WEBVTT 00:00:00.001 --> 00:00:05.020 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:05.020 --> 00:00:09.980 This is episode 139, recorded July 11th, 2019. 00:00:09.980 --> 00:00:11.040 I'm Michael Kennedy. 00:00:11.040 --> 00:00:11.880 And I'm Brian Okken. 00:00:11.880 --> 00:00:14.960 And I want to welcome Ines Montani to the show. 00:00:14.960 --> 00:00:16.340 Ines, it's great to have you here. 00:00:16.340 --> 00:00:18.860 Special guest, going to help us cover the news of the week. 00:00:18.860 --> 00:00:19.440 Thanks for being here. 00:00:19.440 --> 00:00:20.400 Yeah, thanks for having me. 00:00:20.400 --> 00:00:20.980 I'm really excited. 00:00:20.980 --> 00:00:23.020 It's going to be really fun to have you participate in this. 00:00:23.020 --> 00:00:25.740 Also, thank you to DigitalOcean for sponsoring this episode. 00:00:25.740 --> 00:00:28.240 Check them out at pythonbytes.fm/digitalocean. 00:00:28.240 --> 00:00:29.200 More on that later. 00:00:30.000 --> 00:00:32.520 Brian, we've talked a lot about like, how do you install Python? 00:00:32.520 --> 00:00:34.120 How do you manage Python? 00:00:34.120 --> 00:00:35.640 How do you upgrade your stuff? 00:00:35.640 --> 00:00:37.300 There's just so many ways. 00:00:37.300 --> 00:00:40.360 And then we've got things like pipenv, flit, poetry. 00:00:40.360 --> 00:00:41.780 It goes on and on. 00:00:41.780 --> 00:00:43.340 And it still goes on, right? 00:00:43.340 --> 00:00:46.860 This was a contribution from Niels DeBruin sent us this. 00:00:46.860 --> 00:00:50.640 There was an article called Simplify Your Python Developer Environment. 00:00:50.640 --> 00:00:56.460 And it talked about using pyenv, pipex, and pipenv together. 00:00:56.880 --> 00:01:00.240 And my first reaction was like, we've already covered all of these. 00:01:00.240 --> 00:01:02.720 However, I have tried pipex. 00:01:02.720 --> 00:01:05.040 Actually, I did the joke, the pi jokes. 00:01:05.040 --> 00:01:06.260 I did that with pipex. 00:01:06.260 --> 00:01:09.440 And I've tried pipenv once before. 00:01:09.440 --> 00:01:12.520 It doesn't really do anything for me that I really need. 00:01:13.000 --> 00:01:16.800 And the pyenv, I've tried it and it didn't work for me. 00:01:16.800 --> 00:01:21.160 So actually, all these things, I kind of wanted to give them another shot anyway. 00:01:21.160 --> 00:01:23.420 So I went ahead and read this article. 00:01:23.420 --> 00:01:26.240 And it actually is pretty nice. 00:01:26.240 --> 00:01:30.220 It's a nice pros and cons of all the tools and how to set them up. 00:01:30.420 --> 00:01:37.020 And I think for somebody that wants to try these out again, this is a good article to read to try to get back into it. 00:01:37.140 --> 00:01:45.720 So if people don't remember, pyenv is used to install and manage multiple Python versions and flavors on a computer. 00:01:45.720 --> 00:01:57.760 And then pipex is something that allows you to create, take a Python application and have it bundled with its own virtual environment and use it globally on your system without having to activate the environment. 00:01:57.980 --> 00:02:07.720 And then pipenv is for when you're working on project and application, it's a way to manage virtual environments and dependencies on a per project basis. 00:02:07.720 --> 00:02:14.160 These are really interesting because I feel like often they kind of blend like in a blur together, right? 00:02:14.160 --> 00:02:17.940 You're like, well, I know there's all these E and V things I could use and whatnot. 00:02:17.940 --> 00:02:21.840 And it's like, well, when should I use what and what one is relevant for the situation? 00:02:21.840 --> 00:02:22.560 And what do you think, Anis? 00:02:22.560 --> 00:02:26.180 It actually took me a second to remember the one that I've used. 00:02:26.180 --> 00:02:27.940 So I'm totally in the market for stuff like that. 00:02:27.940 --> 00:02:30.220 But actually, I hadn't heard of pipex at all. 00:02:30.220 --> 00:02:32.960 But I think, yeah, pyenv I definitely use. 00:02:32.960 --> 00:02:38.860 And I think it's quite important, at least for the work I'm doing, because I need to run stuff in all kinds of Pythons. 00:02:38.860 --> 00:02:42.960 And I mean, as a library developer, we need to build stuff for like Python 2. 00:02:42.960 --> 00:02:45.780 We're supporting Python 2.7 and 3.5. 00:02:45.780 --> 00:02:54.720 Just so folks know, like you're deeply involved with spaCy and some tools built on top of that, the natural language processing, which will come evident as we go through some of your topics and more. 00:02:54.720 --> 00:02:56.780 But, you know, maybe not everyone knows your background. 00:02:56.880 --> 00:03:02.820 But I feel like pyenv is most relevant for people building libraries. 00:03:02.820 --> 00:03:07.560 And pipenv is really relevant for people building applications. 00:03:07.560 --> 00:03:08.160 I don't know. 00:03:08.160 --> 00:03:08.800 What do you think, Brian? 00:03:08.800 --> 00:03:11.780 Both of them are important for people like us that have to do both. 00:03:11.980 --> 00:03:14.220 I have a question for Ines. 00:03:14.220 --> 00:03:16.100 The pyenv. 00:03:16.100 --> 00:03:28.160 So one of my concerns was, can I use it to install, to have multiple Pythons and still be able to run them all from one project, like with the tox build to be able to access all of them? 00:03:28.320 --> 00:03:31.420 So I haven't tried it all like within the exact same projects. 00:03:31.420 --> 00:03:34.500 I think you might need different virtual environments for that. 00:03:34.500 --> 00:03:35.620 At least that's how I use it. 00:03:35.620 --> 00:03:38.920 So I use venv to create my virtual environments. 00:03:38.920 --> 00:03:44.440 And then they are created with whichever Python version I've configured locally. 00:03:44.440 --> 00:03:46.680 So I do like pyenv local 2.7. 00:03:46.680 --> 00:03:49.260 And then the virtual environment will have 2.7 in it. 00:03:49.260 --> 00:03:49.500 Okay. 00:03:49.500 --> 00:03:50.780 Well, I'll give it a shot. 00:03:50.780 --> 00:03:51.260 Yeah, yeah. 00:03:51.260 --> 00:03:51.720 Really nice. 00:03:52.100 --> 00:03:53.380 I'm a big fan of PipX. 00:03:53.380 --> 00:03:55.140 I think PipX is super cool. 00:03:55.140 --> 00:04:00.740 So PipX is great if you have a thing that you just want as a utility on the command line in your terminal. 00:04:00.740 --> 00:04:02.840 That happens to be Python-based. 00:04:02.840 --> 00:04:07.880 So instead of brew install something or npm install something, PipX does that. 00:04:07.880 --> 00:04:09.740 So, for example, I have like cookie cutter. 00:04:09.740 --> 00:04:12.160 I have the HTTP library. 00:04:12.160 --> 00:04:12.800 I have glances. 00:04:12.800 --> 00:04:15.640 I have PyJokes, of course, because we run this podcast. 00:04:15.640 --> 00:04:19.380 And some other stuff that I'll even talk about later, like PTPython. 00:04:19.380 --> 00:04:21.900 Is that a bit like PEX files or is that something else? 00:04:21.980 --> 00:04:23.600 It's a little bit. 00:04:23.600 --> 00:04:31.080 What it does is you say I want to have, let's say I want the utility PTPython, which is like an Emacs enabled rebel, basically. 00:04:31.080 --> 00:04:31.360 Okay. 00:04:31.360 --> 00:04:32.340 So I want to have that. 00:04:32.340 --> 00:04:33.560 And I want to be able to just type it. 00:04:33.560 --> 00:04:34.920 It's not tied to any project. 00:04:34.920 --> 00:04:37.940 I just want to have it as a command on my computer that I can use. 00:04:37.940 --> 00:04:41.360 So what you can do is you can PipX install PTPython. 00:04:41.360 --> 00:04:48.360 And then it just automatically puts it in a location, modifies your path so that it has its own virtual environment. 00:04:48.360 --> 00:04:51.100 It upgrades itself and its dependencies separately. 00:04:51.280 --> 00:04:55.760 But anytime I want, I can just type PTPython and go crazy or glances or PyJoke or whatever. 00:04:55.760 --> 00:04:56.200 Okay. 00:04:56.200 --> 00:04:57.260 I think I get it. 00:04:57.260 --> 00:04:57.440 Okay. 00:04:57.440 --> 00:04:58.540 That actually sounds quite cool. 00:04:58.540 --> 00:04:58.740 Yeah. 00:04:58.740 --> 00:04:59.320 It's really nice. 00:04:59.320 --> 00:05:04.020 You can just say, what are the updates from my Python libraries that I use as applications or little utilities? 