WEBVTT 00:00:00.001 --> 00:00:04.740 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:04.740 --> 00:00:08.960 This is episode 211, recorded December 2nd, 2020. 00:00:08.960 --> 00:00:10.180 I'm Michael Kennedy. 00:00:10.180 --> 00:00:11.340 And I'm Brian Okken. 00:00:11.340 --> 00:00:12.620 And we have a special guest. 00:00:12.620 --> 00:00:13.160 Yeah. 00:00:13.160 --> 00:00:14.020 Matthew Feigert, welcome. 00:00:14.020 --> 00:00:15.660 Yeah, thanks so much for having me on. 00:00:15.660 --> 00:00:17.020 Yeah, it's great to have you here. 00:00:17.020 --> 00:00:20.160 You've been over on Talk Python before, right? 00:00:20.160 --> 00:00:20.600 Yeah. 00:00:20.600 --> 00:00:24.980 Talking about some cool high energy physics and all that kind of stuff. 00:00:24.980 --> 00:00:28.400 Yeah, I looked that up last night just to try and remind myself. 00:00:28.400 --> 00:00:29.980 That was episode 144. 00:00:29.980 --> 00:00:37.800 I was on with my colleagues, Michaela Paganini and Michael Kagan to talk with you about machine learning applications at the LHC. 00:00:37.800 --> 00:00:38.280 Yeah. 00:00:38.280 --> 00:00:42.120 And you do stuff over with CERN at the Large Hadron Collider and things like that? 00:00:42.120 --> 00:00:42.640 Yeah. 00:00:42.640 --> 00:00:42.980 Yeah. 00:00:42.980 --> 00:00:47.160 So I'm a postdoctoral researcher at the University of Illinois at Urbana-Champaign. 00:00:47.160 --> 00:00:57.400 And so there I split my time between working on the Atlas experiment and working as a software researcher at the Institute for Research and Innovation and Software for High Energy Physics. 00:00:57.800 --> 00:01:07.200 And so like on Atlas, Atlas is this like huge five story tall particle detector that lives 100 meters underground at CERN's Large Hadron Collider. 00:01:07.200 --> 00:01:10.320 That's just outside beautiful Geneva, Switzerland. 00:01:11.100 --> 00:01:21.700 And so there I work with a few thousand of my closest colleagues and friends to try and look for evidence of new physics and make precision measurements of physics we do know about. 00:01:22.500 --> 00:01:37.800 And then my IRISHEP work is kind of focused on working in an interdisciplinary and inter-experimental team to try and improve the necessary cyber infrastructure and software for us to be able to use in upcoming runs of the Large Hadron Collider. 00:01:37.800 --> 00:01:41.900 And in what we call like a high luminosity run, which is going to be way more collisions than normal. 00:01:41.900 --> 00:01:43.480 Have you guys ever turned it up to full power? 00:01:43.480 --> 00:01:45.400 Have you turned it up to full power yet? 00:01:45.900 --> 00:01:53.620 So, yeah, the design luminosity or the design energy of the LHC is at something called 14 tera electron volts, 14 TV. 00:01:53.620 --> 00:02:00.140 And we've been running intentionally at a lower operating energy for the last couple of years, just a little bit below that. 00:02:00.140 --> 00:02:03.740 But in the late 2020s, we're going to suck the entire Earth into it. 00:02:05.300 --> 00:02:08.460 You know, no, no experimental evidence of black hole creation. 00:02:08.460 --> 00:02:08.800 Yeah. 00:02:08.800 --> 00:02:18.860 But kind of the cool thing is that if we even did make a black hole, the LHC due to something called Hawking radiation, it would evaporate well before it could actually ever do anything interesting gravitationally. 00:02:18.860 --> 00:02:20.600 But yeah, it's really exciting. 00:02:20.600 --> 00:02:21.100 Really. 00:02:21.100 --> 00:02:23.680 I'm joking, but it's such a cool place. 00:02:23.680 --> 00:02:25.060 Such cool technology. 00:02:25.060 --> 00:02:27.520 I mean, that's right out of the edge of physics these days. 00:02:27.520 --> 00:02:29.640 And the technology side is neat, too. 00:02:29.640 --> 00:02:30.140 Yeah. 00:02:30.140 --> 00:02:31.060 No, it's super fun. 00:02:31.060 --> 00:02:34.020 Well, welcome over to Python Bytes. 00:02:34.020 --> 00:02:35.100 Yeah, it's great to be here. 00:02:35.100 --> 00:02:36.820 So I yeah, it's good to have you. 00:02:36.820 --> 00:02:37.440 Thanks for coming. 00:02:37.440 --> 00:02:41.640 And Brian, I think let's let's start with another one of my favorite topics. 00:02:41.640 --> 00:02:42.320 Farms. 00:02:42.320 --> 00:02:43.580 I love farming. 00:02:43.580 --> 00:02:45.400 You know, you see the bumper sticker. 00:02:45.400 --> 00:02:46.860 No farms, no food. 00:02:46.860 --> 00:02:48.020 I like food a lot. 00:02:48.020 --> 00:02:49.160 So I love farms. 00:02:49.160 --> 00:02:50.220 No, no. 00:02:50.220 --> 00:02:51.800 But the farm stack. 00:02:51.800 --> 00:02:54.420 We've heard the lamp stack, other stacks. 00:02:54.420 --> 00:02:56.680 Like lamp is not as useful as farm. 00:02:56.680 --> 00:02:57.100 Right. 00:02:57.100 --> 00:02:57.960 Farm sounds more useful. 00:02:57.960 --> 00:02:59.180 So tell us about farm. 00:02:59.180 --> 00:03:02.440 So Aaron Bassett, he's I'm not sure. 00:03:02.680 --> 00:03:08.960 I think he's one of the spokespeople for Mongo or something like advocate or something like 00:03:08.960 --> 00:03:09.120 that. 00:03:09.120 --> 00:03:15.240 Anyway, he's he's doing he he wrote this article, but they've also done. 00:03:15.500 --> 00:03:18.100 I think that there's been some talks given. 00:03:18.100 --> 00:03:19.960 But this is this is a nice article. 00:03:19.960 --> 00:03:24.520 It's a called introducing farm stack, which is FastAPI react and MongoDB. 00:03:25.220 --> 00:03:34.540 So the I really actually appreciated the article and the code with it, because there's a GitHub, 00:03:34.540 --> 00:03:37.740 a little GitHub to do credit app that they've put together. 00:03:37.740 --> 00:03:45.800 And it the article describes basically all of the pieces of the application using it like 00:03:45.800 --> 00:03:46.700 a little to do app. 00:03:46.700 --> 00:03:53.840 But with FastAPI, you've got this is interactive, interactive documentation mode where you can 00:03:53.840 --> 00:03:57.160 interact with the application just almost immediately. 00:03:57.160 --> 00:04:00.080 You don't have to really do much to put it all together. 00:04:00.080 --> 00:04:05.740 And and then for all your endpoints, you can actually interact with them, send send data, 00:04:05.740 --> 00:04:06.420 do queries. 00:04:06.820 --> 00:04:09.880 And there's a little animated gif to show how that's done. 00:04:09.880 --> 00:04:15.460 But the article then, you know, goes through and says, you know, basically how how the endpoints 00:04:15.460 --> 00:04:23.440 and routes get hooked up and then uses you, you, you, a corn to set up an async event loop and 00:04:23.440 --> 00:04:30.480 get that going shows how easy it is to connect to a database and then defining models with and 00:04:30.480 --> 00:04:33.640 how easy it is to set up a schema goes through. 00:04:34.120 --> 00:04:39.660 And then it kind of hooks up, talks through the code discussion of you do have to write 00:04:39.660 --> 00:04:44.860 code for the endpoints and and really how easy those are with all of these these pieces. 00:04:44.860 --> 00:04:49.820 The React application is it's kind of a minimal app React application. 00:04:49.820 --> 00:04:53.480 I'm not sure why they kind of included that, but it's kind of a neat addition. 00:04:53.480 --> 00:04:59.000 There's a React application that's running that just sort of shows some of the interaction with 00:04:59.000 --> 00:05:05.840 the the CRUD app and it gets updated while, you know, while you're changing things through 00:05:05.840 --> 00:05:07.260 the interactive API. 00:05:07.260 --> 00:05:14.120 And I just I like the demonstration of working through working with an API and working through 00:05:14.120 --> 00:05:20.820 changing things and seeing it show up having a like a React app at the other end. 00:05:20.820 --> 00:05:24.040 It's kind of a fun way to kind of experiment with an API. 00:05:24.200 --> 00:05:26.140 This is a really neat thing. 00:05:26.140 --> 00:05:29.680 And one of the other major stacks that's been used around Mongo is the mean stack. 00:05:29.680 --> 00:05:34.080 And the farm stack is way nicer than the mean stack, not just because it uses Python and not 00:05:34.080 --> 00:05:34.380 JavaScript. 00:05:34.380 --> 00:05:36.180 But there's some interesting things here. 00:05:36.180 --> 00:05:43.180 One of the examples is actually kind of blowing my mind in that it's an if statement using the 00:05:43.180 --> 00:05:48.080 walrus operator awaiting an asynchronous call in an API method. 00:05:48.080 --> 00:05:51.960 like the walrus operator and async, the await keyword. 00:05:51.960 --> 00:05:52.940 I've never seen those together. 00:05:52.940 --> 00:05:55.180 And it's kind of like it's inspiring. 00:05:55.180 --> 00:05:56.040 It's nice. 00:05:56.040 --> 00:05:56.660 It's good. 00:05:56.660 --> 00:05:57.100 Yeah. 00:05:57.100 --> 00:05:57.520 Yeah. 00:05:57.520 --> 00:05:59.360 Such succinct code as well. 00:05:59.360 --> 00:06:00.540 It's super nice. 00:06:00.540 --> 00:06:02.640 I mean, it uses FastAPI, which is fantastic. 