00:37:51.360 We organize the design space dimensions here around the steps in the workflow of a data scientist. 00:37:51.360 --> 00:38:02.180 So as a data scientist, we imagine that you might start a project by importing your data and then writing and editing your code and your prose around that code. 00:38:02.180 --> 00:38:04.480 And then you would want to run that code. 00:38:04.480 --> 00:38:07.600 And then finally, you'd want to publish your notebook in some way. 00:38:07.740 --> 00:38:12.980 So it's like the input, editing, and output steps for a data scientist. 00:38:12.980 --> 00:38:16.160 And those are, like, the workflow steps. 00:38:16.160 --> 00:38:28.840 And the 10 dimensions that we pulled out were data sources, as you mentioned, and things like versioning and collaboration, as well as execution models and execution liveness. 00:38:29.300 --> 00:38:38.320 And for publishing a notebook, we talked about, like, you might want a notebook as a static HTML page or maybe as a software package, as we alluded to earlier. 00:38:38.320 --> 00:38:39.880 There is a lot to consider here. 00:38:39.880 --> 00:38:45.240 Let's just maybe grab a couple and to give people a sense, then they can go look at the paper. 00:38:45.240 --> 00:38:48.680 You've got a big full page graph chart type of thing. 00:38:48.680 --> 00:38:49.300 Yep. 00:38:49.300 --> 00:38:50.040 Chart, I guess. 00:38:50.400 --> 00:38:52.320 For example, I mentioned data sources, right? 00:38:52.320 --> 00:38:53.700 We've got local files. 00:38:53.700 --> 00:38:56.060 And you say, okay, this is access local file system. 00:38:56.060 --> 00:38:59.920 And you always give an example of a system that implements that. 00:38:59.920 --> 00:39:01.400 Like, so RStudio, for example. 00:39:01.400 --> 00:39:02.360 Right, right. 00:39:02.360 --> 00:39:03.840 What else is on that access there? 00:39:03.840 --> 00:39:04.240 Yeah. 00:39:04.320 --> 00:39:12.980 So we have local files, which means that the system basically only, like, natively supports opening files that are stored locally on the same computer. 00:39:12.980 --> 00:39:20.760 Some notebook systems also allow you to read in cloud files, like files on their servers as though they were local. 00:39:20.760 --> 00:39:23.100 This is what Google Colab does. 00:39:23.100 --> 00:39:28.520 It allows you to read in files from Google Drive as if they were stored locally alongside a notebook. 00:39:29.500 --> 00:39:38.280 We have some systems that have some special handling for large data sets, like data sets that don't fit in a computer's memory, like Databricks, uses Spark to handle that. 00:39:38.280 --> 00:39:38.660 Right. 00:39:38.660 --> 00:39:41.840 And we also have some systems that support streaming data. 00:39:41.840 --> 00:39:47.080 So, like, you might hook up your notebook to some, like, web socket on the internet. 00:39:47.080 --> 00:39:52.980 And as new data comes in, your notebook will automatically rerun and update to reflect the latest versions of those data. 00:39:52.980 --> 00:40:02.280 Oh, like if you want to set up a dashboard and as things behind the scenes change, like, the notebook is always sort of up to date with the state of the world. 00:40:02.280 --> 00:40:03.540 Polling almost, but better. 00:40:03.540 --> 00:40:04.420 Yeah, exactly. 