00:05:04.020 --> 00:05:04.700 It's pretty cool. 00:05:04.700 --> 00:05:05.520 I like that one as well. 00:05:05.520 --> 00:05:05.960 Yeah. 00:05:05.960 --> 00:05:10.220 I wonder if we can ship our annotation tool Prodigy like that because it's very, very command line heavy. 00:05:10.220 --> 00:05:13.040 And, you know, it's usually kind of a separate thing. 00:05:13.040 --> 00:05:15.120 All it does is build upon pip. 00:05:15.120 --> 00:05:21.360 So if you could pip install the thing and it has an entry point, then you can pipX install it as well. 00:05:21.360 --> 00:05:22.020 Ah, okay. 00:05:22.180 --> 00:05:22.360 Yeah. 00:05:22.360 --> 00:05:22.600 Cool. 00:05:22.600 --> 00:05:22.840 Yeah. 00:05:22.840 --> 00:05:23.160 Pretty cool. 00:05:23.160 --> 00:05:25.220 All right, Brian, are you switching to this? 00:05:25.220 --> 00:05:28.020 Are you going to use PyEnv, pipX, and pipEnv? 00:05:28.020 --> 00:05:29.020 Is this your new plan? 00:05:29.020 --> 00:05:33.460 I definitely want to try PyEnv because I want to try new Python versions. 00:05:33.460 --> 00:05:39.620 And my old process was to just download the regular install and install it. 00:05:39.620 --> 00:05:42.060 And then my path is all weird. 00:05:42.060 --> 00:05:43.660 And, yeah, it's a mess. 00:05:43.660 --> 00:05:44.160 Yeah, classic. 00:05:44.160 --> 00:05:44.400 Cool. 00:05:44.400 --> 00:05:45.360 Yeah, this is nice. 00:05:45.360 --> 00:05:47.260 Definitely for people looking for a different workflow. 00:05:47.260 --> 00:05:49.060 It's something they can check out for sure. 00:05:49.060 --> 00:05:51.440 All right, Ines, what's this next one that you got for us? 00:05:51.440 --> 00:05:56.400 This week was actually, or like the past few weeks were actually super exciting also in my field. 00:05:56.400 --> 00:06:01.500 So there was a new release of a fast AI course for natural language processing. 00:06:02.220 --> 00:06:07.620 So fast AI is a very popular free online course for deep learning by Rachel Thomas and Jeremy Howard. 00:06:07.620 --> 00:06:09.640 And it also comes with a Python library. 00:06:09.640 --> 00:06:12.560 It comes with lots of notebooks, really active communities. 00:06:12.560 --> 00:06:15.640 So if you want to get into like the modern machine learning stuff, 00:06:15.640 --> 00:06:19.360 that's like probably the go-to course that I would also recommend to you. 00:06:19.360 --> 00:06:26.020 And they've also produced some very influential research developed alongside the library and the course. 00:06:26.020 --> 00:06:30.280 So, for example, ULM Fit, which was a very popular algorithm for text classification. 00:06:31.080 --> 00:06:34.860 And yeah, the new thing is they've just released a course of natural language processing. 00:06:34.860 --> 00:06:36.400 And it's a very practical introduction. 00:06:36.800 --> 00:06:40.720 And what I thought was really, really interesting and really cool about it is that it, of course, 00:06:40.720 --> 00:06:44.620 covers like the modern neural network approaches and all the like very hip stuff. 00:06:44.620 --> 00:06:47.440 But it also focuses on traditional techniques. 00:06:47.440 --> 00:06:54.020 So just the whole background of like, okay, what did people do before deep learning, topic modeling, linear models, 00:06:54.020 --> 00:06:58.480 just really all the basics and even like rule-based approaches like regular expressions. 00:06:59.080 --> 00:07:01.660 Like, you know, some people might look at this and be like, what regics? 00:07:01.660 --> 00:07:03.740 That's like, you know, I did that 20 years ago. 00:07:03.740 --> 00:07:09.680 But in fact, like, you know, in real life and in like real life practical applications, that's like super important. 00:07:09.680 --> 00:07:12.480 And you can do a lot with that, that, you know, really gets the job done. 00:07:12.480 --> 00:07:13.780 So I thought that was really cool. 00:07:13.780 --> 00:07:17.900 And of course, another thing, the course has a really strong focus on ethics as well. 00:07:18.000 --> 00:07:26.180 So their videos on bias and disinformation and basically just the topic of like, okay, think about the impact that the work you'll be doing has. 00:07:26.180 --> 00:07:27.900 And I thought that was incredibly refreshing. 00:07:27.900 --> 00:07:31.460 And of course, disclaimer, like I haven't actually watched all the videos yet. 00:07:31.460 --> 00:07:36.460 I don't even think that's like physically possible to do all of that since it was released. 00:07:36.460 --> 00:07:37.940 It's too new and it's too long. 00:07:37.940 --> 00:07:38.160 Yeah. 00:07:38.160 --> 00:07:38.680 Yeah. 00:07:38.680 --> 00:07:42.260 It's like, it's a lot of material, but yeah, I really like the work they're doing. 00:07:42.260 --> 00:07:44.740 And it's like, yeah, it's a very, very significant release. 00:07:44.740 --> 00:07:45.480 And it's all free. 00:07:45.480 --> 00:07:45.720 Yeah. 00:07:45.720 --> 00:07:46.300 It's super cool. 00:07:46.300 --> 00:07:46.420 Yeah. 00:07:46.420 --> 00:07:47.120 It looks like it's free. 00:07:47.260 --> 00:07:50.240 Like I was able to go pull up the videos without even having an account there. 00:07:50.240 --> 00:07:51.440 It just sort of takes you through it. 00:07:51.440 --> 00:07:53.820 More like an online video book. 00:07:53.820 --> 00:07:54.800 You just make your way through, right? 00:07:54.800 --> 00:07:55.100 Yeah. 00:07:55.100 --> 00:07:56.160 And lots of like notebooks. 00:07:56.160 --> 00:07:57.420 So you can open the same notebooks. 00:07:57.420 --> 00:08:04.220 You can play through the examples and of course also use their library to, you know, really work through the things efficiently. 00:08:04.220 --> 00:08:04.580 Sure. 00:08:04.580 --> 00:08:04.820 Yeah. 00:08:04.820 --> 00:08:05.460 That's super cool. 00:08:05.460 --> 00:08:08.340 Does it cover any of the libraries that you all work on? 00:08:08.340 --> 00:08:09.420 Like spaCy or anything? 00:08:09.420 --> 00:08:12.280 spaCy is a much more like high level toolkit and framework. 00:08:12.280 --> 00:08:15.800 So it's like, you know, this is really the basics of the technology. 00:08:16.040 --> 00:08:19.300 So while the Fast AR library, I think it does use spaCy for tokenization. 00:08:19.300 --> 00:08:28.500 But spaCy is really, you know, once you're building applications and you have some problems and you want to construct your pipelines and really ship something into production, that's when you would be using spaCy. 00:08:28.880 --> 00:08:35.600 But actually, in fact, spaCy is not really the best solution if you really want to learn the underlying algorithms and implementations. 00:08:35.600 --> 00:08:37.860 Because we're actually super opinionated. 00:08:37.860 --> 00:08:41.780 You get one implementation and, you know, you kind of take that or you plug in your own. 00:08:41.780 --> 00:08:42.040 I see. 00:08:42.040 --> 00:08:46.020 This is like learning the algorithms and the foundations that maybe spaCy uses so you understand it better. 00:08:46.140 --> 00:08:46.580 Yeah, exactly. 00:08:46.580 --> 00:08:49.320 And also, you know, giving you some, yeah, the background. 00:08:49.320 --> 00:08:54.300 And yeah, even the rule-based ideas, which, yeah, I still think it's so great. 00:08:54.300 --> 00:08:55.160 I scroll through it. 00:08:55.160 --> 00:08:56.460 I'm like, great, regular expressions. 00:08:56.460 --> 00:08:59.380 That's like really what people should think about. 00:08:59.380 --> 00:09:00.500 Yeah, I mean, that's a start, right? 00:09:00.500 --> 00:09:03.280 Like sometimes you just want to pull data out of text and there you go. 00:09:03.400 --> 00:09:11.780 Yeah, and especially then some people really, you know, then they start throwing like a neural network model at it when actually best solution would have been to just write one regular expression. 00:09:11.780 --> 00:09:17.180 Like, I don't know, you work at a company, you have a statistical model that recognizes organization names. 00:09:17.180 --> 00:09:22.700 And then, you know, your manager comes to you and is like, well, it's all great, but like it often gets our own company name wrong. 00:09:22.700 --> 00:09:23.480 Can you fix that? 00:09:23.480 --> 00:09:24.260 That's really embarrassing. 