00:06:02.640 --> 00:06:09.720 It's using motor, which is MongoDB's officially supported Python async library because you need an 00:06:09.720 --> 00:06:13.980 async capable library in order to do things against MongoDB. 00:06:14.640 --> 00:06:17.280 You know, this actually comes from the developer blog at MongoDB. 00:06:17.280 --> 00:06:24.280 There also are some ORM like things, some ODMs, object document mapper stuff that also supports 00:06:24.280 --> 00:06:25.820 async and await from MongoDB. 00:06:25.820 --> 00:06:28.500 So if you're more in the ORM style, you might check that out. 00:06:28.500 --> 00:06:30.340 But other than that, this looks pretty neat to me. 00:06:30.340 --> 00:06:30.480 Yeah. 00:06:30.480 --> 00:06:30.980 Yeah. 00:06:30.980 --> 00:06:35.660 And I do know that a lot of people use the ORMs, but I like I appreciated the example without 00:06:35.660 --> 00:06:41.760 an ORM for people because you throw an ORM example in there and then people that don't 00:06:41.760 --> 00:06:43.740 use that particular one get lost. 00:06:44.140 --> 00:06:44.580 Yeah. 00:06:44.580 --> 00:06:46.540 Matthew, do you guys do anything with MongoDB? 00:06:46.540 --> 00:06:47.780 Any of these kind of things? 00:06:47.780 --> 00:06:48.260 FastAPI? 00:06:48.260 --> 00:06:49.280 Yeah. 00:06:49.280 --> 00:06:50.420 I have some friends that do. 00:06:50.420 --> 00:06:56.820 I personally, myself, I'm not too versed in Mongo, but I've heard it on the show and 00:06:56.820 --> 00:06:58.320 many, many times elsewhere. 00:06:58.320 --> 00:07:03.640 So this is, I think, also just kind of paging through the article as Ryan was talking about. 00:07:03.640 --> 00:07:04.620 It is pretty impressive. 00:07:04.620 --> 00:07:06.780 So it's really concise. 00:07:06.780 --> 00:07:10.040 Like, here's your four lines to completely implement the API. 00:07:10.040 --> 00:07:10.540 Yeah. 00:07:10.540 --> 00:07:11.560 Type of things, right? 00:07:11.560 --> 00:07:13.680 Asynchronous, fast, like all the cool stuff. 00:07:13.760 --> 00:07:13.920 Yeah. 00:07:13.920 --> 00:07:14.760 Yeah. 00:07:14.760 --> 00:07:20.540 There was an example, a case study of MongoDB being used at the Large Hadron Collider, but 00:07:20.540 --> 00:07:22.720 that was many years ago and I don't know if it still is. 00:07:22.720 --> 00:07:24.900 So I've completely forgotten where that is. 00:07:24.900 --> 00:07:25.320 Yeah. 00:07:25.320 --> 00:07:25.960 I don't know. 00:07:25.960 --> 00:07:26.620 Yeah. 00:07:26.620 --> 00:07:27.060 Yeah. 00:07:27.060 --> 00:07:27.540 Cool. 00:07:27.540 --> 00:07:27.780 Cool. 00:07:27.880 --> 00:07:31.840 So next thing I want to talk about another programming language. 00:07:31.840 --> 00:07:37.400 Last time, Brian, I went on and on, maybe the time before, two times ago, about .NET and C-Sharp 00:07:37.400 --> 00:07:41.920 because Anthony Shaw had done that work on Pigeon to get Python to run on .NET. 00:07:41.920 --> 00:07:45.900 And we're like, well, why are we talking about C-Sharp on this project, right? 00:07:45.900 --> 00:07:46.680 On this podcast. 00:07:47.100 --> 00:07:49.680 Well, I want to talk about something even more advanced, AppleScript. 00:07:49.680 --> 00:07:52.040 Wow. 00:07:52.040 --> 00:07:52.660 Cutting edge. 00:07:52.660 --> 00:07:53.280 Yes. 00:07:53.280 --> 00:07:57.280 It's like the CMD shell script of Apple. 00:07:57.280 --> 00:07:59.520 Have you ever programmed an AppleScript? 00:07:59.520 --> 00:08:00.200 It's painful. 00:08:00.620 --> 00:08:01.240 No, I have not. 00:08:01.240 --> 00:08:06.500 It's like you say, like, tell this application that to like make a command. 00:08:06.500 --> 00:08:08.180 Oh, it's bad news bears. 00:08:08.180 --> 00:08:08.700 Let me tell you. 00:08:08.700 --> 00:08:15.880 So what I've come across is this thing called PyAppleScript. 00:08:15.880 --> 00:08:18.840 Now, this is not brand new, but it's brand new to me. 00:08:18.840 --> 00:08:23.140 And there's a lot of talk about Macs and people may be getting new Macs. 00:08:23.140 --> 00:08:27.780 So I thought I would say, hey, look, here's a cool way to automate your Mac or, you know, 00:08:27.780 --> 00:08:29.380 Macs within your company or whatever. 00:08:29.700 --> 00:08:32.900 With Python instead of this dreaded NS AppleScript. 00:08:32.900 --> 00:08:33.520 Okay. 00:08:33.520 --> 00:08:33.780 All right. 00:08:33.780 --> 00:08:39.620 So basically it's a Python wrapper around NS AppleScript, allowing Python scripts or applications 00:08:39.620 --> 00:08:44.580 to communicate with AppleScript and AppleScriptable applications. 00:08:44.580 --> 00:08:48.880 So apps for which they basically implement AppleScript and let you do that. 00:08:48.880 --> 00:08:52.800 So scripts get compiled either from source or they can be loaded from disk. 00:08:52.800 --> 00:08:57.740 They have these, some of these ideas are from AppleScript as a standard run handler and user 00:08:57.740 --> 00:09:00.820 defined handlers can be invoked with or without arguments. 00:09:00.820 --> 00:09:02.440 They're automatically converted. 00:09:02.660 --> 00:09:08.060 The responses to and from AppleScript are automatically converted either from AppleScript to Python 00:09:08.060 --> 00:09:12.400 types like Python string versus AppleScript one or vice versa. 00:09:12.400 --> 00:09:12.680 Right. 00:09:12.680 --> 00:09:14.940 So you don't have to do the type coercion, which is cool. 00:09:14.940 --> 00:09:16.320 And they're persistent. 00:09:16.320 --> 00:09:20.920 So you can call your handle multiple times and it retains a state like AppleScript would. 00:09:20.920 --> 00:09:28.800 And it also has no dependency on the legacy AppleScript library or the so-called flawed scripting 00:09:28.800 --> 00:09:32.640 bridge framework, which is limited to OSA script executables. 00:09:32.840 --> 00:09:33.600 So that's pretty cool. 00:09:33.600 --> 00:09:38.900 If you want to automate things on your Mac, you obviously could use Bash. 00:09:38.900 --> 00:09:43.840 But if you're talking to some kind of application that implements one of these scripts, like for 00:09:43.840 --> 00:09:48.780 example, you want to tell this other application to grab something out of the clipboard and then 00:09:48.780 --> 00:09:50.840 tell it to do something or something like that. 00:09:50.840 --> 00:09:51.040 Right. 00:09:51.040 --> 00:09:52.860 Like you couldn't reasonably do that with Bash. 00:09:52.860 --> 00:09:53.260 Right. 00:09:53.260 --> 00:09:56.500 Once it starts up, you kind of want to go back and forth with it. 00:09:56.500 --> 00:09:58.800 So it sounds like AppleScript might be the thing to do. 00:09:58.800 --> 00:09:59.300 Pretty cool. 00:09:59.500 --> 00:09:59.940 Yeah. 00:09:59.940 --> 00:10:00.500 Yeah. 00:10:00.500 --> 00:10:00.980 Yeah. 00:10:00.980 --> 00:10:02.400 I mean, not a lot to it. 00:10:02.400 --> 00:10:07.760 Like if you've got a script, your Apple macOS stuff, do it with Python. 00:10:07.760 --> 00:10:10.260 You don't have to do it with that AppleScript stuff. 00:10:10.260 --> 00:10:11.060 No, it's neat. 00:10:11.060 --> 00:10:11.460 Yeah. 00:10:11.460 --> 00:10:11.880 Yeah. 00:10:11.880 --> 00:10:16.380 So Matthew, you probably brought something to do with physics, data science, I'm guessing. 00:10:16.380 --> 00:10:17.380 What's your first one here? 00:10:17.380 --> 00:10:18.160 Yeah, a bit. 00:10:18.160 --> 00:10:24.780 So we currently live in this like really nice age of having awesome CI services and all these 00:10:24.780 --> 00:10:27.240 really nice metrics for all your GitHub projects and everything. 00:10:27.240 --> 00:10:30.920 So, you know, if you're, I'm thinking of like coverage. 00:10:30.920 --> 00:10:35.600 So if you're, you know, using pytest and, you know, making sure that you're reporting your 00:10:35.600 --> 00:10:39.880 coverage, you have all these really great services to also track your coverage and report that 00:10:39.880 --> 00:10:40.880 in a nice shiny badge. 00:10:41.120 --> 00:10:46.540 But let's say you're developing some tool or some library and you have some sort of performance 00:10:46.540 --> 00:10:48.380 metric that you care about. 00:10:48.380 --> 00:10:53.500 Let's say like how fast some, the speed of evaluation for certain expensive functions. 00:10:53.500 --> 00:10:57.040 And you actually want to try and like track that through the entire history of your code 00:10:57.040 --> 00:10:57.320 base. 00:10:57.320 --> 00:11:00.680 And that's not something that's like traditionally very super easy to do. 00:11:01.040 --> 00:11:03.220 So recently I was really happy to find something. 00:11:03.220 --> 00:11:04.500 So like if you're making changes. 00:11:04.500 --> 00:11:09.160 So if you're going to be adding some feature or whatever, you are refactoring it. 00:11:09.160 --> 00:11:12.160 So it's easier to write, but you're not sure if that makes it faster or slower. 00:11:12.160 --> 00:11:16.120 And this would sort of give you that information from week to week or something like that. 00:11:16.120 --> 00:11:16.680 Exactly. 