00:40:04.420 --> 00:40:11.260 So you can kind of see from the way you described that, that every single, like, dimension here has some use case. 00:40:11.260 --> 00:40:19.360 And the use case is kind of motivated by some real world, like, somebody sat down and was like, man, I don't like rerunning my notebook to update my dashboard every time. 00:40:19.360 --> 00:40:23.620 I just wish I had some way to, like, have it automatically, like, streaming the data and update itself. 00:40:23.620 --> 00:40:23.900 Yeah. 00:40:23.900 --> 00:40:25.140 Yeah, for sure. 00:40:25.140 --> 00:40:29.640 So let's go just through two more and then I'll let people go check out the details for the rest. 00:40:29.640 --> 00:40:31.620 So how about execution order? 00:40:31.620 --> 00:40:42.940 Yeah, so execution order goes directly to the in order, out of order problem that we were talking about earlier, where the Drupal notebooks allow users to run cells in any order they choose. 00:40:42.940 --> 00:40:44.640 So I can jump around cells freely. 00:40:44.640 --> 00:40:50.220 Other systems like Observable require a certain order of cells. 00:40:50.220 --> 00:41:02.760 So what Observable does is it takes a cell, looks at all the dependencies, like, the variable dependencies of that cell, and then forces a notebook to run in the topically sorted order of cells, right? 00:41:02.760 --> 00:41:06.540 Depending on, like, what cells to run should depend on other cells. 00:41:06.540 --> 00:41:17.660 We also have systems like Streamlit, which force, like, this in order execution model that we discussed earlier, where it looks to the user, like, the cell, like, the notebook always runs from top to bottom every time. 00:41:17.800 --> 00:41:22.600 So those are the three main variations we found of ordering execution. 00:41:22.600 --> 00:41:23.860 Yeah, interesting. 00:41:23.860 --> 00:41:29.680 Let's go to the very last part, the end of this lifecycle, if you will, and talk about notebook outputs. 00:41:29.680 --> 00:41:39.080 Yeah, so notebook outputs, we found the most common use case, I think, reflected across all of our systems was taking a notebook and publishing it as a static report, so to speak. 00:41:39.080 --> 00:41:47.100 So it's essentially like taking a notebook and converting it to HTML and emailing the HTML file or putting it up as a web page somewhere. 00:41:47.540 --> 00:41:58.220 The next step up would be, like, a dashboard where you don't have, where you don't show viewers the code that you might, like, where you might require to produce to charts or graphs. 00:41:58.220 --> 00:42:01.220 You just show the charts or graphs, like, themselves. 00:42:01.220 --> 00:42:01.740 Right. 00:42:01.740 --> 00:42:02.680 Okay. 00:42:02.680 --> 00:42:13.600 And then finally, within that space, we have, like, what we call software artifacts, which encapsulate the idea of you can use notebooks as cron jobs, or you can use notebooks as software packages. 00:42:13.960 --> 00:42:17.240 Right, right, like NB dev creating something you could put on PyPI. 00:42:17.240 --> 00:42:18.120 Yeah, exactly. 00:42:18.120 --> 00:42:18.780 Exactly. 00:42:18.780 --> 00:42:19.220 Yeah. 00:42:19.220 --> 00:42:19.920 Yeah. 00:42:19.920 --> 00:42:28.060 I really think it's cool that you broke it down like this, but then each one has an example type of thing. 00:42:28.060 --> 00:42:30.740 So you're like, okay, well, how does, say, Gigantum do it? 00:42:30.740 --> 00:42:32.780 Or how does Streamlit do this thing? 00:42:32.780 --> 00:42:33.840 Or Observable? 00:42:33.840 --> 00:42:34.640 And so on. 00:42:34.840 --> 00:42:40.900 Yeah, I think actually seeing all this stuff together might inspire features to start cross-pollinating. 