00:09:24.260 --> 00:09:30.740 And, you know, you can spend like hours trying to tune your model and like update it and fine tune it with more examples. 00:09:30.740 --> 00:09:38.140 Or you can just add one regular expression or one rule on top that says, okay, whenever I come across this string, don't get it wrong. 00:09:38.140 --> 00:09:42.640 And that will likely take you like five minutes and it's much more effective in the real world. 00:09:42.640 --> 00:09:46.220 So this is the practical applied natural language processing, right? 00:09:46.220 --> 00:09:46.520 Yeah. 00:09:46.520 --> 00:09:51.320 Speaking of language, this next item that I want to talk about, it just scares me. 00:09:51.320 --> 00:09:52.500 So let me tell you quick about it. 00:09:52.500 --> 00:09:54.180 I'll get your two opinions. 00:09:54.180 --> 00:10:02.960 The idea is that we can clone the human voice by giving it a sample and using some sort of neural network type thing. 00:10:02.960 --> 00:10:03.700 I'm not sure exactly. 00:10:03.700 --> 00:10:05.680 So this was sent in by Brendan. 00:10:05.680 --> 00:10:06.740 Thank you for sending that over. 00:10:06.740 --> 00:10:16.500 And, you know, in just a couple of minutes, you can load up somebody's voice, use some Python, a pre-trained machine learning model, and then type in text. 00:10:16.640 --> 00:10:20.980 It will speak whatever that text is in the voice of that person. 00:10:20.980 --> 00:10:23.560 It's just, I don't know. 00:10:23.560 --> 00:10:26.600 I feel like public discourse is in serious jeopardy here. 00:10:26.600 --> 00:10:27.340 What do you all think? 00:10:27.340 --> 00:10:28.560 This is pretty interesting. 00:10:28.820 --> 00:10:35.000 And I was a little frightened, actually, watching this video of how easy it was to copy somebody's voice. 00:10:35.000 --> 00:10:39.240 I mean, I don't know what the code behind it's probably not easy, but it's a little creepy. 00:10:39.240 --> 00:10:44.460 My first thought was I could use it for if I have solo episodes, I could be my own co-host. 00:10:44.460 --> 00:10:46.580 But that'd be cool. 00:10:46.580 --> 00:10:47.500 That'd be pretty funny. 00:10:47.500 --> 00:10:47.820 Yeah. 00:10:47.820 --> 00:10:52.380 No, but I think I would say, I think it's still quite compute intensive, like, to do that, right? 00:10:52.380 --> 00:11:01.040 I think you still, you know, if you really want to have good results and really want to do it right, but it's true that this, you know, this really, this is a really good, like, example of, wow, that's possible. 00:11:01.040 --> 00:11:09.580 And that's the type of stuff that's been possible for quite a while, especially across, like, you know, video image and, you know, also voice, audio. 00:11:09.580 --> 00:11:09.900 Right. 00:11:09.900 --> 00:11:13.260 I mean, we've heard of the deepfake stuff for videos and whatnot, and that's kind of scary. 00:11:13.260 --> 00:11:16.480 But this is, it's almost like this could be worse, right? 00:11:16.480 --> 00:11:24.840 I can imagine somebody just putting a little static, a little muffling filter on it and saying, oh, here's a hot mic take behind the scenes. 00:11:24.840 --> 00:11:33.340 Somebody said something they weren't supposed to during a presidential debate or some kind of public figure they're trying to discredit and just make them say stuff. 00:11:33.340 --> 00:11:34.800 Now, I don't know what you all think. 00:11:34.800 --> 00:11:36.760 To me, it still sounds a little bit off. 00:11:36.760 --> 00:11:41.820 Like, it doesn't sound exactly like the person, but, you know, it's pretty close. 00:11:42.000 --> 00:11:43.340 But I think you would still use it. 00:11:43.340 --> 00:11:43.940 Yeah. 00:11:43.940 --> 00:11:52.700 And also, I mean, I do think this type of technology will definitely lead to a situation where we will all just take things we hear in, like, recordings or videos much less seriously. 00:11:52.700 --> 00:12:00.400 Like, you know, I do think it will develop into a culture where we don't necessarily trust an audio recording because, you know, it could have been produced by something like this. 00:12:00.400 --> 00:12:10.960 But I do think just for everyday life scam, like, I think it's very timely because, yeah, there are all these news articles about all these, like, financial scams and companies using deepfake audio. 00:12:11.740 --> 00:12:15.280 Or even before, like, using spoofed emails that were quite effective. 00:12:15.280 --> 00:12:22.720 And actually reading that, like, you know, the trick there, I'm like, I can totally see how, like, an accounting department falls for that and thinks, oh, it's their boss in a hurry. 00:12:22.720 --> 00:12:26.660 And then imagine that with a voice of, like, you know, CEO. 00:12:26.660 --> 00:12:27.420 So scary. 00:12:27.420 --> 00:12:28.080 Exactly. 00:12:28.080 --> 00:12:28.420 Yeah. 00:12:28.420 --> 00:12:29.660 So I'm going to link to the video. 00:12:29.660 --> 00:12:30.580 You all can watch it. 00:12:30.580 --> 00:12:33.060 I'm also linking to the software that did this. 00:12:33.060 --> 00:12:40.480 Apparently, it uses something called transfer learning from speaker verification to multi-speaker text-to-speech synthesis. 00:12:41.000 --> 00:12:44.980 And that even has an acronym, SV2-TTS, of course. 00:12:44.980 --> 00:12:47.200 And so this was based on someone's thesis. 00:12:47.200 --> 00:12:49.240 And you can watch the video and get a good sense. 00:12:49.240 --> 00:12:57.220 But, yeah, you just imagine, like, I call, I somehow get the number of the CEO and I call them up and I record that call. 00:12:57.560 --> 00:13:00.720 And then I take their, like, I don't need only five seconds or so with their voice. 00:13:00.720 --> 00:13:04.540 And then I take that and I use some text to, like, real-time generate. 00:13:04.540 --> 00:13:07.100 I'll call up the accounting department and say, hey, this is so-and-so. 00:13:07.100 --> 00:13:08.180 We've got a super emergency. 00:13:08.180 --> 00:13:09.860 Really important client. 00:13:09.860 --> 00:13:10.800 We forgot to pay them. 00:13:10.800 --> 00:13:12.280 We owe them $10,000. 00:13:12.280 --> 00:13:17.900 And you just, like, type and then replay what you type live back to them. 00:13:17.900 --> 00:13:24.880 Yeah, and then you overlay that, yeah, with some background music, background noise and just, like, oh, I'm, like, in a taxi right now. 00:13:24.880 --> 00:13:25.300 Exactly. 00:13:25.300 --> 00:13:26.780 Like, please get this done ASAP. 00:13:26.840 --> 00:13:27.540 Otherwise, we're in trouble. 00:13:27.540 --> 00:13:27.980 Bye. 00:13:27.980 --> 00:13:29.960 Sometimes what usually doesn't work for email. 00:13:29.960 --> 00:13:34.760 But if it's literally a voice that's interacting with you that sounds like the boss, well, it might work. 00:13:34.760 --> 00:13:36.300 Hopefully we didn't give anyone ideas. 00:13:36.300 --> 00:13:38.920 I guess you guys are better criminals than me. 00:13:38.920 --> 00:13:42.680 I was just thinking, like, a different version of Ferris Bueller's Day Off. 00:13:42.680 --> 00:13:45.800 You could just use this to call in and excuse yourself from school. 00:13:45.800 --> 00:13:47.060 Oh, my gosh, you're right. 00:13:47.060 --> 00:13:50.740 This would be beautiful when I was in, like, middle school or high school. 00:13:50.740 --> 00:13:51.640 Oh, my goodness. 00:13:51.640 --> 00:13:51.960 Yeah. 00:13:51.960 --> 00:13:54.060 No, Michael's not feeling well. 00:13:54.060 --> 00:13:54.800 Is he going to be okay? 00:13:54.920 --> 00:13:57.500 He may be out tomorrow, but he'll be back pretty soon. 00:13:57.500 --> 00:13:59.680 All right, then. 00:13:59.680 --> 00:14:01.280 Yeah, that's really horrible. 00:14:01.280 --> 00:14:04.200 Yeah, I mean, pretty soon you have, like, kids recording their parents' voices. 00:14:04.200 --> 00:14:04.980 Yes. 00:14:04.980 --> 00:14:06.020 I don't know if that works. 00:14:06.100 --> 00:14:07.040 These are all bad. 00:14:07.040 --> 00:14:12.600 This is a good example of why, yes, a focus on ethics, you know, when you're learning these technologies is, like, incredibly important. 00:14:12.600 --> 00:14:19.120 Because, you know, we have that technology and every developer should think about, okay, what's the impact of having this and using this? 00:14:19.120 --> 00:14:24.160 And, okay, we can release it to really also, you know, make everyone aware that this exists. 