00:11:16.680 --> 00:11:16.980 Yeah. 00:11:17.060 --> 00:11:21.220 So you might like, you might go ahead and say like, okay, well, you know, I have like 00:11:21.220 --> 00:11:25.860 some, some tests that make sure that this function evaluates and under some period of time, if 00:11:25.860 --> 00:11:28.560 it's an expensive function for your test. 00:11:28.560 --> 00:11:33.700 But let's say you actually want to like track across like different, different parameter, 00:11:33.700 --> 00:11:39.060 parameterizations, how that function actually is about is being performing and evaluating it 00:11:39.060 --> 00:11:39.940 in your whole code base. 00:11:39.940 --> 00:11:45.580 So I've recently found this super cool tool written in Python called airspeed velocity. 00:11:46.600 --> 00:11:53.040 And so from the docs, ASV airspeed velocity is a tool for benchmarking Python packages over 00:11:53.040 --> 00:11:53.700 their lifetime. 00:11:53.700 --> 00:11:59.180 So it deals with runtime memory consumption and even custom compute values. 00:11:59.180 --> 00:12:05.320 And the results are then displayed in a super nice web front end that's interactive and basically 00:12:05.320 --> 00:12:07.560 just requires like static webpage hosting. 00:12:07.560 --> 00:12:10.760 So it's, it's pretty impressive. 00:12:10.760 --> 00:12:15.120 And just if you click on the docs, you can see that's developed by a community of people, 00:12:15.120 --> 00:12:18.800 but led by Michael Dorritboom. 00:12:18.800 --> 00:12:20.800 I'm probably getting your, your name wrong. 00:12:20.800 --> 00:12:21.500 Very sorry. 00:12:21.500 --> 00:12:23.900 And Pauly Burton. 00:12:23.900 --> 00:12:25.860 But if you look at some... 00:12:25.860 --> 00:12:29.560 He's the guy that who was behind a pie oxidizer at Mozilla. 00:12:29.560 --> 00:12:30.060 Oh, really? 00:12:30.060 --> 00:12:30.920 Oh, okay. 00:12:30.920 --> 00:12:31.260 Yeah. 00:12:31.260 --> 00:12:32.780 That's a super cool project. 00:12:32.780 --> 00:12:34.120 Yeah, for sure. 00:12:34.240 --> 00:12:34.680 Yeah. 00:12:34.680 --> 00:12:34.800 Yeah. 00:12:34.800 --> 00:12:35.640 yeah. 00:12:35.640 --> 00:12:40.240 And so, I mean, if you look at the other people that are on the contributor list, you can spot 00:12:40.240 --> 00:12:44.480 a lot of names that are common in the SciPy and Jupyter ecosystem. 00:12:44.480 --> 00:12:48.980 So it's, you already know that this is a nice community built tool. 00:12:48.980 --> 00:12:54.320 And then also, as kind of some example cases, they give, current projects that 00:12:54.320 --> 00:12:56.840 are using it like NumPy and SciPy and AstroPy. 00:12:56.840 --> 00:12:59.300 So pretty well established projects. 00:12:59.300 --> 00:13:03.940 And just as kind of like an example, if you click on like the SciPy project and go 00:13:03.940 --> 00:13:08.380 to the interpolate function there, you can, you can just kind of look at a very nice 00:13:08.380 --> 00:13:14.320 visualization of the actual evaluation, in time on the vertical axis across a whole 00:13:14.320 --> 00:13:19.460 bunch of parameterizations, such as like CPython version and number of samples that are being 00:13:19.460 --> 00:13:19.800 run. 00:13:19.800 --> 00:13:24.460 And you can see this for the entire lifetime of the code base and you can zoom in on any 00:13:24.460 --> 00:13:25.640 section just with the mouse. 00:13:25.820 --> 00:13:30.040 And something I think that is super, super cool is if you, if you're looking at the visualization 00:13:30.040 --> 00:13:34.420 of the plot and you see that, oh, there's like one commit where all of a sudden things go 00:13:34.420 --> 00:13:36.700 funky and the evaluation time just jumps up. 00:13:36.700 --> 00:13:42.100 You can just click on that node and it immediately opens up to that commit in GitHub, which is, 00:13:42.100 --> 00:13:45.880 I think, super awesome that you don't have to go and like search through your commit history 00:13:45.880 --> 00:13:48.280 to figure out what, like where that corresponds to. 00:13:48.280 --> 00:13:49.240 It's just boom, right? 00:13:49.240 --> 00:13:52.160 I'm looking at, it shows the Shaw from GitHub. 00:13:52.160 --> 00:13:52.760 Yeah. 00:13:53.100 --> 00:13:56.240 The, the unique identifier of the commit. 00:13:56.240 --> 00:13:57.560 That's crazy. 00:13:57.560 --> 00:13:57.920 Yeah. 00:13:57.920 --> 00:13:58.880 So, wow. 00:13:58.880 --> 00:13:59.300 Yeah. 00:13:59.300 --> 00:14:03.740 So I've, I've, you know, a project that I'm working on, we've been interested in trying 00:14:03.740 --> 00:14:07.200 to have the sort of like metric tracking for some of our, for some of our work. 00:14:07.200 --> 00:14:11.720 So this is something that I'm actively kind of, looking at how we might be able to deploy 00:14:11.720 --> 00:14:14.120 this for one of my projects with my coauthors. 00:14:14.240 --> 00:14:16.440 but it's openly developed on GitHub. 00:14:16.440 --> 00:14:19.020 It's up on pipe, pipe behind as well. 00:14:19.020 --> 00:14:20.660 So just pip install ASV. 00:14:20.660 --> 00:14:25.400 and then I think something that's kind of very cute and very kind of Pythonic is that, 00:14:25.400 --> 00:14:29.680 uh, if you, when you go to the reporting dashboard for the different libraries that you're actually 00:14:29.680 --> 00:14:35.540 benchmarking, it will, up at the top, say the airspeed velocity of an unladen X. 00:14:35.540 --> 00:14:40.460 So the airspeed velocity of an unladen like numpy or an unladen scipy. 00:14:40.460 --> 00:14:44.380 So, you know, keeping very true to the, you know, Python's roots there. 00:14:44.380 --> 00:14:50.120 There's some Monty Python, the show Zen in there for sure. 00:14:50.120 --> 00:14:50.620 Exactly. 00:14:50.620 --> 00:14:51.080 Yeah. 00:14:51.080 --> 00:14:52.240 This is impressive. 00:14:52.240 --> 00:14:55.020 I mean, Brian, how do you see this fitting into like testing and stuff? 00:14:55.020 --> 00:14:56.300 I actually love this. 00:14:56.300 --> 00:14:57.960 I, I could use this right away. 00:14:57.960 --> 00:15:02.020 There's, there's lots of, well, a lot of times it's, it's not, yeah. 00:15:02.020 --> 00:15:05.240 Performance of performance is always something we care about. 00:15:05.240 --> 00:15:11.300 And, and benchmarking systems, and, you know, testing, it's always, it's something 00:15:11.300 --> 00:15:17.380 you forget about sometimes, like running, running stuff and it still works, but like 00:15:17.380 --> 00:15:21.060 over time things slow down and it's good to, good to know that. 00:15:21.060 --> 00:15:21.580 Yeah. 00:15:21.580 --> 00:15:26.340 And if this could just be automatic and just part of your CI, you just go back and see the 00:15:26.340 --> 00:15:26.700 updates. 00:15:26.700 --> 00:15:27.640 That'd be very cool. 00:15:27.640 --> 00:15:28.260 Definitely. 00:15:28.260 --> 00:15:28.840 Yeah. 00:15:28.840 --> 00:15:33.400 I don't think that this is something that at the moment, and I'm happy to be corrected about 00:15:33.400 --> 00:15:33.620 this. 00:15:33.620 --> 00:15:37.960 I don't think at the moment there is some way that this is currently being, given 00:15:37.960 --> 00:15:41.860 as like a CI service, but I think that this is something that you could like set up and 00:15:41.860 --> 00:15:43.260 run for yourself pretty easily. 00:15:43.260 --> 00:15:43.660 Yeah. 00:15:43.660 --> 00:15:45.220 You could probably plug it in. 00:15:45.220 --> 00:15:45.720 Yeah. 00:15:45.720 --> 00:15:46.640 Yeah, exactly. 00:15:46.640 --> 00:15:53.000 But you could probably do some kind of webhook when, when there's a check-in automatically 00:15:53.000 --> 00:15:54.580 kick it off and then save a result. 00:15:54.580 --> 00:15:54.840 Right. 00:15:54.860 --> 00:15:59.340 You could just hook into the GitHub actions and then have it just call you back and start 00:15:59.340 --> 00:16:02.420 your, you know, let's take a, take a record of this or whatever. 00:16:02.420 --> 00:16:03.060 Yeah. 00:16:03.060 --> 00:16:03.360 Yeah. 00:16:03.360 --> 00:16:03.760 Very cool. 00:16:03.760 --> 00:16:04.720 This is a great idea. 00:16:04.720 --> 00:16:05.260 Yeah. 00:16:05.260 --> 00:16:09.800 Something else that I'm, I haven't really investigated yet, but that I'm looking into is if this can 00:16:09.800 --> 00:16:12.380 also be used to do like GPU benchmarking. 00:16:12.380 --> 00:16:16.960 So like, let's say you have a library that, you know, also that is going to be, you 00:16:16.960 --> 00:16:21.420 can transparently, use the APIs to transparently move from CPU to GPU. 00:16:21.420 --> 00:16:26.480 Like you have something like Jack's or TensorFlow or PyTorch, then this might be kind of a nice 00:16:26.480 --> 00:16:26.820 way. 00:16:26.820 --> 00:16:31.280 if it's, if it's based on those to be able to like benchmark your GPU performance 00:16:31.280 --> 00:16:31.700 as well. 00:16:31.700 --> 00:16:32.160 Yeah. 00:16:32.460 --> 00:16:34.460 Well, and that's one of the things you might not test, right? 