00:42:40.900 --> 00:42:42.940 Hopefully not every single feature, right? 00:42:42.940 --> 00:42:44.080 That would be too much, but still. 00:42:44.080 --> 00:42:44.860 The kitchen sink. 00:42:44.860 --> 00:42:45.100 Yeah. 00:42:45.100 --> 00:42:45.420 It seems powerful. 00:42:45.420 --> 00:42:46.960 Kitchen sink notebook. 00:42:46.960 --> 00:42:48.380 Yeah. 00:42:48.380 --> 00:42:49.780 Yeah. 00:42:49.780 --> 00:42:52.380 I mean, maybe kitchen sink is available on PyPI. 00:42:52.380 --> 00:42:53.420 Like, you could just go for that. 00:42:53.420 --> 00:42:53.540 Yeah, yeah. 00:42:53.540 --> 00:42:54.380 Pick install kitchen sink. 00:42:54.380 --> 00:42:54.600 Yeah. 00:42:54.600 --> 00:42:55.140 Exactly. 00:42:55.140 --> 00:42:57.480 Gives you half the kitchen sink. 00:42:57.480 --> 00:42:58.180 Takes half an hour. 00:42:58.180 --> 00:42:58.820 Yeah. 00:42:58.820 --> 00:42:59.660 Kitchen sink. 00:42:59.660 --> 00:43:00.860 Wait five minutes. 00:43:00.860 --> 00:43:01.980 Awesome. 00:43:01.980 --> 00:43:02.520 All right. 00:43:02.520 --> 00:43:04.380 Well, this is a really interesting view. 00:43:04.380 --> 00:43:08.640 I guess, you know, maybe let's kind of close out the conversation around this stuff by asking, 00:43:08.640 --> 00:43:10.820 I'll let you both each way in on this. 00:43:10.820 --> 00:43:12.160 What surprised you? 00:43:12.160 --> 00:43:16.860 Like, as you went through this, what did you expect and what was different than that? 00:43:16.860 --> 00:43:18.100 What surprised you as you went through this? 00:43:18.100 --> 00:43:21.860 I mean, first thing that surprised me is how many there were, but what else? 00:43:21.860 --> 00:43:27.100 For me, I think it was this tension between we hear about the problems of notebooks quite often. 00:43:27.360 --> 00:43:33.420 And yet we, from this analysis, we don't often see those desires reflected in actual products. 00:43:33.420 --> 00:43:43.880 So the in order, out of order execution model, for example, the vast majority of the systems we looked at only allowed any order execution where the user defines execution. 00:43:43.880 --> 00:43:48.920 Collaboration between users and notebook authors is relatively common. 00:43:49.060 --> 00:43:53.440 But in a company, you'd rarely find someone working on notebook all by their loans themselves. 00:43:53.440 --> 00:44:00.160 But at the same time, we see very few systems that support like the sort of real-time collaboration that Google Docs has made popular. 00:44:00.160 --> 00:44:00.440 Right. 00:44:00.440 --> 00:44:06.260 And it seems so, so easily on the cusp of that because it's already in the browser. 00:44:06.260 --> 00:44:08.060 Often it's already hosted. 00:44:08.060 --> 00:44:09.240 Right? 00:44:09.240 --> 00:44:09.660 Right. 00:44:09.660 --> 00:44:10.220 Right. 00:44:10.220 --> 00:44:10.560 Exactly. 00:44:10.800 --> 00:44:18.840 So it kind of reveals, I think as a whole, the analysis reveals that there's still room for improvement for a lot of these products. 00:44:18.840 --> 00:44:19.260 Yeah. 00:44:19.260 --> 00:44:19.980 Yeah. 00:44:19.980 --> 00:44:20.180 Good. 00:44:20.180 --> 00:44:20.420 Yeah. 00:44:20.420 --> 00:44:34.080 And I guess, I don't know if it was a surprise as much as, you know, something that, you know, we mentioned toward the end of the paper was that, you know, on the, you know, Sam kind of mentioned the product side of, you know, practically speaking, things are pretty tied to Jupyter-like things because people are used to that. 00:44:34.080 --> 00:44:41.200 And a lot of products are Jupyter extensions or Jupyter hosted on Azure, Jupyter hosted on AWS, Jupyter hosted on Google Cloud. 