00:14:24.160 --> 00:14:26.280 And, you know, we're talking about this right now, but still. 00:14:26.280 --> 00:14:26.540 Yeah. 00:14:26.540 --> 00:14:27.000 Yeah. 00:14:27.060 --> 00:14:28.540 I guess final thought on this one, Ines. 00:14:28.540 --> 00:14:37.080 What do you think the chances of some sort of fingerprinting or, like, system that can determine that this was faked, right? 00:14:37.080 --> 00:14:48.400 Like, not a human, but if I could take this and feed it to, say, another ML model that knows, like, the little glitches that show up in the system, like, will we be able to verify stuff or not in the future? 00:14:48.400 --> 00:14:49.300 Are we just lost? 00:14:49.300 --> 00:14:49.800 I think so. 00:14:49.800 --> 00:14:52.120 I think there have been, like, some experiments where they tried that. 00:14:52.120 --> 00:14:59.780 And, I mean, another approach would be, okay, you can always encode things in the model that, like, only show up under very, very certain circumstances. 00:14:59.780 --> 00:15:01.800 So, that's how, you know, you can watermark that model. 00:15:01.800 --> 00:15:03.680 You can release that. 00:15:03.680 --> 00:15:12.940 And then if you say a very, very, very specific sequence or if you type nonsense secrets in there, it will always produce something nonsense but very differently. 00:15:12.940 --> 00:15:15.260 And then you're like, ah, that's the system that was used. 00:15:15.260 --> 00:15:17.040 And, you know, that's how we can. 00:15:17.280 --> 00:15:23.400 It won't work on the output, but, like, at least, you know, we can at least have some way of kind of finding out what type of system. 00:15:23.400 --> 00:15:25.260 Yeah, that's part of the ethics part, right? 00:15:25.260 --> 00:15:29.520 Like, that you embed these little watermarks rather than just, like, put it out there. 00:15:29.520 --> 00:15:30.020 I don't know. 00:15:30.020 --> 00:15:35.160 It's pretty scary to me, but I think as a society, we'll come around to figure out what to do about it. 00:15:35.160 --> 00:15:35.580 Cool. 00:15:35.580 --> 00:15:37.500 Well, not scary is DigitalOcean. 00:15:37.500 --> 00:15:39.360 I just want to tell you quickly about them. 00:15:39.360 --> 00:15:40.420 And they're supporting the show. 00:15:40.420 --> 00:15:41.560 So, thanks to DigitalOcean. 00:15:41.560 --> 00:15:44.940 All of our software runs on top of DigitalOcean infrastructure. 00:15:45.260 --> 00:15:49.880 You get the MP3s, it delivers either streams or downloads out of there, things like that. 00:15:49.880 --> 00:15:51.020 So, they're really, really great. 00:15:51.020 --> 00:15:54.920 You can get started for as little as $5 per month for a server. 00:15:54.920 --> 00:15:58.420 And they got a bunch of cool services, managed databases, load balancers, and whatnot. 00:15:59.120 --> 00:16:02.600 And it's, you know, it's not like EC2, which is so complicated. 00:16:02.600 --> 00:16:03.380 It could run Netflix. 00:16:03.380 --> 00:16:06.940 It's like the simple thing that you just need to build your app and get it going. 00:16:06.940 --> 00:16:10.220 So, check them out at pythonbytes.fm/DigitalOcean. 00:16:10.220 --> 00:16:12.640 And you get a $50 credit for new users there. 00:16:12.640 --> 00:16:14.280 And definitely highly recommend it. 00:16:14.280 --> 00:16:16.860 Brian, what's this next one that you're working on here? 00:16:16.860 --> 00:16:17.220 Okay. 00:16:17.300 --> 00:16:20.700 Well, another contributed by a listener, this one from Andrew. 00:16:20.700 --> 00:16:26.320 He contributed a little snippet that was on a Reddit stream. 00:16:26.320 --> 00:16:35.480 And it was, I'm going to just guess, just describe it as abusing the pyproject.tomo file by putting any file stuff in it. 00:16:35.480 --> 00:16:44.240 So, the example that he gave was you can have the pytest any file and the mypy any file or two any files for tools. 00:16:44.240 --> 00:16:46.560 And you can, they're in any file format. 00:16:46.560 --> 00:16:51.020 And Tommel files kind of look like any files, but they're not. 00:16:51.020 --> 00:16:51.940 They're different. 00:16:51.940 --> 00:16:57.060 And you can break any with Tommel syntax and you can break Tommel with any syntax. 00:16:57.060 --> 00:17:04.740 However, you can write them such that they are, if you're careful, you can write them such that they comply with both. 00:17:04.740 --> 00:17:07.000 And I went ahead and tried this out. 00:17:07.000 --> 00:17:13.380 I was able to try putting pytest any, like, options within the Tommel file. 00:17:13.380 --> 00:17:17.800 And, you know, both pytest and mypy do not support doing this. 00:17:17.800 --> 00:17:21.960 But they do support passing in a path of where their config file is. 00:17:21.960 --> 00:17:27.000 And if you pass in the project.tommel file, you can get it to work. 00:17:27.000 --> 00:17:32.780 And if all you're trying to do is reduce the number of files in your project, yeah, this kind of works. 00:17:32.780 --> 00:17:33.940 Why would you want to do that? 00:17:34.000 --> 00:17:37.520 Just to try to reduce the number of files in your top level directory. 00:17:37.520 --> 00:17:37.940 Okay. 00:17:37.940 --> 00:17:38.960 I don't know. 00:17:38.960 --> 00:17:41.380 I should, like, use all of these, like, files more. 00:17:41.380 --> 00:17:44.820 And I feel like, you know, I love this idea of imagine if there was one config file. 00:17:44.820 --> 00:17:47.900 Like, really, you know, only one place where you put everything. 00:17:47.900 --> 00:17:49.040 Also, your dependencies. 00:17:49.460 --> 00:17:51.080 Everything just goes in one file. 00:17:51.080 --> 00:17:52.300 And then you have that. 00:17:52.300 --> 00:17:55.400 But for some reason, it's never actually worked out that way in practice. 00:17:55.400 --> 00:17:57.520 You can put it in the setup.cfg. 00:17:57.520 --> 00:17:59.060 That's a possibility. 00:17:59.060 --> 00:18:01.640 But those are any file syntax also. 00:18:01.640 --> 00:18:04.700 I have to admit, like, our projects don't even have a setup.cfg. 00:18:04.700 --> 00:18:06.160 We have a setup.py. 00:18:06.160 --> 00:18:07.840 And then we have the requirements.txt. 00:18:07.840 --> 00:18:08.260 Okay. 00:18:08.420 --> 00:18:10.300 I have an admission as well. 00:18:10.300 --> 00:18:15.840 Like, I know we're talking about pyproject.toml, but I'm still just using requirements.txt as well for some of my projects. 00:18:15.840 --> 00:18:17.080 Because, you know what? 00:18:17.080 --> 00:18:18.100 The workflow works. 00:18:18.260 --> 00:18:24.420 I've got, like, external systems, like PyUp that are out there, like, automatically doing PRs for changes. 00:18:24.420 --> 00:18:26.280 I mean, it's just, like, it's super cool. 00:18:26.280 --> 00:18:28.680 But at the same time, I already got a flow working. 00:18:28.680 --> 00:18:29.960 And I just, you know. 00:18:29.960 --> 00:18:30.280 Yeah. 00:18:30.280 --> 00:18:31.340 No, I can relate. 00:18:31.340 --> 00:18:35.940 And so a user once, I think, contributed a pyproject.toml to spaCy. 00:18:35.940 --> 00:18:38.600 And, you know, I really appreciated that because I'm like, great. 00:18:38.600 --> 00:18:43.640 But it's still, I don't think, yeah, we can't really ditch requirements.txt yet. 00:18:43.640 --> 00:18:45.800 And so now we also have that. 00:18:45.940 --> 00:18:50.580 And now if we update, like, a dependency, I have to manually edit that in three places. 00:18:50.580 --> 00:18:50.940 Yeah. 00:18:50.940 --> 00:18:51.640 Yeah, yeah. 00:18:51.640 --> 00:18:53.860 That's how it goes. 00:18:53.860 --> 00:18:54.920 I'm on the bandwagon. 00:18:54.920 --> 00:18:56.080 I'm switching. 00:18:56.080 --> 00:18:57.380 I'm using Flit now. 00:18:57.380 --> 00:19:00.760 So I'm using pyproject.toml and Flit. 00:19:00.760 --> 00:19:01.040 Okay. 00:19:01.040 --> 00:19:04.100 And that works cross-platform, cross-Python? 00:19:04.100 --> 00:19:04.780 I don't know. 00:19:04.780 --> 00:19:06.560 It works for me. 00:19:06.560 --> 00:19:07.300 Well, that's always good. 00:19:07.300 --> 00:19:07.640 Yeah. 00:19:07.640 --> 00:19:12.200 I'm using it for mostly 3.6 and above, 3.6, 3.7, 3.8. 00:19:12.200 --> 00:19:14.860 My personal projects, I think it's sufficient. 