00:16:34.460 --> 00:16:38.580 If it could run either way, you might just run it on your machine, whichever one of those 00:16:38.580 --> 00:16:40.560 it is and forget to try the other style. 00:16:40.560 --> 00:16:40.840 Right. 00:16:40.840 --> 00:16:41.500 Exactly. 00:16:41.500 --> 00:16:41.940 Yeah. 00:16:41.940 --> 00:16:46.020 And I don't think there's too many CI services that are going to, you know, generously give 00:16:46.020 --> 00:16:48.980 you some like really nice GPUs to be doing benchmarking on. 00:16:48.980 --> 00:16:50.320 Yeah, that's for sure. 00:16:50.320 --> 00:16:50.880 For sure. 00:16:50.880 --> 00:16:51.520 All right. 00:16:51.520 --> 00:16:54.820 Now for the next item, let me tell you about our sponsor. 00:16:54.820 --> 00:16:59.520 This episode is brought to you by Tech Meme, the Tech Meme Ride Home Podcast. 00:17:00.240 --> 00:17:05.460 They've been for two years recording episodes every single day. 00:17:05.460 --> 00:17:11.300 And so they're Silicon Valley's favorite tech news podcast, and you can get them daily 15 00:17:11.300 --> 00:17:15.380 to 20 minutes, exactly by 5 PM Eastern, all the tech news you want. 00:17:15.380 --> 00:17:15.720 Right. 00:17:15.720 --> 00:17:17.960 But it's not just headlines, much like by them bites. 00:17:17.960 --> 00:17:21.560 Actually, it's a very similar show, but for the broader tech industry, you could have a robot 00:17:21.560 --> 00:17:26.120 read the headlines or just flip through them, but it has the context and the analysis all 00:17:26.120 --> 00:17:26.560 around it. 00:17:26.560 --> 00:17:29.520 So it's like tech news as a service, if you will. 00:17:29.680 --> 00:17:35.020 So the folks over at Tech Meme, they're online all day reading to catch you up and just search 00:17:35.020 --> 00:17:42.220 your podcast app for the Ride Home and subscribe to the Tech Meme Ride Home Podcast or just visit 00:17:42.220 --> 00:17:45.160 pythonbytes.fm/ride to subscribe. 00:17:45.160 --> 00:17:48.200 I have a theory, a hypothesis about this. 00:17:48.200 --> 00:17:53.820 I think that probably actually be a ton of work to put together a show daily on a time like 00:17:53.820 --> 00:17:54.020 that. 00:17:54.020 --> 00:17:55.040 But it's great that they're doing it. 00:17:55.040 --> 00:17:57.160 Do you have any other hypotheses, Brian? 00:17:57.480 --> 00:17:57.860 Yes. 00:17:57.860 --> 00:18:04.020 My hypothesis is that there's not enough examples out in the world of how people are using a 00:18:04.020 --> 00:18:07.120 hypothesis in the field in real world applications. 00:18:07.120 --> 00:18:12.320 So I'm excited that Parsec put it together. 00:18:12.320 --> 00:18:13.020 So Parsec... 00:18:13.020 --> 00:18:15.820 Well, let's take a real quick step back just for people who don't know. 00:18:15.820 --> 00:18:17.180 What is Hypothesis? 00:18:17.180 --> 00:18:18.080 Oh, okay. 00:18:18.080 --> 00:18:18.380 Right. 00:18:18.380 --> 00:18:20.240 Hypothesis is a testing framework. 00:18:20.240 --> 00:18:21.540 Well, it's not really... 00:18:21.540 --> 00:18:23.140 It attaches to other testing frameworks. 00:18:23.140 --> 00:18:25.840 So you can use it with unit test or py test. 00:18:25.840 --> 00:18:27.720 You probably should use it with py test. 00:18:27.720 --> 00:18:35.760 But it's a way, instead of writing a declarative single test or test case, you can... 00:18:35.760 --> 00:18:37.340 It's a property-based testing. 00:18:37.340 --> 00:18:40.220 So you describe kind of... 00:18:40.220 --> 00:18:41.400 It's not like... 00:18:41.400 --> 00:18:43.420 I expect one plus two equals three. 00:18:43.980 --> 00:18:50.940 I expect if I add two integers and they're both positive that the result is going to be 00:18:50.940 --> 00:18:52.360 greater than both of them. 00:18:52.360 --> 00:18:56.600 You have these properties that you describe what the answer is. 00:18:57.260 --> 00:18:58.600 There's a... 00:18:58.600 --> 00:19:06.420 The examples that Hypothesis and other tutorials on how to use Hypothesis have given are more 00:19:06.420 --> 00:19:08.180 of these like A plus B sort of things. 00:19:08.180 --> 00:19:09.280 They're simplistic things. 00:19:09.280 --> 00:19:14.680 And I do see a lot of value in Hypothesis and I know a lot of people are using it. 00:19:14.680 --> 00:19:19.960 But there haven't been a lot of good descriptions for really how it's being used. 00:19:19.960 --> 00:19:23.100 Like a real world example of how it's being used. 00:19:23.100 --> 00:19:26.700 Because I'm probably not going to... 00:19:26.700 --> 00:19:28.960 I don't have those little tiny algorithm things. 00:19:28.960 --> 00:19:30.300 I've got big chunks of stuff. 00:19:30.300 --> 00:19:34.080 And Hypothesis does have to run the test many times. 00:19:34.080 --> 00:19:37.220 So how do you do this effectively on a large project? 00:19:37.220 --> 00:19:38.940 So I love seeing this article. 00:19:38.940 --> 00:19:41.700 So Parsec is a... 00:19:41.700 --> 00:19:45.160 It's a client-side encrypted file sharing service. 00:19:45.160 --> 00:19:46.440 I'd never heard of them before. 00:19:46.440 --> 00:19:48.240 This blog. 00:19:48.240 --> 00:19:49.220 But it sounds cool. 00:19:49.220 --> 00:19:49.480 It's cool. 00:19:49.580 --> 00:19:54.700 They describe themselves as the zero trust file sharing service like Dropbox where it's 00:19:54.700 --> 00:19:56.420 end-to-end encryption for Dropbox. 00:19:56.420 --> 00:20:00.000 You could share the files, but it only matters if you actually have the key, right? 00:20:00.000 --> 00:20:00.480 Right. 00:20:00.480 --> 00:20:02.320 Actually, I have no idea. 00:20:02.320 --> 00:20:03.920 Sure. 00:20:03.920 --> 00:20:05.200 I suspect so. 00:20:05.200 --> 00:20:05.420 Yeah. 00:20:05.420 --> 00:20:07.120 It sounds like a cool service, actually. 00:20:07.120 --> 00:20:08.280 It sounds pretty neat. 00:20:08.280 --> 00:20:13.420 So they describe kind of what they're doing and some of the problems. 00:20:13.420 --> 00:20:17.780 It's a large four-year-old asynchronous Python project. 00:20:18.480 --> 00:20:24.820 And then they describe this RAID redundancy algorithm that they need. 00:20:24.820 --> 00:20:31.580 It's fairly complex with a bunch of servers and stuff, a bunch of data stores going on. 00:20:31.680 --> 00:20:40.820 And what they need to test is they need to check things like if the blocks can be split into chunks and if the blocks can be rebuilt from the chunks that were split up before. 00:20:41.060 --> 00:20:44.300 And then if you can rebuild them if you've got missing chunks. 00:20:44.300 --> 00:20:50.740 And so this all sounds fairly, you know, yeah, I can understand how you could try to test that. 00:20:50.740 --> 00:20:52.360 But there's a lot of variables in there. 00:20:52.360 --> 00:20:53.820 How big is the chunk size? 00:20:53.820 --> 00:20:54.900 How many chunks? 00:20:55.260 --> 00:20:56.760 How much stuff should be missing? 00:20:56.760 --> 00:20:58.480 And all that sort of stuff. 00:20:58.480 --> 00:21:04.060 And then they're thinking, yeah, hypothesis would be good for that. 00:21:04.060 --> 00:21:12.320 The normal tutorials talk about a stateless way to test with hypothesis. 00:21:12.320 --> 00:21:21.760 But they're saying that for them, the stateful method that is supported is very useful because they're an asynchronous system. 00:21:21.760 --> 00:21:23.080 And they describe how to do that. 00:21:23.080 --> 00:21:25.780 It's actually a fairly complex description. 00:21:25.780 --> 00:21:28.460 And it's kind of a lot to get through. 00:21:28.460 --> 00:21:30.580 But it's neat that the power's there. 00:21:30.660 --> 00:21:37.620 So it does, you know, walks through how they exactly how they set up a test like this. 00:21:37.620 --> 00:21:43.720 And this is something I think the testing community of considering hypothesis has been missing. 00:21:43.720 --> 00:21:45.060 So this is great. 00:21:45.060 --> 00:21:51.600 They end with some recommendations, which it's great. 00:21:51.600 --> 00:21:57.820 So the recommendation is for parts of your system that which parts should you throw hypothesis at? 00:21:57.820 --> 00:22:00.520 That's a really good question because you don't want to throw it at everything. 00:22:00.520 --> 00:22:01.120 Right. 00:22:01.120 --> 00:22:04.420 Because there is some expense to set it up and also to run everything. 00:22:04.420 --> 00:22:15.540 So they describe it as if the piece you're testing is kind of an encoder decoder thing, like theirs is you're splitting things into chunks and then rebuilding things. 00:22:15.540 --> 00:22:21.020 It's a hypothesis is a no brainer for that because you can compare. 00:22:21.020 --> 00:22:26.400 Is my input the same as the encoded then decoded output? 00:22:26.400 --> 00:22:27.500 Yeah. 00:22:27.500 --> 00:22:35.260 The other case is if you have a simple oracle, like it's simple to test the answer, but it's complex to come up with the answer. 00:22:35.260 --> 00:22:38.080 I'm not sure what that is. 00:22:38.080 --> 00:22:43.240 But in the case, you know, some of the cases are, you know, I've got a complex system. 