00:44:41.200 --> 00:44:48.320 And so my kind of focus is thinking about the academic side, think about the research papers that people have been prototyping for this. 00:44:48.320 --> 00:44:58.160 And, you know, I don't know if it was a surprise, but it's something that was kind of a call to action we had at the end of the paper is that, you know, I felt personally that the academic work was not sufficiently weird, maybe. 00:44:58.360 --> 00:44:58.480 Right. 00:44:58.480 --> 00:45:06.020 So it's like you want, you kind of, you, you know, one of the roles of academic research and prototypes is you want to push the bounds and try stuff that's very different. 00:45:06.020 --> 00:45:06.220 Right. 00:45:06.220 --> 00:45:09.800 So imagine experimental programming language or experimental toolkits. 00:45:09.800 --> 00:45:11.540 You know, the point isn't to be practical. 00:45:11.540 --> 00:45:14.120 The point is to stretch the limits of what we know. 00:45:14.120 --> 00:45:22.020 So because I think notebook research is still pretty early, it's all these systems in academia have been done probably within the last four or five years. 00:45:22.200 --> 00:45:28.200 It makes sense that the first wave is fairly kind of, you know, making the extensions that people expect, right? 00:45:28.200 --> 00:45:31.220 Versioning, in order execution, collaboration and stuff. 00:45:31.220 --> 00:45:39.120 So I think that what would excite me more to see in the coming years is just to see the academic projects or the R&D projects becoming more weird, right? 00:45:39.120 --> 00:45:46.100 So one of the things that we pitched at the end is just thinking about multi-device, different kinds of off-desktop devices, right? 00:45:46.100 --> 00:45:57.760 I mean, already we're computing with phones and smartwatches and augmented reality and displays, but we're still just hunched over in front of our laptop, typing in this, literally just typing in a text box in the web browser, right? 00:45:57.760 --> 00:45:58.840 Like that's all the notebooks are. 00:45:58.840 --> 00:46:00.600 They're just little text boxes in web browsers. 00:46:00.600 --> 00:46:04.720 So I think we can really expand our imagination thinking, you know, way beyond. 00:46:04.720 --> 00:46:06.540 That would excite me a lot more. 00:46:06.540 --> 00:46:07.360 Yeah. 00:46:07.360 --> 00:46:08.320 Very interesting. 00:46:08.320 --> 00:46:08.660 Yeah. 00:46:08.660 --> 00:46:14.400 And to echo that point, what we drew out of the notebook systems was like this data science workflow. 00:46:14.900 --> 00:46:24.680 But the reason why we drew out the data science workflow is because I think every single system that we've evaluated here was designed for data scientists for the most part. 00:46:24.680 --> 00:46:35.280 And so with data scientists as the audience of choice, I think they're necessarily like, it comes with like a limitation on what we can imagine notebook be used for. 00:46:35.280 --> 00:46:45.020 So in the paper, we talk about broadening notebook use cases for other types of users, like perhaps instructors for perhaps like maybe like artists, let's say. 00:46:45.020 --> 00:46:52.660 And each of those audiences has probably an entirely different set of desires and things they want to accomplish with programming. 