00:19:15.060 --> 00:19:19.360 And the ones I'm supporting for other people, I think it's fine for an individual project 00:19:19.360 --> 00:19:21.840 owner to say, I'm not supporting 2.7. 00:19:21.840 --> 00:19:22.220 So. 00:19:22.220 --> 00:19:22.820 No, of course. 00:19:22.820 --> 00:19:25.640 Like, I would never, you know, and I also would never go and just, like, whine about, like, 00:19:25.640 --> 00:19:29.580 oh, there's, of course, there's, like, an edge case for 2.7 on Windows. 00:19:29.580 --> 00:19:30.600 And there's something there. 00:19:30.600 --> 00:19:34.620 Like, I understand it's not, like, as easy if, like, you know, there was, like, the Python 00:19:34.620 --> 00:19:36.500 that could just, like, magically fix everything. 00:19:36.500 --> 00:19:36.800 Yeah. 00:19:36.800 --> 00:19:37.700 You know, I appreciate it. 00:19:37.700 --> 00:19:39.360 There's a lot of work that went into all of this. 00:19:39.360 --> 00:19:41.640 And it's like, yeah, something's a bit tricky. 00:19:41.800 --> 00:19:44.880 One of the interesting things was coverage.py. 00:19:44.880 --> 00:19:45.960 I use that a lot. 00:19:45.960 --> 00:19:51.780 And it got, there was a request to put pyproject.toml support on coverage. 00:19:52.180 --> 00:19:56.960 The reason why it isn't there isn't because of any sort of, like, not that it would be 00:19:56.960 --> 00:20:01.060 cool, but the Toml parsing is not part of the standard library. 00:20:01.060 --> 00:20:07.980 And coverage has the strict policy that the only dependencies that it has are standard library 00:20:07.980 --> 00:20:08.460 dependencies. 00:20:08.460 --> 00:20:11.280 That's a reasonable desire also. 00:20:11.280 --> 00:20:11.720 Well, yeah. 00:20:11.720 --> 00:20:12.040 Yeah. 00:20:12.040 --> 00:20:12.260 Yeah. 00:20:12.260 --> 00:20:12.580 That's cool. 00:20:12.580 --> 00:20:15.480 You just get coverage.py and just run the file or whatever, right? 00:20:15.480 --> 00:20:15.720 Yeah. 00:20:15.720 --> 00:20:19.580 But maybe we should get Toml support added to the standard library and then it wouldn't be 00:20:19.580 --> 00:20:19.940 an issue. 00:20:19.940 --> 00:20:20.460 There you go. 00:20:20.580 --> 00:20:20.780 Yeah. 00:20:20.780 --> 00:20:21.080 Yeah. 00:20:21.080 --> 00:20:22.100 That's a whole different discussion. 00:20:22.100 --> 00:20:26.060 I know that's quite a heated debate about what should be on the standard library these 00:20:26.060 --> 00:20:26.340 days. 00:20:26.340 --> 00:20:27.980 And the trend is less, not more. 00:20:27.980 --> 00:20:29.820 I think if you take the poll. 00:20:29.820 --> 00:20:30.760 Yeah. 00:20:30.760 --> 00:20:31.020 Yeah. 00:20:31.020 --> 00:20:32.100 Yeah. 00:20:32.100 --> 00:20:38.800 So some of the tools that you build are absolutely about making machine learning easier and do that 00:20:38.800 --> 00:20:39.260 across teams. 00:20:39.260 --> 00:20:42.220 So I know that you're really turned on to that space and pay a lot of attention. 00:20:42.460 --> 00:20:46.740 So this Polyaxon one that you found for our next item must be pretty interesting. 00:20:46.740 --> 00:20:47.020 Yeah. 00:20:47.020 --> 00:20:50.360 So basically, it's actually quite funny because, you know, I've obviously like, you know, thought 00:20:50.360 --> 00:20:53.320 about what I was going to talk about and had like something else planned. 00:20:53.320 --> 00:20:58.520 And then really today, earlier today, that release came out version 0.5 of Polyaxon. 00:20:58.520 --> 00:20:59.340 So I was like, okay, great. 00:20:59.340 --> 00:21:00.160 This is perfect. 00:21:00.160 --> 00:21:05.660 It's like as if like, you know, they'd waited for my podcast recordings. 00:21:06.260 --> 00:21:06.380 Yeah. 00:21:06.380 --> 00:21:08.360 So basically, we've been using Polyaxon internally. 00:21:08.360 --> 00:21:11.480 And essentially, it's a tool for experiment management. 00:21:11.480 --> 00:21:14.920 So, you know, if you work in machine learning and you train your models, you have to like 00:21:14.920 --> 00:21:16.440 run tons of experiments. 00:21:16.440 --> 00:21:20.540 And you have to, you know, you run an experiment, train a model, look at the results, then you 00:21:20.540 --> 00:21:21.140 stop it. 00:21:21.140 --> 00:21:24.380 Then you tweak some other knobs, then you try again. 00:21:24.380 --> 00:21:26.480 And you keep doing that until you have a good result. 00:21:26.480 --> 00:21:31.020 And one thing we always do when we travel and like visit universities and research labs, 00:21:31.020 --> 00:21:35.060 we usually always ask them like, hey, how do you run your experiments? 00:21:35.060 --> 00:21:39.800 And usually they're like, well, yeah, you know, we got this GPU and it sits on my desk. 00:21:39.800 --> 00:21:43.220 And then I start an experiment and then I sit around and then I wait. 00:21:43.220 --> 00:21:47.880 And, you know, that's at like the top, some of the top like labs and like, you know, people 00:21:47.880 --> 00:21:51.100 where you'd think like, oh, they must have everything taken care of, tons of money. 00:21:51.100 --> 00:21:53.580 It's like, no, they're sitting there with their little GPU on their desk. 00:21:53.580 --> 00:21:56.120 And that's how it's done. 00:21:56.120 --> 00:21:59.460 And basically, Polyaxon basically, you know, helps you solve this. 00:21:59.460 --> 00:22:01.580 So it's like super, you know, it's built on Kubernetes. 00:22:01.580 --> 00:22:03.140 It's like very easy to set up. 00:22:03.400 --> 00:22:07.820 And especially, you know, if you're already set up with cloud computing and yeah, and you 00:22:07.820 --> 00:22:11.420 can also do stuff like hyperparameter search where like, you know, every hyperparameter 00:22:11.420 --> 00:22:13.500 is a tiny knob and you have like tons of them. 00:22:13.500 --> 00:22:17.300 And then you want to find the one combination that gives you better accuracy. 00:22:17.300 --> 00:22:22.340 And so you can run lots of experiments, see them, you know, in their little graphs and like 00:22:22.340 --> 00:22:23.220 try things out. 00:22:23.220 --> 00:22:25.080 So it's been a very great tool. 00:22:25.180 --> 00:22:28.000 It's all open source, which is, you know, very much in our spirit. 00:22:28.000 --> 00:22:35.180 And yeah, they just released 0.5 and which comes with a plugin system, which is also great. 00:22:35.180 --> 00:22:36.240 It's very much in our spirit. 00:22:36.240 --> 00:22:39.260 That's also how we like to do things with spaCy. 00:22:39.260 --> 00:22:40.920 You can run it locally. 00:22:40.920 --> 00:22:44.320 And like, you know, it comes with some new features for chaining stuff together. 00:22:44.320 --> 00:22:46.300 If you have your experiments have lots of steps. 00:22:46.300 --> 00:22:47.980 So yeah, it's a great tool. 00:22:47.980 --> 00:22:49.300 If you're working in the field. 00:22:49.300 --> 00:22:50.120 It looks super cool. 00:22:50.240 --> 00:22:51.200 Yeah, I can definitely recommend it. 00:22:51.200 --> 00:22:51.380 Yeah. 00:22:51.380 --> 00:22:53.220 So it's got a platform as a service offering. 00:22:53.220 --> 00:22:55.560 So you just, you know, kick it off and have it go. 00:22:55.560 --> 00:22:59.180 But also like on-premise enterprise option too. 00:22:59.180 --> 00:22:59.920 Yeah, that's cool. 00:22:59.920 --> 00:23:00.780 I think that's the focus. 00:23:00.780 --> 00:23:04.140 It's like, you know, you run it and then it gives you like a little UI and you just like, 00:23:04.140 --> 00:23:08.340 you know, you set it up yourself on your service and then it manages that. 00:23:08.340 --> 00:23:09.380 Yeah, this looks really great. 00:23:09.380 --> 00:23:12.820 Runs on Google's Kubernetes engine, among other things. 00:23:12.820 --> 00:23:16.520 Probably pretty much any Kubernetes cluster, I guess. 00:23:16.520 --> 00:23:17.500 I haven't tried it. 00:23:17.500 --> 00:23:19.900 Yeah, I think you might have to do a bit more like setup. 00:23:20.200 --> 00:23:24.060 If, you know, you're bringing your own, but like it should be quite very straightforward. 00:23:24.060 --> 00:23:26.880 Google actually makes this quite straightforward. 00:23:26.880 --> 00:23:30.300 And yeah, and another nice thing here is with the hyperparameter search. 00:23:30.300 --> 00:23:35.640 Like if you, yeah, most machine learning stuff is done on GPU, but not everyone has GPUs. 00:23:35.640 --> 00:23:36.380 They're very expensive. 