00:22:43.240 --> 00:22:47.580 And I just there's properties about the output that are easy to describe. 00:22:48.020 --> 00:22:53.180 The other one is, yeah, I guess similar is if it's hard to compute, but easy to check. 00:22:53.180 --> 00:23:01.560 Well, one example that just jumps out at me right away is anytime you have a file format, I'm going to save this thing, be able to save and load these files. 00:23:01.560 --> 00:23:02.120 Right. 00:23:02.180 --> 00:23:06.120 Because all you got to do is load up a whole bunch of random data, say save, load. 00:23:06.120 --> 00:23:06.880 Is it the same? 00:23:06.880 --> 00:23:09.060 If it's not, that's a problem. 00:23:09.060 --> 00:23:09.460 Yeah. 00:23:09.460 --> 00:23:10.040 Yeah. 00:23:10.040 --> 00:23:10.620 Yeah. 00:23:10.620 --> 00:23:21.900 And actually, I have talked with some people that that have thrown this at some of the standard library modules just on the side to test. 00:23:22.360 --> 00:23:29.500 Because there's a lot of standard library stuff that's like kind of encoding, decoding sort of thing or two way conversions. 00:23:29.500 --> 00:23:30.100 Yeah. 00:23:30.100 --> 00:23:30.380 Cool. 00:23:30.380 --> 00:23:30.780 Yeah. 00:23:30.780 --> 00:23:31.800 This is super nice. 00:23:31.800 --> 00:23:34.700 I'm going to have to really dig into this article in more detail. 00:23:34.700 --> 00:23:42.620 I remember the first time I like learned about hypothesis was when one of the core devs gave a talk at PyCypys 2019. 00:23:42.620 --> 00:23:44.780 And it just blew my mind then. 00:23:44.780 --> 00:23:49.280 And so this is so cool to see this like very, very interesting application here. 00:23:49.280 --> 00:23:49.800 Yeah. 00:23:49.800 --> 00:23:50.260 Yeah. 00:23:50.380 --> 00:23:52.500 It seems like there's a lot of uses in data science. 00:23:52.500 --> 00:24:00.340 Data science seems tough to test like that scientific computation side because slight variations, you might not get perfect equality. 00:24:00.340 --> 00:24:01.160 Exactly. 00:24:01.160 --> 00:24:02.160 Close enough. 00:24:02.160 --> 00:24:02.640 Right. 00:24:02.640 --> 00:24:08.140 It's like, well, it's off, but it's like, you know, 10 to the negative 10th or something off. 00:24:08.140 --> 00:24:08.580 Right. 00:24:08.580 --> 00:24:10.780 That doesn't actually matter, but the equality fails. 00:24:10.780 --> 00:24:11.320 Yeah. 00:24:11.320 --> 00:24:11.660 Yeah. 00:24:11.660 --> 00:24:19.060 You end up using NumPy's, you know, NumPy's approximation comparison schemes quite a bit in your. 00:24:19.800 --> 00:24:20.040 Yeah. 00:24:20.040 --> 00:24:20.760 And your pie test. 00:24:20.760 --> 00:24:21.500 I can imagine. 00:24:21.500 --> 00:24:22.840 I can imagine. 00:24:22.840 --> 00:24:23.580 Very cool. 00:24:23.580 --> 00:24:24.400 All right. 00:24:24.400 --> 00:24:25.640 Next one, Brian. 00:24:25.640 --> 00:24:31.380 I told you about last time I talked about I'm still waiting on my Mac mini, right? 00:24:31.380 --> 00:24:31.920 Yes. 00:24:31.920 --> 00:24:35.680 I ordered the Apple, the M1 Mac mini maxed out. 00:24:35.680 --> 00:24:37.060 And I'm a little bit jealous. 00:24:37.060 --> 00:24:40.020 My daughter is getting a new Mac mini. 00:24:40.020 --> 00:24:42.700 She doesn't, or Mac Air. 00:24:42.700 --> 00:24:45.940 She doesn't know what about, but it's supposed to show up tomorrow and mine's still weeks away. 00:24:45.940 --> 00:24:47.120 And I don't think that that's very fair. 00:24:47.480 --> 00:24:53.820 But if you are an organization that depends on cloud computing and, you know, what organizations 00:24:53.820 --> 00:24:55.100 don't these days, right? 00:24:55.100 --> 00:24:56.060 They almost all do. 00:24:56.060 --> 00:25:02.760 It was just announced at reInvent that AWS is going to be offering Mac instances as a type 00:25:02.760 --> 00:25:03.100 of VM. 00:25:03.220 --> 00:25:06.920 So until now you've been able to get Windows, Linux, that's it. 00:25:06.920 --> 00:25:11.900 So for all those people out there who are offering some kind of iOS app, even if they're not like 00:25:11.900 --> 00:25:16.580 a Mac shop, they still have to have Macs around because you can't compile and sign your IPA, 00:25:16.580 --> 00:25:19.700 your Mac, whatever iPhone app format is. 00:25:19.780 --> 00:25:21.800 You can't create those without a Mac. 00:25:21.800 --> 00:25:26.340 So there's all these Macs that are around for like continuous, you know, CI/CD or checking those 00:25:26.340 --> 00:25:27.700 things and whatnot. 00:25:27.700 --> 00:25:32.080 So now you can go to AWS and say, I'll take a Mac mini, please. 00:25:32.080 --> 00:25:32.940 That's pretty cool. 00:25:32.940 --> 00:25:33.460 That's cool. 00:25:33.460 --> 00:25:33.980 Yeah. 00:25:33.980 --> 00:25:36.580 So you can do your tests up there and they don't have M1 yet. 00:25:36.580 --> 00:25:40.000 Those are the Intel ones, but the M1 chips are coming later. 00:25:40.000 --> 00:25:41.020 So you'll be able to do it. 00:25:41.020 --> 00:25:46.480 What's interesting about this offering from AWS as is basically any cloud service, you would 00:25:46.480 --> 00:25:47.720 imagine it's a VM, right? 00:25:47.940 --> 00:25:52.420 But these, when you say I want one of these, you actually get a dedicated Mac mini. 00:25:52.420 --> 00:25:53.900 That's you get pure hardware. 00:25:53.900 --> 00:25:57.160 Well, that's why you can't get yours because Amazon bought them. 00:25:57.160 --> 00:25:58.540 They did. 00:25:58.540 --> 00:26:00.480 They had a huge truck full of them. 00:26:00.480 --> 00:26:01.720 Well, they bought the Intel one. 00:26:01.720 --> 00:26:03.280 So those were on sale, I bet anyway. 00:26:03.280 --> 00:26:09.020 But no, they have some interesting, what do they call it? 00:26:09.020 --> 00:26:09.500 Nitro. 00:26:09.500 --> 00:26:14.140 I think they call it their Nitro service or something like that, which allows them to virtualize 00:26:14.140 --> 00:26:15.460 actual real hardware. 00:26:15.460 --> 00:26:16.980 So this is pretty neat. 00:26:16.980 --> 00:26:17.620 You can sign up. 00:26:17.620 --> 00:26:18.960 The billing is interesting. 00:26:18.960 --> 00:26:24.100 You have to pay for at least one day's worth if you get it, which I think is like $24. 00:26:24.100 --> 00:26:29.060 If you're going to run it continuously all the time, this is one pricey sucker. 00:26:29.060 --> 00:26:33.220 Like the Mac mini you can get now is $700. 00:26:33.220 --> 00:26:36.220 This is $770 a month. 00:26:36.720 --> 00:26:37.640 Oh, okay. 00:26:37.640 --> 00:26:42.580 So if what you need is like a couple of Mac minis, you're probably, and you need them 00:26:42.580 --> 00:26:43.420 on all the time. 00:26:43.420 --> 00:26:47.340 You're probably better off just buying a few and sticking them in a closet, especially the 00:26:47.340 --> 00:26:48.000 M1s. 00:26:48.000 --> 00:26:51.600 But if you just need one on demand every now and then, or you need to burst into them or 00:26:51.600 --> 00:26:52.860 something like that, that could be interesting. 00:26:52.860 --> 00:26:53.280 Yeah. 00:26:53.280 --> 00:26:53.720 Yeah. 00:26:53.720 --> 00:26:56.780 If you're back old school and you only release like once every three months. 00:26:56.780 --> 00:27:03.700 Well, there was some conversations like, well, if your data is already stored in S3 and you 00:27:03.700 --> 00:27:08.080 have like huge quantity of data and what you need to run is actually running like some video 00:27:08.080 --> 00:27:12.040 processing on the Mac, you could do it by the data instead of transferring that kind of stuff. 00:27:12.040 --> 00:27:13.800 Things like that might be interesting. 00:27:13.800 --> 00:27:14.760 I don't know. 00:27:14.760 --> 00:27:18.720 I would go ahead and throw out there also that this is all interesting. 00:27:18.720 --> 00:27:23.440 I have links to this kind of stuff and whatnot, like the blog post announcing it and so on. 00:27:23.440 --> 00:27:26.460 But there's also this thing called Mac Stadium. 00:27:26.460 --> 00:27:28.740 And if you look at Mac Stadium, it's pretty interesting. 00:27:28.740 --> 00:27:33.900 You go over there and say, give me a dedicated bare mini, a bare metal Mac mini in their data 00:27:33.900 --> 00:27:35.400 center, $60 a month. 00:27:35.400 --> 00:27:41.520 So you can actually get like a decent one for a decent price over there. 00:27:41.520 --> 00:27:43.980 So if you just want one running all the time, it might be good. 00:27:43.980 --> 00:27:48.360 But the thing is, if you're already like deeply integrated to AWS, maybe this is a good thing. 00:27:48.360 --> 00:27:48.780 Yeah. 00:27:48.780 --> 00:27:49.240 Yeah. 00:27:49.240 --> 00:27:51.480 Is there anything you, yeah, go ahead. 00:27:51.480 --> 00:27:53.220 I was just going to say, this seems pretty interesting. 00:27:53.220 --> 00:27:58.380 I mean, I know one of the reasons that I love using GitHub Actions and Azure Pipelines 00:27:58.380 --> 00:28:04.240 is the ability to be able to get access to Mac VMs for builds. 