00:46:53.060 --> 00:46:58.540 And I think with that comes like a whole nother set of trade-offs and perhaps design dimensions for notebook systems. 00:46:58.540 --> 00:46:59.280 That's a good point. 00:46:59.280 --> 00:46:59.520 Yeah. 00:46:59.520 --> 00:47:02.900 Like DJs, musicians, orchestras. 00:47:03.280 --> 00:47:04.280 Like what could you do? 00:47:04.280 --> 00:47:08.860 Like what kind of world would you build if you were trying to, you know, coordinate that kind of stuff? 00:47:08.860 --> 00:47:13.260 Philip, you brought up our phones, which are pretty incredible these days. 00:47:13.260 --> 00:47:15.160 Did you guys look at Carnet? 00:47:15.260 --> 00:47:28.620 I mean, it's not really something that needs separate analysis because it's just basically Jupyter, but Carnet, C-A-R-N-E-T-S, which is a open source thing on at least a free thing. 00:47:28.620 --> 00:47:31.980 I'm pretty sure it's open source for Jupyter running natively on iOS. 00:47:31.980 --> 00:47:32.720 Oh, interesting. 00:47:32.720 --> 00:47:34.260 I haven't looked at that. 00:47:34.260 --> 00:47:35.860 Yeah, it's pretty new. 00:47:35.860 --> 00:47:37.540 I just learned about it a few months ago. 00:47:37.540 --> 00:47:38.920 Oh, Stan, that's on GitHub. 00:47:38.920 --> 00:47:39.740 Yeah. 00:47:39.740 --> 00:47:43.220 We should add that to our analysis if we had time. 00:47:43.220 --> 00:47:44.320 That's pretty cool. 00:47:44.320 --> 00:47:49.320 It is, but I mean, it's not really fundamentally different, but it's interesting that it has NumPy. 00:47:49.320 --> 00:47:50.200 Right. 00:47:50.200 --> 00:47:51.280 It's just on the platform. 00:47:51.280 --> 00:47:51.560 Yeah. 00:47:51.560 --> 00:47:58.900 And a lot of those libraries like natively executed on iOS, not like an interface to some cloud thing, but like legitimately there, you know? 00:47:58.900 --> 00:47:59.240 Yeah. 00:47:59.240 --> 00:48:02.900 I mean, already people are talking about, you know, on our mobile phones, right? 00:48:02.900 --> 00:48:04.560 That, you know, for machine learning, right? 00:48:04.560 --> 00:48:09.740 You download these pre-trained models so then you can just do inference on your phone without going to the cloud, right? 00:48:09.740 --> 00:48:10.980 So both for privacy, right? 00:48:10.980 --> 00:48:13.560 If you want to do image recognition, both for privacy and for bandwidth. 00:48:13.560 --> 00:48:16.180 You can do all this computation on your phone locally. 00:48:16.180 --> 00:48:20.720 So you could imagine, you know, what would it mean if you could do data science out in the world, right? 00:48:20.720 --> 00:48:22.240 You have your smartwatch, you have your phone. 00:48:22.240 --> 00:48:25.840 You could just, you know, do your analysis, your Google Glass or whatever. 00:48:25.840 --> 00:48:27.820 There's this kind of sci-fi future. 00:48:27.820 --> 00:48:28.000 Yeah. 00:48:28.020 --> 00:48:35.980 I guess like, you know, the thought that we have as kind of human computer interaction or HCI researchers is, you know, what would the future of computing look like, right? 00:48:35.980 --> 00:48:40.120 Right now, it still looks like people hunched over on their laptops, typing in text boxes, right? 00:48:40.120 --> 00:48:42.220 Like, you want to go beyond that somehow. 00:48:42.220 --> 00:48:45.820 Still waiting to just plug an interface right into the neck, right? 00:48:45.820 --> 00:48:47.460 No more typing or massing. 00:48:47.460 --> 00:48:49.140 Projects onto your retina. 