00:23:36.380 --> 00:23:41.220 And we actually say it's not always the best choice necessarily, because if you just want 00:23:41.220 --> 00:23:43.540 to run lots of experiments, you can run them all in parallel. 00:23:43.540 --> 00:23:46.580 You can run like thousands in parallel. 00:23:46.580 --> 00:23:50.360 And then, you know, if you have a tool like Polyaxon that can help you do that. 00:23:50.360 --> 00:23:52.520 So, you know, you don't have to kick them all off manually. 00:23:52.520 --> 00:23:55.840 It's actually going to be much cheaper and much more efficient. 00:23:56.300 --> 00:23:58.220 And you don't need a fancy GPU. 00:23:58.220 --> 00:23:59.760 You can just run it on CPU. 00:23:59.760 --> 00:24:00.100 Right. 00:24:00.100 --> 00:24:00.300 Yeah. 00:24:00.300 --> 00:24:03.000 The GPUs are great, but they're much harder to come by. 00:24:03.000 --> 00:24:04.020 So that's cool. 00:24:04.020 --> 00:24:04.220 Yeah. 00:24:04.220 --> 00:24:04.440 Yeah. 00:24:04.440 --> 00:24:07.200 It also has something about you can run it on your laptop as well. 00:24:07.200 --> 00:24:07.540 Yeah. 00:24:07.540 --> 00:24:10.040 It's a little data science of the box thing they talk about there at the end. 00:24:10.040 --> 00:24:11.520 So, yeah, super cool. 00:24:11.520 --> 00:24:12.400 A nice one. 00:24:12.400 --> 00:24:12.900 All right. 00:24:12.900 --> 00:24:14.920 This last one I want to talk about here. 00:24:14.920 --> 00:24:18.400 Actually, the way I got it onto my system is I use PipX. 00:24:18.400 --> 00:24:20.380 It's my PipX installed. 00:24:20.380 --> 00:24:21.700 This thing called Flint. 00:24:21.700 --> 00:24:26.160 So we've heard about linting and we've heard about f-strings. 00:24:26.160 --> 00:24:31.820 And I'm guessing some combination thereof is where the name of the thing called Flint came 00:24:31.820 --> 00:24:32.000 from. 00:24:32.000 --> 00:24:32.720 It's quite new. 00:24:32.720 --> 00:24:36.280 It's not super popular yet, but it works really, really well. 00:24:36.280 --> 00:24:38.700 So the idea is I've got some code. 00:24:38.700 --> 00:24:41.220 Maybe it's old code. 00:24:41.220 --> 00:24:44.060 Maybe I just haven't bothered to write everything using f-strings. 00:24:44.060 --> 00:24:47.800 And I would like to modernize it in its string processing. 00:24:47.800 --> 00:24:53.160 So this tool, what you can do is you can point it at a single Python file, or you can just point 00:24:53.160 --> 00:24:57.140 it as a directory, like a top level directory to just go to every, you know, traverse the 00:24:57.140 --> 00:25:03.580 whole directory tree and find all the Python files and then rewrite all the string operations 00:25:03.580 --> 00:25:04.780 to be f-strings. 00:25:04.780 --> 00:25:05.320 Nice. 00:25:05.320 --> 00:25:06.640 And it does a pretty good job. 00:25:06.640 --> 00:25:13.780 It'll do the percent, you know, Python 2 style formats, as well as the, you know, dot format 00:25:13.780 --> 00:25:14.260 style. 00:25:14.260 --> 00:25:16.200 And it'll just replace all those with f-strings. 00:25:16.200 --> 00:25:18.100 There's a couple of things it doesn't do. 00:25:18.100 --> 00:25:22.900 If it's like multi-line, really long stuff, it won't replace those. 00:25:22.900 --> 00:25:28.660 And when I first tried it, it actually was making a mistake on digit grouping format. 00:25:28.660 --> 00:25:34.200 So if you have curly braces colon comma, because you want, you know, thousands, millions grouping 00:25:34.200 --> 00:25:37.080 and so on, that like just went insane and broke my code. 00:25:37.080 --> 00:25:41.480 But I submitted a bug over to the guys working on it, fixed it. 00:25:41.480 --> 00:25:43.300 I believe a new release is already out. 00:25:43.300 --> 00:25:47.160 So that shouldn't be there, but just, you know, run it on something you have under source 00:25:47.160 --> 00:25:52.160 control and just look at the lines that have changed before you do the commit or, you know, 00:25:52.160 --> 00:25:53.240 run your task, something like that. 00:25:53.240 --> 00:25:53.400 Yeah. 00:25:53.400 --> 00:25:55.140 Well, you're such a good open source user. 00:25:55.140 --> 00:26:01.260 Like, you know, found a bug, you know, reported it probably with like a nice, you know, description 00:26:01.260 --> 00:26:01.760 fixed. 00:26:01.760 --> 00:26:06.480 I'm like, I ran this on all of Talk Python training and all of Python bytes and some of it broke. 00:26:06.480 --> 00:26:08.860 So what I found out is it's exactly this. 00:26:08.860 --> 00:26:10.160 If it has digit grouping, it broke. 00:26:10.160 --> 00:26:12.480 And so then they fixed it, but it was no big deal. 00:26:12.480 --> 00:26:18.540 I think I converted about 500 to 700 string formats over to f-strings. 00:26:18.540 --> 00:26:21.100 And it just, it's cleaner, shorter, nicer. 00:26:21.100 --> 00:26:26.060 The thing with f-strings is I always, I don't know if you all use it, I'll ask you a sec, 00:26:26.060 --> 00:26:28.560 but like, I'm always like, okay, I'm going to write the string. 00:26:28.640 --> 00:26:30.380 I'd say quote, type, type, type. 00:26:30.380 --> 00:26:31.900 Oh, I want to put something in here. 00:26:31.900 --> 00:26:33.160 Curly brace. 00:26:33.160 --> 00:26:37.460 I wish I would have done the F back, back, back, back, back, back, back, put the F and then 00:26:37.460 --> 00:26:38.360 back, back, back, back, back. 00:26:38.360 --> 00:26:39.020 And then type the thing. 00:26:39.020 --> 00:26:40.900 I'm like, well, that was more work than just dot format. 00:26:40.900 --> 00:26:43.960 Cause the IDE will auto-complete us like dot F and then boom. 00:26:43.960 --> 00:26:48.740 Like, so a lot of times I ended up using the format anyway, but I still prefer to have the 00:26:48.740 --> 00:26:49.540 f-strings and read them. 00:26:49.540 --> 00:26:53.080 So this way I can write it however I want and then just hit it with this before I do a check-in. 00:26:53.240 --> 00:26:58.740 I think that we should ask VS Code and PyCharm to detect when we put a curly brace in a string 00:26:58.740 --> 00:27:00.360 and automatically add the F. 00:27:00.360 --> 00:27:00.740 Yes. 00:27:00.740 --> 00:27:01.500 Just like a hockey. 00:27:01.500 --> 00:27:03.700 I was going to say the exact same thing. 00:27:03.700 --> 00:27:04.060 So it's not. 00:27:04.060 --> 00:27:05.420 That's awesome. 00:27:05.420 --> 00:27:07.020 You all are in the same boat. 00:27:07.020 --> 00:27:08.820 So yeah, this is, this is really cool. 00:27:08.820 --> 00:27:10.720 I definitely think my code is nicer. 00:27:10.720 --> 00:27:14.860 I originally created like when I created the Python bytes website and I created Talk Python 00:27:14.860 --> 00:27:15.220 training. 00:27:15.220 --> 00:27:20.200 This was when the latest version of Python and Ubuntu was three, five. 00:27:20.200 --> 00:27:21.700 So we didn't have f-strings. 00:27:22.000 --> 00:27:25.880 And I actually took the server down once on accident because I used an F string and a 00:27:25.880 --> 00:27:30.200 little utility file that was like in the same directory and like the scanning path, looking 00:27:30.200 --> 00:27:34.000 for the routes, found that, couldn't parse it and the website couldn't start. 00:27:34.000 --> 00:27:35.060 I'm like, why is it down? 00:27:35.060 --> 00:27:35.640 What have I done? 00:27:35.640 --> 00:27:36.620 I didn't even change it. 00:27:36.620 --> 00:27:41.380 So anyway, I'm really happy now that I could just take all that code that I used to like 00:27:41.380 --> 00:27:45.100 leave in the format style and just flint space, you know, just run on this directory. 00:27:45.100 --> 00:27:45.940 Boom, it's done. 00:27:45.940 --> 00:27:47.220 So yeah, it's really nice. 00:27:47.220 --> 00:27:50.680 I can't wait to just like intuitively just use like f-strings and all that stuff. 00:27:50.680 --> 00:27:53.060 It's still kind of, I don't know, just ingrained in like my brain. 00:27:53.060 --> 00:27:56.920 Like even, I know, I go to conferences and I see people use all the new syntax and I'm 00:27:56.920 --> 00:27:58.680 like, yeah, oh, that's so nice. 00:27:58.680 --> 00:28:03.920 But it's just like, you know, in my day to day work, you know, even if we don't support 00:28:03.920 --> 00:28:05.660 two, seven, we support three, five. 00:28:05.660 --> 00:28:07.700 And they're just like a lot of these. 