00:28:04.240 --> 00:28:08.520 But if you, I could also see this being really interesting and useful if you have like some 00:28:08.520 --> 00:28:13.700 very huge application or some like very large stack that you want to be able to be able 00:28:13.700 --> 00:28:18.360 to do CI or tests on, that this could be really, really nice. 00:28:18.360 --> 00:28:23.380 Especially if you don't just want to be like, you know, pounding and destroying like one Mac 00:28:23.380 --> 00:28:24.520 over and over and over again. 00:28:25.180 --> 00:28:25.320 Yeah. 00:28:25.320 --> 00:28:25.720 Yeah. 00:28:25.720 --> 00:28:26.440 This is nice. 00:28:26.440 --> 00:28:28.040 Especially if you have a distributed team. 00:28:28.040 --> 00:28:29.380 Yeah. 00:28:29.380 --> 00:28:32.200 Which every team is basically a distributed team. 00:28:32.200 --> 00:28:32.400 Yeah. 00:28:32.400 --> 00:28:32.680 Yeah. 00:28:32.680 --> 00:28:33.520 Welcome to 2020. 00:28:33.520 --> 00:28:37.780 One thing that's interesting about this is you can literally press a button or even just 00:28:37.780 --> 00:28:40.020 through the AWS, probably the Bodo API. 00:28:40.660 --> 00:28:42.920 You can just make a new Mac instantly. 00:28:42.920 --> 00:28:46.380 Like within seconds, you can have a clean, pre-configured Mac. 00:28:46.380 --> 00:28:52.580 You can create AMIs, the Amazon machine image, which are like, install a bunch of stuff and 00:28:52.580 --> 00:28:56.580 get it set up and then like save it so I can respawn new machines from it. 00:28:56.580 --> 00:28:59.900 Those are pretty interesting options that just having a Mac mini in the closet. 00:28:59.900 --> 00:29:03.340 You know, push a button, make a brand new one, try this, throw it away, make it a different 00:29:03.340 --> 00:29:04.160 way, throw it away. 00:29:04.160 --> 00:29:06.720 Like there are some use cases here that could be interesting. 00:29:06.720 --> 00:29:08.380 That said, I won't be using it. 00:29:10.000 --> 00:29:12.020 I'm just going to buy a Mac mini if I can ever get it. 00:29:12.020 --> 00:29:14.040 All right, Matthew, what's this last one you got for us? 00:29:14.040 --> 00:29:17.700 Yeah, I don't have any clever transition, but all right. 00:29:17.700 --> 00:29:24.520 So maybe, I don't know about you, but I end up having to deal with a lot of JSON serializations 00:29:24.520 --> 00:29:29.260 of different statistical models and different, and, you know, sometimes also getting, you know, 00:29:29.260 --> 00:29:33.280 CSVs of different data sets that I want to be doing analysis on. 00:29:33.280 --> 00:29:38.020 And, you know, your first instinct might just be to say, okay, I'm just going to open 00:29:38.020 --> 00:29:41.180 this up in Pandas and start to get to work on it. 00:29:41.180 --> 00:29:47.060 But if you kind of are used and comfortable to working in the Linux command line kind of 00:29:47.060 --> 00:29:51.500 ecosystem of data tools, you might be itching a little bit and want to kind of just, you 00:29:51.500 --> 00:29:54.840 know, peek inside at the command line level and kind of get to work there. 00:29:54.840 --> 00:29:59.960 And so in that case, you might be really interested in this tool called VisiData. 00:29:59.960 --> 00:30:01.940 So VisiData is written on this. 00:30:01.940 --> 00:30:03.740 This is blowing my mind, actually. 00:30:03.940 --> 00:30:07.300 Like, when I saw this, my jaw was kind of on the floor. 00:30:07.300 --> 00:30:12.260 So we'll make sure that this is linked in the show notes because it has some really cool 00:30:12.260 --> 00:30:12.740 videos. 00:30:12.740 --> 00:30:18.540 But so from the docs, so it's VisiData is described as data science without the drudgery. 00:30:18.820 --> 00:30:22.040 So it's an interactive multi-tool for tabular data. 00:30:22.040 --> 00:30:27.460 It combines clarity of spreadsheets with efficiencies of being at the terminal and also, you know, 00:30:27.460 --> 00:30:33.360 the power of Python 3 on a really lightweight utility that can handle millions of rows with 00:30:33.360 --> 00:30:33.700 ease. 00:30:33.700 --> 00:30:35.180 I can attest to that personally. 00:30:35.360 --> 00:30:40.260 I've opened up to like four gigabyte CSV files before and it just, you know, drops 00:30:40.260 --> 00:30:43.020 right in and starts asynchronously loading like a champ. 00:30:43.020 --> 00:30:47.540 It's in addition to that, it supports kind of a really astounding number of file formats 00:30:47.540 --> 00:30:48.280 that it supports. 00:30:48.280 --> 00:30:52.540 Currently on the website, it says it supports 42 different file formats. 00:30:52.540 --> 00:30:58.040 So it, you know, supports things that you would expect like CSV and JSON, but then it also 00:30:58.040 --> 00:30:59.720 supports things like JIRA. 00:30:59.720 --> 00:31:05.040 I guess like whatever JIRA uses for their sort of like tabular stuff. 00:31:05.040 --> 00:31:07.320 It also can like read my, my SQL. 00:31:07.320 --> 00:31:13.220 And I guess it can also even deal with PNG, the image file format, which I was, you know, 00:31:13.220 --> 00:31:13.880 impressed by. 00:31:13.880 --> 00:31:15.720 So this is all openly developed. 00:31:15.720 --> 00:31:17.580 The output is a terminal, right? 00:31:17.580 --> 00:31:17.920 Yeah. 00:31:17.920 --> 00:31:19.640 Not a thing like text. 00:31:19.640 --> 00:31:20.540 Yeah. 00:31:21.540 --> 00:31:22.020 Yeah. 00:31:22.020 --> 00:31:27.560 So this is all openly developed on GitHub by a guy named Saul Pawson, I think. 00:31:27.560 --> 00:31:36.240 And if you go to the, if you go to the Visadata website, it also has plenty of links to live 00:31:36.240 --> 00:31:40.000 demos of him doing kind of interactive examples of visualizations. 00:31:40.000 --> 00:31:45.900 There's one lightning talk that he's given at, I think, PyCascades 2018 or something like 00:31:45.900 --> 00:31:51.520 that, where he's able to just call up a CSV file of, of like, 311 components. 00:31:51.520 --> 00:31:58.380 And then through the, through using Visadata, just kind of hone down onto certain boroughs 00:31:58.380 --> 00:32:05.000 and then be able to do a filter on, on different complaint types to be able to basically find 00:32:05.000 --> 00:32:10.720 complaints about rodents and then filter on rat complaints and then plot that inside of Visadata 00:32:10.720 --> 00:32:17.040 still on the terminal to basically make a visualization of like rodent distribution in the New York City 00:32:17.040 --> 00:32:17.460 boroughs. 00:32:17.560 --> 00:32:20.600 So I thought that was, you know, quite amusing and really cool. 00:32:20.600 --> 00:32:24.540 It's also, you know, this is a Python application. 00:32:24.540 --> 00:32:30.300 So you might not want to, you know, continuously install this in every single virtual environment you make. 00:32:30.920 --> 00:32:34.580 So, I mean, it is up on PyPI, so you can just do pip install Visadata. 00:32:34.580 --> 00:32:40.340 But since it's an application, you probably might also want it just kind of as a generic tool on your machine. 00:32:40.340 --> 00:32:45.600 So it's distributed through a lot of, you know, nice common package managers. 00:32:45.600 --> 00:32:52.040 So if you're on Linux, they've got it on apt, as well as things like NX and GUX. 00:32:52.440 --> 00:32:54.160 But I didn't see it on Yum. 00:32:54.160 --> 00:32:57.620 So if you're on Fedora or CentOS, you might be a little bit out of luck. 00:32:57.620 --> 00:32:58.460 You might have to do it manually. 00:32:58.460 --> 00:33:02.040 It's, of course, on Homebrew and even Condo Forge. 00:33:02.040 --> 00:33:08.060 And it's not listed there, but a very, very cool tool that's been featured on the show before, 00:33:08.060 --> 00:33:10.340 which is PipX by Chad Smith. 00:33:10.340 --> 00:33:11.060 Yeah, PipX is awesome. 00:33:11.060 --> 00:33:11.820 It's so good. 00:33:11.820 --> 00:33:12.460 I love it. 00:33:12.460 --> 00:33:14.160 I tested this last night. 00:33:14.160 --> 00:33:21.100 I just fired up a Python 3.8 Docker container and, you know, went ahead and installed PipX 00:33:21.100 --> 00:33:25.420 and then used PipX to install Vizidata and was able to drop right into Vizidata as expected. 00:33:25.420 --> 00:33:28.040 So it's very, very cool. 00:33:28.040 --> 00:33:33.680 And just the power that you can have with it, I think, is worth checking out for anybody who 00:33:33.680 --> 00:33:36.040 is doing data analysis with tabular data. 00:33:36.040 --> 00:33:37.240 This is super cool. 00:33:37.240 --> 00:33:41.240 I love when people build these tools that are kind of, you don't really expect them to be 00:33:41.240 --> 00:33:41.720 so powerful. 00:33:41.720 --> 00:33:44.920 And you talked about how you just dropped in and grabbed some random data and started answering 00:33:44.920 --> 00:33:45.480 questions. 00:33:45.480 --> 00:33:46.700 And that's super neat. 00:33:46.700 --> 00:33:47.120 Yeah. 00:33:47.120 --> 00:33:47.620 Yeah. 00:33:47.620 --> 00:33:48.720 The number of inputs. 