00:48:49.700 --> 00:48:51.900 You can have Jupyter in your eyes. 00:48:51.900 --> 00:48:52.480 Exactly. 00:48:52.480 --> 00:48:53.820 Sign me up. 00:48:53.820 --> 00:48:55.780 You'll be the first best subject. 00:48:55.780 --> 00:48:56.820 Yeah. 00:48:56.820 --> 00:49:00.320 Maybe not the first, but early adopter. 00:49:00.320 --> 00:49:01.280 Let's put me in that category. 00:49:01.280 --> 00:49:02.120 All right, you guys. 00:49:02.120 --> 00:49:03.140 This was super interesting. 00:49:03.140 --> 00:49:08.000 And it definitely opened my eyes to some things that are going on out there that I had no idea about. 00:49:08.000 --> 00:49:09.460 So super cool research. 00:49:09.460 --> 00:49:09.820 Thanks. 00:49:10.000 --> 00:49:11.360 It was really fun for us to talk about it. 00:49:11.360 --> 00:49:12.020 Yeah, for sure. 00:49:12.020 --> 00:49:17.440 Now, before you get out of here, I'm going to ask you the two questions that go at the end of the show all the time. 00:49:17.440 --> 00:49:19.840 And I'll, Philip, I'll pick on you since you're a veteran. 00:49:19.840 --> 00:49:23.000 If you're going to write some Python code, what editor do you use? 00:49:23.000 --> 00:49:26.200 I still use Vim because I haven't learned anything new. 00:49:26.200 --> 00:49:28.720 So I still use Vim with whatever defaults. 00:49:28.720 --> 00:49:29.740 Right on. 00:49:29.740 --> 00:49:30.160 All right. 00:49:30.160 --> 00:49:34.160 And then notable PyPI package, something that you've used that is interesting. 00:49:34.160 --> 00:49:36.000 I'll do a plug for Sam's actually. 00:49:36.140 --> 00:49:42.760 So Sam, for his master's thesis at Berkeley, he made this package called MB Interact, which is related to Jupyter Notebooks. 00:49:42.760 --> 00:49:44.000 It's called MB Interact. 00:49:44.000 --> 00:49:45.120 And you can pip install it. 00:49:45.120 --> 00:49:46.780 So I'll let Sam talk more about that. 00:49:46.780 --> 00:49:49.140 That's a lead into Sam's stuff. 00:49:49.140 --> 00:49:50.460 Thanks. 00:49:50.460 --> 00:49:50.980 Yeah, yeah. 00:49:50.980 --> 00:49:51.200 Cool. 00:49:51.200 --> 00:49:51.760 All right. 00:49:51.760 --> 00:49:53.360 Sam, if you're going to write some Python code editor? 00:49:53.360 --> 00:49:53.860 Yeah. 00:49:53.860 --> 00:49:54.980 So I use VS Code. 00:49:54.980 --> 00:49:59.880 But I've hopped from Sublime to Emacs to Vim and then not to VS Code. 00:49:59.880 --> 00:50:02.320 So now I'm a happy VS Code user. 00:50:02.320 --> 00:50:02.800 All right. 00:50:02.800 --> 00:50:03.020 Cool. 00:50:03.020 --> 00:50:04.460 Do you do any bindings? 00:50:04.620 --> 00:50:07.480 Like you set up Vim bindings or Emacs bindings in there or something? 00:50:07.480 --> 00:50:08.660 I set up Vim bindings. 00:50:08.660 --> 00:50:14.080 But if the VS Code team is listening, there are some problems I have with the Vim bindings in VS Code that I'd love to talk about. 00:50:14.080 --> 00:50:14.780 Oh, we're calling them out. 00:50:14.780 --> 00:50:16.560 This is the call-out section. 00:50:16.560 --> 00:50:18.160 This is where we start the beefs. 00:50:18.160 --> 00:50:20.440 The call-out section at the end of the episode. 00:50:20.440 --> 00:50:20.820 That's right. 00:50:20.820 --> 00:50:23.160 It's the underhanded compliment. 00:50:23.160 --> 00:50:24.320 Not exactly. 00:50:24.320 --> 00:50:26.400 But it's like, I love it, but beautiful. 00:50:26.400 --> 00:50:28.020 No, VS Code is a good one. 00:50:28.020 --> 00:50:28.600 All right. 00:50:28.600 --> 00:50:29.880 You want to tell us about MB Interact? 