00:28:07.700 --> 00:28:08.560 Yeah, exactly. 00:28:08.560 --> 00:28:12.180 You know, there is this thing called, I can't remember what it was called, Brian. 00:28:12.260 --> 00:28:15.640 We covered it where it lets you add F string support to Python two. 00:28:15.640 --> 00:28:18.380 It may also work for Python three, five. 00:28:18.380 --> 00:28:22.600 You can definitely retroactively add f-strings to the format. 00:28:22.600 --> 00:28:23.560 It's some weird way. 00:28:23.560 --> 00:28:26.380 But then you need like another runtime dependency, which. 00:28:26.380 --> 00:28:26.740 Yeah. 00:28:26.740 --> 00:28:26.960 Yeah. 00:28:26.960 --> 00:28:27.740 It's not worth it. 00:28:27.740 --> 00:28:28.820 It does some weird thing. 00:28:28.820 --> 00:28:29.520 That's how it works. 00:28:29.520 --> 00:28:29.720 Yeah. 00:28:29.720 --> 00:28:30.800 That's a bit unattractive. 00:28:30.880 --> 00:28:35.380 It like rewrites like the file loader with a certain weird encoder. 00:28:35.380 --> 00:28:36.620 It's like, it's pretty sketchy. 00:28:36.620 --> 00:28:36.900 Yeah. 00:28:36.900 --> 00:28:37.180 Okay. 00:28:37.180 --> 00:28:39.640 Now I wouldn't want to ship that in like our libraries. 00:28:39.640 --> 00:28:40.540 Oh, come on. 00:28:40.540 --> 00:28:40.960 Why not? 00:28:40.960 --> 00:28:42.180 Just run your tests. 00:28:42.180 --> 00:28:43.200 Make sure they pass. 00:28:43.200 --> 00:28:43.720 That's good. 00:28:43.720 --> 00:28:44.080 Yeah. 00:28:44.080 --> 00:28:44.440 Good. 00:28:44.440 --> 00:28:47.880 But it's cool that something like this is out so that when you decide like we're no 00:28:47.880 --> 00:28:51.120 longer supporting three, five, you just hit it with this and like, you know, a quick 00:28:51.120 --> 00:28:53.440 scan through the files and it's, it's F stringified. 00:28:53.440 --> 00:28:54.580 That's going to be so, so nice. 00:28:54.580 --> 00:28:55.520 Like I can't wait. 00:28:55.520 --> 00:29:00.200 I'm yet all that and all the type hints and like, you know, once we can drop all of the, 00:29:00.200 --> 00:29:03.420 the older versions, like I can be so, also nice. 00:29:03.420 --> 00:29:06.760 So yeah, I'm really, I won't even mind like rewriting all of the code. 00:29:06.760 --> 00:29:09.980 Like I think with our team, we're just going to sit down and be like, yay, let's do this. 00:29:09.980 --> 00:29:10.700 We're going to do it. 00:29:10.700 --> 00:29:11.080 Yes. 00:29:11.080 --> 00:29:11.420 Here we come. 00:29:11.420 --> 00:29:11.640 Yeah. 00:29:11.640 --> 00:29:12.660 It's going to be so satisfying. 00:29:12.660 --> 00:29:14.080 Like years later. 00:29:14.080 --> 00:29:14.380 Yeah. 00:29:14.380 --> 00:29:16.340 Well, it's the curse of success, right? 00:29:16.340 --> 00:29:20.340 You have so many people using your libraries that you just got to keep it sort of a little 00:29:20.340 --> 00:29:20.760 bit backwards. 00:29:20.760 --> 00:29:21.020 Yeah. 00:29:21.020 --> 00:29:21.260 Sure. 00:29:21.260 --> 00:29:23.180 And like some people are still stuck on legacy code. 00:29:23.180 --> 00:29:26.820 Like, I mean, I'm not, it's not, you know, some people like look down on like companies 00:29:26.820 --> 00:29:29.320 that are still on like Python too, but it's like, you know, it's not 00:29:29.520 --> 00:29:35.180 like many of them like really enjoy using all this legacy software and legacy like stuff. 00:29:35.180 --> 00:29:37.200 They're just like, it just exists. 00:29:37.200 --> 00:29:40.260 And we might as well keep supporting it if we can. 00:29:40.260 --> 00:29:40.520 Yeah. 00:29:40.520 --> 00:29:42.600 What's the Python 2 story for you all? 00:29:42.600 --> 00:29:44.160 You're still supporting it for now? 00:29:44.160 --> 00:29:44.480 Yeah. 00:29:44.480 --> 00:29:45.780 And I think we will for a while. 00:29:45.780 --> 00:29:46.560 After January? 00:29:46.560 --> 00:29:46.940 Yeah. 00:29:47.020 --> 00:29:47.780 We probably will. 00:29:47.780 --> 00:29:52.580 Like there will just naturally be a point where like we cannot upgrade any of our dependencies. 00:29:52.580 --> 00:29:54.120 Like, I don't know, NumPy for example. 00:29:54.120 --> 00:29:54.500 Okay. 00:29:54.500 --> 00:29:59.140 If we ever, you know, it is a good reason why we want to use a newer version of that. 00:29:59.140 --> 00:30:00.320 We just can't. 00:30:00.320 --> 00:30:03.700 And if everyone else drops it, we just have to be like, okay, that's it. 00:30:03.700 --> 00:30:04.200 We can't. 00:30:04.200 --> 00:30:06.320 And it's also not like the old versions are going away. 00:30:06.400 --> 00:30:12.280 Like if we make sure we don't have any major bugs, like you can still use an old version 00:30:12.280 --> 00:30:14.360 of spaCy and like, we're not going to take that away from you. 00:30:14.360 --> 00:30:14.580 Right. 00:30:14.580 --> 00:30:16.320 Just pin the version and you'll be good. 00:30:16.320 --> 00:30:16.540 Yeah. 00:30:16.540 --> 00:30:17.160 Very cool. 00:30:17.160 --> 00:30:17.580 Very cool. 00:30:17.580 --> 00:30:18.100 All right. 00:30:18.100 --> 00:30:19.860 Well, that's it for all of our main items. 00:30:19.860 --> 00:30:22.980 Brian, you want to, I know we've got a few little extra things just thrown here at the 00:30:22.980 --> 00:30:23.080 end. 00:30:23.080 --> 00:30:23.660 Do you want to kick it off? 00:30:23.660 --> 00:30:23.900 Yeah. 00:30:23.960 --> 00:30:26.300 So we had an email from Andre Janisch. 00:30:26.300 --> 00:30:27.880 I think that's how you say his name. 00:30:27.880 --> 00:30:33.320 Saying that we were, in one of our episodes, we talked about regular expressions taking down, 00:30:33.320 --> 00:30:35.720 and I even forgot, took down something. 00:30:35.720 --> 00:30:36.580 It was something major. 00:30:36.580 --> 00:30:36.920 Yeah. 00:30:36.920 --> 00:30:40.140 I can't remember, but some major cloud provider went down because of it. 00:30:40.140 --> 00:30:40.300 Yeah. 00:30:40.300 --> 00:30:41.740 And how that could happen. 00:30:41.740 --> 00:30:46.960 So there's an interesting video talking about regular expression denial of service attacks 00:30:46.960 --> 00:30:48.420 and how it happens. 00:30:48.420 --> 00:30:49.800 It was just an interesting video. 00:30:49.800 --> 00:30:52.620 We'll have a link to it in the show notes if anybody wants to watch. 00:30:52.620 --> 00:30:52.940 Okay. 00:30:52.940 --> 00:30:53.260 Yeah. 00:30:53.360 --> 00:30:53.540 Yeah. 00:30:53.540 --> 00:30:54.040 Super cool. 00:30:54.040 --> 00:30:55.900 So I got a couple I want to throw out there. 00:30:55.900 --> 00:31:02.980 One is if you're doing any work where you're working with like microservices or you want to 00:31:02.980 --> 00:31:07.700 have some kind of application that's talking to some API endpoint, you want to debug it. 00:31:07.700 --> 00:31:13.760 There's a new thing called HTTP toolkit, and it has like special Python support. 00:31:13.760 --> 00:31:19.260 So this is like a proxy you can run on your computer and say, start recording, and it'll 00:31:19.260 --> 00:31:21.800 start recording all the requests that you're making. 00:31:22.140 --> 00:31:29.260 So it integrates with url.lib.request, url.lib2, request, pip, Python 2 and 3, photo, all those 00:31:29.260 --> 00:31:32.860 things, and specifically catches traffic from those. 00:31:33.260 --> 00:31:39.020 And it does interesting stuff by changing the Python path and environment variables. 00:31:39.020 --> 00:31:44.760 And then all of these libraries that I talked about apparently respect certain proxy settings 00:31:44.760 --> 00:31:45.640 and things you can set. 00:31:45.640 --> 00:31:47.520 So you don't have to change your app at all. 00:31:47.520 --> 00:31:52.260 You just start a terminal with HTTP toolkit and then run your code, and it can record it. 00:31:52.260 --> 00:31:52.680 That's cool. 00:31:52.880 --> 00:31:53.000 Yeah. 00:31:53.000 --> 00:31:55.240 So if you're like, why is this crashing? 00:31:55.240 --> 00:31:59.980 You know, I run my request thing, and then I get some kind of crash. 00:31:59.980 --> 00:32:03.900 And how am I supposed to, you know, you don't have like the developer tools in your web browser 00:32:03.900 --> 00:32:05.820 to like look at the headers and whatnot. 00:32:05.820 --> 00:32:07.060 So you can do it with this. 00:32:07.060 --> 00:32:07.500 It's pretty cool. 00:32:07.500 --> 00:32:09.800 It's free, but there's also a paid version. 