00:33:48.940 --> 00:33:54.560 And because it's open source and because of all the other examples of data types, I 00:33:54.560 --> 00:33:59.840 think even if you have a different data type, it shouldn't be too hard to modify this to 00:33:59.840 --> 00:34:01.060 handle something different. 00:34:01.060 --> 00:34:03.040 I do notice I'm excited about it. 00:34:03.040 --> 00:34:05.580 It does have PCAP files for packet capture. 00:34:05.580 --> 00:34:08.800 These are for communication packets. 00:34:08.800 --> 00:34:12.460 Talking to all your devices and all your hardware at your company, right? 00:34:12.840 --> 00:34:16.320 Well, this is like even Wi-Fi packets and cellular packets. 00:34:16.320 --> 00:34:17.980 That's how we debug those. 00:34:17.980 --> 00:34:18.980 So nice. 00:34:18.980 --> 00:34:20.000 It's very cool. 00:34:20.000 --> 00:34:20.760 Yeah. 00:34:20.760 --> 00:34:21.200 Very cool. 00:34:21.200 --> 00:34:22.240 And PipX is great. 00:34:22.240 --> 00:34:28.060 I install a bunch of apps like Glances, which is a fantastic, like visualize the state, you 00:34:28.060 --> 00:34:29.420 know, like top, but way, way better. 00:34:29.420 --> 00:34:34.140 The ACPy, which is great for it's a better, but much, much better curl. 00:34:34.320 --> 00:34:38.660 But the most important thing I install that way is a Py joke. 00:34:38.660 --> 00:34:42.600 So now I can type Py joke on my command line and we're always right there. 00:34:42.600 --> 00:34:47.360 So speaking of which, move on to our extras. 00:34:47.360 --> 00:34:48.860 That's all of our main topics. 00:34:48.860 --> 00:34:50.200 Brian, you got anything this week? 00:34:50.200 --> 00:34:51.220 Oh, I did. 00:34:51.220 --> 00:34:52.780 I haven't dropped them in. 00:34:52.780 --> 00:34:53.860 Where'd my extras go? 00:34:53.860 --> 00:34:56.160 Yeah, well, you got it. 00:34:56.160 --> 00:35:03.540 I just wanted to bring up that the PyCon 2021 is going to be virtual and there's a website 00:35:03.540 --> 00:35:04.000 up. 00:35:04.000 --> 00:35:07.060 It's us.pycon.org slash 2021. 00:35:07.060 --> 00:35:12.440 And there's not a lot there yet, but you can check out what's going to happen. 00:35:12.440 --> 00:35:19.620 It's not surprising that they have to start planning it and they may as well plan it as a virtual 00:35:19.620 --> 00:35:20.020 event. 00:35:20.720 --> 00:35:23.940 I'm just kind of hoping that we would have live, but I understand. 00:35:23.940 --> 00:35:24.820 Yeah. 00:35:24.820 --> 00:35:27.360 I mean, Hycon is my geek holiday. 00:35:27.360 --> 00:35:28.360 I love it. 00:35:28.360 --> 00:35:32.860 It's both work, but it's also just such a nice getaway to connect with everybody. 00:35:32.860 --> 00:35:38.340 You, everyone else we know from the community, listeners, I'm going to miss not having it. 00:35:38.340 --> 00:35:38.800 Yeah. 00:35:38.800 --> 00:35:39.300 I'm glad. 00:35:39.300 --> 00:35:40.060 Do you attend? 00:35:40.060 --> 00:35:40.800 Sorry, Brian. 00:35:40.800 --> 00:35:45.660 No, it's good that they, I always check whenever they announce the date to make sure it doesn't 00:35:45.660 --> 00:35:46.720 overlap Mother's Day. 00:35:46.720 --> 00:35:48.780 And it does not. 00:35:48.780 --> 00:35:49.860 That's not good. 00:35:50.660 --> 00:35:50.840 Yeah. 00:35:50.840 --> 00:36:00.120 So I have unfortunately not attended PyCon yet in person or, I mean, it was canceled this 00:36:00.120 --> 00:36:00.320 year. 00:36:00.320 --> 00:36:06.500 So maybe, maybe I'll attend this year remote, but I'm a regular attendee of the SciPy conference, 00:36:06.500 --> 00:36:11.720 which this, so this past year, SciPy 2020 was moved online. 00:36:11.720 --> 00:36:16.680 And I thought that the organizers did a fantastic job of actually writing it online while still, 00:36:16.680 --> 00:36:19.000 you know, keeping kind of that SciPy community feel. 00:36:19.000 --> 00:36:23.080 So that was helped a lot also by, you know, plenty of bad puns. 00:36:23.080 --> 00:36:28.840 So I think that might, might be something that still comes through for PyCon 2021, maybe. 00:36:28.840 --> 00:36:30.060 Yeah, absolutely. 00:36:30.060 --> 00:36:36.780 One of the live listeners, Mohammed said, ask if it's going to cost money or if it's going to 00:36:36.780 --> 00:36:38.400 be free this year to attend. 00:36:38.400 --> 00:36:40.080 Did you notice anything, Brian? 00:36:40.080 --> 00:36:41.680 I haven't looked. 00:36:41.680 --> 00:36:45.780 I'm looking around and I don't know that it costs anything. 00:36:45.780 --> 00:36:46.920 It's from what I can tell. 00:36:46.920 --> 00:36:48.000 I don't see any pricing. 00:36:48.000 --> 00:36:53.920 What I saw was sponsor information to get sponsors to sign up to be part of whatever they're doing 00:36:53.920 --> 00:36:54.140 there. 00:36:54.140 --> 00:36:55.480 but I can't tell. 00:36:55.480 --> 00:36:56.280 Yeah, not sure. 00:36:56.280 --> 00:37:01.360 Somebody else throw in the chat or put it into the, you know, visit pythonbytes.fm/211 00:37:01.360 --> 00:37:02.560 and put it in the comments down there. 00:37:02.560 --> 00:37:03.040 All right. 00:37:03.040 --> 00:37:04.460 I got a couple here. 00:37:04.460 --> 00:37:08.560 First of all, we're trying out live streaming here and I think it's going pretty well. 00:37:08.560 --> 00:37:10.460 Seems like it's working out. 00:37:10.460 --> 00:37:11.460 There's a bunch of people watching. 00:37:11.460 --> 00:37:17.420 So if you want to get notified and we happen to keep doing this, just visit pythonbytes.fm 00:37:17.420 --> 00:37:21.160 slash YouTube and it should have like the scheduled upcoming live stream. 00:37:21.160 --> 00:37:22.400 You can like get notified there. 00:37:22.400 --> 00:37:23.920 So we'll, maybe we'll keep doing this. 00:37:23.920 --> 00:37:24.480 It's been fun. 00:37:24.480 --> 00:37:26.720 Thanks for everyone out there who's watching right now. 00:37:26.720 --> 00:37:33.440 And in addition to PyCon, which you just announced or mentioned the announcement of, that is the 00:37:33.440 --> 00:37:40.020 main way that the PSF is funded, but they're also doing a dedicated offering sort of fundraiser 00:37:40.020 --> 00:37:46.620 thing with six companies to help raise some money for the PSF and Talk Python training is 00:37:46.620 --> 00:37:47.240 being part of that. 00:37:47.240 --> 00:37:53.440 And 50% of the revenue of a certain set of our courses that are sold during the month of 00:37:53.440 --> 00:37:55.680 December goes directly to the PSF. 00:37:55.680 --> 00:38:00.940 And people who buy those courses through the PSF fundraiser also get like 20% of a discount. 00:38:00.940 --> 00:38:06.360 So there's a link in the show notes for people to take some of our courses and donate to the 00:38:06.360 --> 00:38:06.740 PSF. 00:38:06.740 --> 00:38:08.800 If you'd rather just directly donate, that's fine. 00:38:08.880 --> 00:38:12.920 But if you're looking to get some of our courses anyway, you can do it this way and support the PSF. 00:38:12.920 --> 00:38:14.320 They're hoping to raise $60,000. 00:38:14.320 --> 00:38:17.320 You know, hopefully we can do that for them and we'll see. 00:38:17.800 --> 00:38:20.060 Brian, you announced Big PyCon. 00:38:20.060 --> 00:38:24.840 Another thing that got announced is Small PyCon, PyCascades. 00:38:25.300 --> 00:38:28.800 PyCascades being the mountain range that connects Portland, Seattle, and Vancouver. 00:38:28.800 --> 00:38:32.080 And traditionally, this conference is cycled between those three cities. 00:38:32.080 --> 00:38:34.920 I don't even remember anymore what it's supposed to be this year. 00:38:34.920 --> 00:38:38.780 I think it's supposed to go back to Vancouver, but it's not going to Vancouver because nobody's 00:38:38.780 --> 00:38:39.300 going anywhere. 00:38:39.300 --> 00:38:41.740 So PyCascades is online. 00:38:42.020 --> 00:38:42.940 And those do cost money. 00:38:42.940 --> 00:38:48.020 It's $10 for students, $20 for individuals, and $50 for professionals to support that conference. 00:38:48.020 --> 00:38:52.140 But I'll link to that one since that's one of our local conferences, if you will. 00:38:52.140 --> 00:38:53.200 Yeah, they're trying to push. 00:38:53.200 --> 00:38:57.240 They often push what's going on, try new things. 00:38:57.240 --> 00:38:58.680 So it's a neat conference. 00:38:58.680 --> 00:38:59.280 Yeah. 00:38:59.280 --> 00:38:59.600 Yeah. 00:38:59.600 --> 00:39:00.640 I enjoy my time there as well. 00:39:00.640 --> 00:39:02.220 All right, Matthew, what are you got for us? 00:39:02.220 --> 00:39:03.800 Anything else you want to get a shout out to? 00:39:03.800 --> 00:39:05.720 Yeah, just a few items. 00:39:05.720 --> 00:39:09.520 So Advent of Code 2020 is started now. 00:39:09.520 --> 00:39:13.380 It's day two, but there's still plenty of time to get involved with that if you want to. 00:39:13.380 --> 00:39:19.700 And for those of you who might not know, Advent of Code is just an annual coding challenge that 00:39:19.700 --> 00:39:21.020 takes place every December. 