00:50:29.880 --> 00:50:30.320 Sure. 00:50:30.320 --> 00:50:34.580 MB Interact, it lets you take a Jupyter Notebook with some widgets. 00:50:35.400 --> 00:50:38.440 And normally those widgets don't work outside of the notebook environment. 00:50:38.440 --> 00:50:46.320 Like if you convert the notebook to a web page, to an HTML page by default, the widgets will just break because you don't have a Python service running underneath the notebook. 00:50:46.540 --> 00:51:00.600 But MB Interact allows, it kind of like takes your HTML page and hooks it up to the binder service from the Jupyter team so that you can have a, you can have like this HTML page with interactive widgets, but it doesn't require you to have notebook, a notebook server running locally. 00:51:00.600 --> 00:51:02.220 So you can kind of like share those things. 00:51:02.220 --> 00:51:05.760 You can kind of share your interactive notebooks more readily. 00:51:05.760 --> 00:51:06.540 That's the idea. 00:51:06.540 --> 00:51:06.860 I see. 00:51:06.860 --> 00:51:09.960 You don't have to set up a proper Jupyter server with all the execution and stuff. 00:51:10.040 --> 00:51:10.260 Yeah. 00:51:10.260 --> 00:51:14.060 You're going to send someone a link to a web page and then they can run your interactive widgets. 00:51:14.060 --> 00:51:15.360 That's the idea. 00:51:15.360 --> 00:51:15.620 Okay. 00:51:15.620 --> 00:51:16.280 Yeah. 00:51:16.280 --> 00:51:16.600 Yeah. 00:51:16.600 --> 00:51:17.460 Ah, excellent. 00:51:17.460 --> 00:51:20.280 As for like personal use, I like the TQDM package. 00:51:20.280 --> 00:51:23.600 It gives you like a progress bar when you run, when you run the four loops. 00:51:23.600 --> 00:51:24.460 Really handy. 00:51:24.460 --> 00:51:24.840 Yeah. 00:51:24.840 --> 00:51:25.060 Yeah. 00:51:25.060 --> 00:51:25.820 That's really cool. 00:51:25.820 --> 00:51:31.880 There's a couple of interesting progress bars that you can use, but it's those little touches, 00:51:31.880 --> 00:51:35.100 you know, color, progress bars, maybe a few emojis. 00:51:35.100 --> 00:51:38.000 It just makes it feel so much nicer to work in a CLI app. 00:51:38.000 --> 00:51:38.220 Yeah. 00:51:38.220 --> 00:51:41.540 Our next paper is analyzing 60 progress bar libraries. 00:51:41.540 --> 00:51:41.980 Yeah. 00:51:41.980 --> 00:51:44.100 60, 60 progress bar libraries. 00:51:44.100 --> 00:51:44.720 And emojis. 00:51:44.720 --> 00:51:46.100 What emojis they use. 00:51:46.100 --> 00:51:46.900 And emojis. 00:51:46.900 --> 00:51:47.340 Yeah. 00:51:47.340 --> 00:51:47.920 Perfect. 00:51:47.920 --> 00:51:48.300 Perfect. 00:51:48.300 --> 00:51:49.300 All right. 00:51:49.300 --> 00:51:50.520 Well, you guys, this is really fun. 00:51:50.520 --> 00:51:52.880 People want to know more about your research, your paper. 00:51:52.880 --> 00:51:53.560 What do they do? 00:51:53.560 --> 00:51:54.240 How do they find it? 00:51:54.400 --> 00:51:56.180 I mean, I'll put a link in the show notes, but what else? 00:51:56.180 --> 00:51:56.440 Yeah. 00:51:56.440 --> 00:51:57.520 They can just go to the links. 00:51:57.520 --> 00:52:00.220 Michael, I sent you a link in the chat about MB Interact. 00:52:00.220 --> 00:52:03.140 And then we sent you the paper links and such. 00:52:03.140 --> 00:52:06.400 We can put both of our homepages on there and such. 00:52:06.400 --> 00:52:07.960 Sam, any other plugs? 00:52:07.960 --> 00:52:08.760 Yeah. 00:52:08.760 --> 00:52:13.340 I think you remind me that I haven't put up the paper online. 