00:39:21.020 --> 00:39:26.400 And it's just basically 25 days of fun and interesting programming challenges. 00:39:26.400 --> 00:39:31.900 So it's always a great opportunity to try and brush up on your Python and maybe learn about 00:39:31.900 --> 00:39:37.420 some interesting collections that you might not have known about in the standard library. 00:39:38.200 --> 00:39:39.840 So that's going on right now. 00:39:39.840 --> 00:39:41.000 Worth checking out, I think. 00:39:41.000 --> 00:39:48.420 And then I'm going to sneak in some very small physics-related follow-up to Python Bytes 00:39:48.420 --> 00:39:52.680 episode 205, in which awkward arrays were talked about. 00:39:52.680 --> 00:39:59.440 So the lead developer of awkward arrays is my friend and colleague, Jim Povarsky, who is one 00:39:59.440 --> 00:40:04.700 of my scikit-hep co-collaborators, as well as also a member of iris-hep. 00:40:04.700 --> 00:40:13.320 And as of today, which is recording December 2nd, awkward v1.0 is a release can that is up on PyPI. 00:40:13.320 --> 00:40:20.080 So by the time that this goes live, if you just do pip install awkward, you should get awkward 1.0 releases 00:40:20.080 --> 00:40:21.740 instead of having to do the... 00:40:21.740 --> 00:40:22.500 No more awkward 1. 00:40:22.620 --> 00:40:22.980 Exactly. 00:40:22.980 --> 00:40:24.360 No more awkward 1.0. 00:40:24.360 --> 00:40:25.000 No more awkward 0.0. 00:40:25.000 --> 00:40:25.820 All that jazz. 00:40:25.820 --> 00:40:28.980 It's so good to have the actual install statement be awkward itself. 00:40:28.980 --> 00:40:29.560 Exactly. 00:40:29.560 --> 00:40:32.060 So that's a nice little tidbit. 00:40:32.060 --> 00:40:38.060 And I think there's some nice links in episode 205 if people want to learn more about awkward. 00:40:38.480 --> 00:40:43.660 But that's kind of a backbone of kind of the Pythonic ecosystem for physics right now. 00:40:43.660 --> 00:40:49.980 And then finally, I just want to give some kudos to Python Bytes as well, specifically for making 00:40:49.980 --> 00:40:54.360 full transcripts of the shows available to view on pythonbytes.fm. 00:40:54.360 --> 00:40:58.960 Not only is this, I think, like a cool idea in general, but I think this also makes the show 00:40:58.960 --> 00:41:03.120 more inclusive to the deaf Python community, which is definitely out there. 00:41:03.120 --> 00:41:09.880 And one of my good friends and co-authors is deaf, and I know that he definitely appreciates 00:41:09.880 --> 00:41:10.140 this. 00:41:10.140 --> 00:41:14.460 So good job, Ben, you guys for being more inclusive of the wider community. 00:41:14.460 --> 00:41:15.500 Oh, that's so cool. 00:41:15.500 --> 00:41:17.360 I didn't know anybody was utilizing it. 00:41:17.360 --> 00:41:18.520 Yeah, that's awesome. 00:41:18.520 --> 00:41:19.240 Thank you. 00:41:19.240 --> 00:41:24.200 I think it's absolutely critical for that because the format is only audio. 00:41:24.200 --> 00:41:28.240 But a lot of folks have reached out and said they also appreciate it if they're English as 00:41:28.240 --> 00:41:31.900 a second language and they're not as good with English as well. 00:41:31.900 --> 00:41:33.880 So that also helps, I think. 00:41:33.880 --> 00:41:35.800 They're like, what was I saying again? 00:41:35.800 --> 00:41:36.620 What a weird word. 00:41:36.620 --> 00:41:37.860 Awkward array? 00:41:37.860 --> 00:41:38.860 Why would they talk about that? 00:41:38.860 --> 00:41:40.240 It doesn't make sense. 00:41:40.240 --> 00:41:44.540 Yeah, transcripts and closed captioning is just more inclusive for everyone. 00:41:44.540 --> 00:41:45.740 So that's awesome. 00:41:45.740 --> 00:41:46.680 Yeah, thanks. 00:41:46.680 --> 00:41:47.660 All right. 00:41:47.660 --> 00:41:51.100 Well, let's wrap it up with a joke, Brian. 00:41:51.100 --> 00:41:51.760 Yeah. 00:41:51.760 --> 00:41:52.300 All right. 00:41:52.300 --> 00:41:54.560 So you guys, I'm going to need your help here. 00:41:54.560 --> 00:41:57.320 I'm going to let Matthew, I'm going to let you pick. 00:41:57.320 --> 00:41:58.780 Do you want to be Windows or Apple? 00:41:58.780 --> 00:42:00.220 I'll be Windows. 00:42:00.220 --> 00:42:01.980 All right, Brian, you'd be Apple. 00:42:01.980 --> 00:42:07.880 So the idea is like the title here is how to fix a computer, any computer. 00:42:07.880 --> 00:42:10.200 So instructions for Windows. 00:42:10.200 --> 00:42:10.900 Go ahead, Matthew. 00:42:10.900 --> 00:42:12.660 So step one, reboot. 00:42:12.660 --> 00:42:15.320 And then the flowchart goes to, did that fix it? 00:42:15.320 --> 00:42:17.060 If no, proceed to step two. 00:42:17.060 --> 00:42:20.880 Step two, format your hard drive and then reinstall Windows. 00:42:20.880 --> 00:42:23.420 Lose all of your files and quietly leap. 00:42:23.420 --> 00:42:26.360 Brian, Apple doesn't have that problem. 00:42:26.360 --> 00:42:28.120 There's some totally different solution there. 00:42:28.360 --> 00:42:28.760 Okay. 00:42:28.760 --> 00:42:30.260 For Apple, it's step one. 00:42:30.260 --> 00:42:31.400 Take it to an Apple store. 00:42:31.400 --> 00:42:32.560 Did that fix it? 00:42:32.560 --> 00:42:34.300 If no, proceed to step two. 00:42:34.300 --> 00:42:36.240 Step two is buy a new Mac. 00:42:36.240 --> 00:42:39.220 Overdraw your account and quietly weep. 00:42:39.220 --> 00:42:41.540 That's me right now. 00:42:41.540 --> 00:42:41.840 All right. 00:42:41.840 --> 00:42:43.000 I got the Linux fix. 00:42:43.000 --> 00:42:43.520 It's so easy. 00:42:43.520 --> 00:42:45.380 It's totally like you don't need those things. 00:42:45.780 --> 00:42:48.140 So you learn to code in C++. 00:42:48.140 --> 00:42:49.740 You recompile the kernel. 00:42:49.740 --> 00:42:53.360 You build your own microprocessor out of spare silicon you have laying around. 00:42:53.360 --> 00:42:55.180 You recompile the kernel again. 00:42:55.180 --> 00:42:56.180 You switch distros. 00:42:56.180 --> 00:43:02.380 You recompile the kernel again, but this time using a CPU powered by the reflected light from 00:43:02.380 --> 00:43:02.800 Saturn. 00:43:02.800 --> 00:43:04.360 You grow a giant beard. 00:43:04.360 --> 00:43:06.120 You blame Sun Microsystems. 00:43:06.300 --> 00:43:10.160 You turn your bedroom into a server closet and spend 10 years falling asleep to the sound 00:43:10.160 --> 00:43:10.840 of wearing fans. 00:43:10.840 --> 00:43:11.960 You switch distros again. 00:43:11.960 --> 00:43:13.280 You abandon all hygiene. 00:43:13.280 --> 00:43:17.740 You write a regular expression that would make any other programmers cry blood. 00:43:17.740 --> 00:43:19.120 You learn to code in Java. 00:43:19.120 --> 00:43:22.080 You recompile again, but this time while wearing your lucky socks. 00:43:22.080 --> 00:43:22.960 Did that fix it? 00:43:22.960 --> 00:43:23.600 No. 00:43:23.600 --> 00:43:24.620 Proceed to step two. 00:43:24.620 --> 00:43:26.660 Revert back to using Windows and Mac. 00:43:26.660 --> 00:43:27.640 Or Mac. 00:43:27.640 --> 00:43:28.520 Quietly weep. 00:43:30.080 --> 00:43:31.540 There's really no good outcome here. 00:43:31.540 --> 00:43:33.380 They all in and quietly weep. 00:43:33.380 --> 00:43:38.780 As a Linux user for the better part of a decade, I can neither confirm nor deny how accurate 00:43:38.780 --> 00:43:39.680 that last part is. 00:43:39.680 --> 00:43:44.440 Yeah, they all have their own special angle. 00:43:44.440 --> 00:43:49.320 It just takes longer to get there with Linux to get to your destination, I guess. 00:43:49.320 --> 00:43:49.860 Yeah. 00:43:49.860 --> 00:43:50.440 All right. 00:43:50.440 --> 00:43:51.920 Well, that's fun as always. 00:43:51.920 --> 00:43:54.940 And everyone watching on YouTube, thanks for being here live and everyone listening. 00:43:54.940 --> 00:43:55.940 Just thank you for listening. 00:43:55.940 --> 00:43:57.380 Matthew, thanks for joining us. 00:43:57.380 --> 00:43:58.820 Hey, thanks so much for having me. 00:43:58.820 --> 00:43:59.780 This was really fun. 00:43:59.780 --> 00:44:00.500 Yeah, yeah. 00:44:00.500 --> 00:44:01.880 Great for the items you brought. 00:44:01.880 --> 00:44:02.680 Enjoyed them. 00:44:02.680 --> 00:44:04.460 And Brian, thanks as always, man. 00:44:04.460 --> 00:44:05.120 Thank you. 00:44:05.120 --> 00:44:05.720 It's been fun. 00:44:05.720 --> 00:44:06.240 Yep. 00:44:06.240 --> 00:44:06.760 See ya. 00:44:06.760 --> 00:44:07.160 Bye. 00:44:07.160 --> 00:44:09.180 Thank you for listening to Python Bytes. 00:44:09.180 --> 00:44:11.740 Follow the show on Twitter via at Python Bytes. 00:44:11.740 --> 00:44:14.620 That's Python Bytes as in B-Y-T-E-S. 00:44:14.620 --> 00:44:17.960 And get the full show notes at pythonbytes.fm. 00:44:17.960 --> 00:44:22.220 If you have a news item you want featured, just visit pythonbytes.fm and send it our way. 00:44:22.220 --> 00:44:24.920 We're always on the lookout for sharing something cool. 00:44:24.920 --> 00:44:28.100 On behalf of myself and Brian Okken, this is Michael Kennedy. 00:44:28.180 --> 00:44:31.540 Thank you for listening and sharing this podcast with your friends and colleagues.