00:52:13.340 --> 00:52:14.940 I don't know if you have the paper online anywhere. 00:52:14.940 --> 00:52:15.240 Yeah. 00:52:15.240 --> 00:52:16.140 I have it on my page. 00:52:16.140 --> 00:52:16.280 Yeah. 00:52:16.280 --> 00:52:16.740 Oh, you do. 00:52:16.740 --> 00:52:17.120 Okay. 00:52:17.120 --> 00:52:19.320 So we can maybe link to the paper from Philip's page. 00:52:19.320 --> 00:52:19.640 Yep. 00:52:19.640 --> 00:52:20.140 Sounds good. 00:52:20.140 --> 00:52:21.240 Well, I'll definitely do that. 00:52:21.240 --> 00:52:22.120 All right, you guys. 00:52:22.120 --> 00:52:23.420 Thank you so much for being on the show. 00:52:23.640 --> 00:52:24.800 It's been really interesting. 00:52:24.800 --> 00:52:30.760 And yeah, there's a lot of notebook exploration and flowers blooming. 00:52:30.760 --> 00:52:34.480 So thanks for putting it all together into like one analysis here. 00:52:34.480 --> 00:52:34.780 Yeah. 00:52:34.780 --> 00:52:37.180 Really big thanks to you, Michael, for having us on. 00:52:37.180 --> 00:52:41.300 And it's been really fun to talk about all the flowers, all the fields of flowers. 00:52:42.240 --> 00:52:43.560 Likewise. 00:52:43.560 --> 00:52:44.840 Thank you so much, Michael. 00:52:44.840 --> 00:52:45.500 Have a good day. 00:52:45.500 --> 00:52:45.860 See you later. 00:52:45.860 --> 00:52:46.860 Bye. 00:52:46.860 --> 00:52:49.700 This has been another episode of Talk Python To Me. 00:52:49.700 --> 00:52:52.700 Our guests on this episode were Sam Lau and Philip Guo. 00:52:53.180 --> 00:52:55.420 And it's been brought to you by Linode and Sentry. 00:52:55.420 --> 00:53:00.340 Start your next Python project on Linode's state-of-the-art cloud service. 00:53:00.340 --> 00:53:03.120 Just visit talkpython.fm/Linode. 00:53:03.120 --> 00:53:04.640 L-I-N-O-D-E. 00:53:04.640 --> 00:53:07.960 You'll automatically get a $20 credit when you create a new account. 00:53:07.960 --> 00:53:10.640 Take some stress out of your life. 00:53:10.640 --> 00:53:15.240 Get notified immediately about errors in your web applications with Sentry. 00:53:15.660 --> 00:53:19.320 Just visit talkpython.fm/sentry and get started for free. 00:53:19.320 --> 00:53:21.480 Want to level up your Python? 00:53:21.480 --> 00:53:26.340 If you're just getting started, try my Python Jumpstart by Building 10 Apps course. 00:53:26.340 --> 00:53:31.440 Or if you're looking for something more advanced, check out our new async course that digs into 00:53:31.440 --> 00:53:34.500 all the different types of async programming you can do in Python. 00:53:34.500 --> 00:53:38.460 And of course, if you're interested in more than one of these, be sure to check out our 00:53:38.460 --> 00:53:39.160 Everything Bundle. 00:53:39.160 --> 00:53:41.040 It's like a subscription that never expires. 00:53:41.040 --> 00:53:43.200 Be sure to subscribe to the show. 00:53:43.200 --> 00:53:45.700 Open your favorite podcatcher and search for Python. 00:53:45.700 --> 00:53:46.840 We should be right at the top. 00:53:46.840 --> 00:53:51.660 You can also find the iTunes feed at /itunes, the Google Play feed at /play, 00:53:51.660 --> 00:53:55.820 and the direct RSS feed at /rss on talkpython.fm. 00:53:55.820 --> 00:53:57.900 This is your host, Michael Kennedy. 00:53:57.900 --> 00:53:59.400 Thanks so much for listening. 00:53:59.400 --> 00:54:00.440 I really appreciate it. 00:54:00.440 --> 00:54:02.200 Now get out there and write some Python code. 00:54:02.200 --> 00:54:22.960 I'll see you next time.