00:32:09.800 --> 00:32:10.400 Just heads up. 00:32:10.560 --> 00:32:13.300 Also, there's a nice little link. 00:32:13.300 --> 00:32:16.880 Last time, time before, Brian, we were talking about magic, Python magic. 00:32:16.880 --> 00:32:22.020 You're like, well, that's a pretty strong, strong name to be magic. 00:32:22.020 --> 00:32:23.800 And all it does is detect file types. 00:32:23.800 --> 00:32:24.780 Right? 00:32:24.780 --> 00:32:25.540 Do you remember that thing? 00:32:25.540 --> 00:32:32.140 David Martinez said, well, the reason it's called magic is there's basically these magic 00:32:32.140 --> 00:32:36.360 number signatures that appear at the beginning of files. 00:32:36.360 --> 00:32:39.100 And that's like teaches you about the syntax. 00:32:39.320 --> 00:32:44.120 So for example, if you had a SQLite file, it would start with 53, 51, 4C, 69, et cetera. 00:32:44.120 --> 00:32:47.220 If you see those numbers up the front, that means SQLite, right? 00:32:47.220 --> 00:32:49.100 So wait, did you actually know this by heart? 00:32:49.100 --> 00:32:50.660 Or did you write this down? 00:32:50.660 --> 00:32:51.920 No, no. 00:32:51.920 --> 00:32:53.340 I don't know. 00:32:53.340 --> 00:32:54.620 Oh, of course. 00:32:54.620 --> 00:32:55.360 Yeah, I know that one. 00:32:55.360 --> 00:32:55.660 Sure. 00:32:55.660 --> 00:32:56.860 No, I just, I pulled it up. 00:32:56.860 --> 00:32:58.240 Let's see. 00:32:58.240 --> 00:33:01.160 Really quickly, Python 374 is out. 00:33:01.160 --> 00:33:06.660 So I brew and upgraded my Python 37, which is how I'm getting on my machine. 00:33:06.660 --> 00:33:08.580 And it's already on homebrew as well. 00:33:08.700 --> 00:33:11.040 So that was like, I don't know, six, seven hours delay there. 00:33:11.040 --> 00:33:12.640 It was really nice to see that come out real quick. 00:33:12.640 --> 00:33:15.360 Or if you have pyenv, you can do pyenv. 00:33:15.360 --> 00:33:18.620 I don't even know the command, but like install something and then... 00:33:18.620 --> 00:33:19.560 Right away have it. 00:33:19.560 --> 00:33:20.140 That's awesome. 00:33:20.140 --> 00:33:20.540 Yeah. 00:33:20.540 --> 00:33:24.280 And then the last one, I want to point out this thing called flying. 00:33:24.280 --> 00:33:25.920 I just call it flying fractals. 00:33:26.460 --> 00:33:31.120 This person put together this project using Py Wonderland. 00:33:31.120 --> 00:33:35.880 And it automates some other libraries that are based on C. 00:33:35.880 --> 00:33:42.800 So they made some really cool videos of like flying through three-dimensional, like animated 00:33:42.800 --> 00:33:45.260 Mandelbrot sets and other kinds of stuff. 00:33:45.320 --> 00:33:50.500 When I just watched it, I thought just, wow, this is super cool computational stuff over 00:33:50.500 --> 00:33:50.760 here. 00:33:50.760 --> 00:33:54.600 I've done a bunch of work in like complex dynamics and like trying to visualize that. 00:33:54.600 --> 00:33:56.320 And like this kind of blew me away. 00:33:56.320 --> 00:33:59.640 So if you care at all about that stuff, I think you'll just enjoy like a minute of that video. 00:33:59.860 --> 00:34:00.120 Yeah. 00:34:00.120 --> 00:34:00.500 Yeah. 00:34:00.500 --> 00:34:04.600 So yeah, actually, you might have actually noticed that my voice still isn't like 100% 00:34:04.600 --> 00:34:06.500 and it's a bit rough. 00:34:06.500 --> 00:34:11.820 And that's because last weekend we had our very, very first conference here in Berlin called 00:34:11.820 --> 00:34:14.500 spaCy IRL, like, you know, spaCy in real life. 00:34:14.500 --> 00:34:18.960 So our community came together and I'm still like absolutely blown away. 00:34:18.960 --> 00:34:20.700 Like there was, you know, the vibe was amazing. 00:34:20.700 --> 00:34:26.160 We had like 200 people there and a lovely like old theater, 12 really amazing talks by people 00:34:26.160 --> 00:34:27.940 from research, industry, community. 00:34:28.620 --> 00:34:33.100 So it was really a lot of fun and all talks were recorded and we're currently uploading 00:34:33.100 --> 00:34:34.240 them to our YouTube channel. 00:34:34.240 --> 00:34:37.480 So probably by the time this airs, they might've already been released. 00:34:37.480 --> 00:34:42.000 And yeah, it'd be a bunch of, yeah, if you're interested in natural language processing, spaCy, 00:34:42.000 --> 00:34:43.740 those, the talks were really, really great. 00:34:43.740 --> 00:34:46.340 And yeah, I hope you, you enjoy watching them. 00:34:46.340 --> 00:34:46.820 That's awesome. 00:34:46.820 --> 00:34:50.260 I'm so glad you're putting them on YouTube and wow, congratulations. 00:34:50.260 --> 00:34:55.600 You must be just blown away at how awesome it is to put on a conference about your own stuff, 00:34:55.600 --> 00:34:55.800 right? 00:34:55.800 --> 00:34:58.560 And so many people came and the energy and like you could tell your voices. 00:34:58.560 --> 00:34:59.280 And it's great. 00:34:59.280 --> 00:35:00.600 I know that was like, yeah. 00:35:00.600 --> 00:35:01.120 Wow. 00:35:01.120 --> 00:35:04.060 And it was also, it's actually quite refreshing to organize your own conference. 00:35:04.060 --> 00:35:08.300 Like, you know, we were like, okay, let's do all the things that we think a conference 00:35:08.300 --> 00:35:09.800 should do and like try them out. 00:35:09.800 --> 00:35:11.220 And it actually worked quite well. 00:35:11.220 --> 00:35:13.760 Like, you know, only one track, for example, stuff like that. 00:35:13.760 --> 00:35:19.840 No questions from the audience, much more social time, some, you know, really healthy food, 00:35:19.840 --> 00:35:20.500 stuff like that. 00:35:20.540 --> 00:35:21.700 So it was, it was, it was great. 00:35:21.700 --> 00:35:25.720 Like my voice suffered a bit, but you know, we're giving a lot back to the community as 00:35:25.720 --> 00:35:27.540 in the videos, photos, stuff. 00:35:27.540 --> 00:35:28.160 So yeah. 00:35:28.160 --> 00:35:29.220 Sounds great. 00:35:29.220 --> 00:35:29.560 Awesome. 00:35:29.560 --> 00:35:30.140 Congratulations. 00:35:30.140 --> 00:35:35.360 A couple of episodes ago, Brian actually talked about FastAPI, not to be confused with fast 00:35:35.360 --> 00:35:37.320 AI, which I talked about earlier. 00:35:37.880 --> 00:35:42.500 So yeah, it's a great, very modern, cool Python library for FastAPIs. 00:35:42.500 --> 00:35:44.980 And we're like, yeah, we've been big fans. 00:35:44.980 --> 00:35:46.900 We've started switching all our APIs over. 00:35:46.900 --> 00:35:52.580 The exciting news here is that their core developer, Sebastian Ramirez is actually going to join our 00:35:52.580 --> 00:35:53.440 team here in Berlin. 00:35:53.440 --> 00:35:58.280 This means like a lot of cool development for us, but also we obviously, since we love the 00:35:58.280 --> 00:36:03.080 FastAPI project, we'll keep supporting that and we'll definitely give him enough time to 00:36:03.080 --> 00:36:03.780 keep working on it. 00:36:03.780 --> 00:36:04.480 That's great news. 00:36:04.480 --> 00:36:05.580 We found him through the project. 00:36:05.580 --> 00:36:07.580 We saw, oh, he's doing like some consulting work. 00:36:07.660 --> 00:36:09.760 We're like, hey, we'd love to work with him. 00:36:09.760 --> 00:36:13.660 And one thing came to another and now, yeah, he'd be part of our Explosion team, which is 00:36:13.660 --> 00:36:14.800 still growing, by the way. 00:36:14.800 --> 00:36:19.440 So we've been very lucky that we were able to work with more people and expand our team. 00:36:19.440 --> 00:36:19.740 Yeah. 00:36:19.740 --> 00:36:22.020 That's so cool that, yeah, that your business is growing. 00:36:22.020 --> 00:36:28.560 And I guess it's worth pointing out that back on Talk Python To Me, we talked about Explosion 00:36:28.560 --> 00:36:28.800 AI. 00:36:28.800 --> 00:36:31.100 I interviewed you about building a software business. 00:36:31.100 --> 00:36:32.760 So back on episode 202. 00:36:32.760 --> 00:36:36.360 So, you know, this is just like more, more evidence that that's all good advice. 00:36:36.360 --> 00:36:36.700 Thanks. 00:36:36.860 --> 00:36:37.780 Yeah, super cool. 00:36:37.780 --> 00:36:39.380 Well, great, great news. 00:36:39.380 --> 00:36:42.560 I guess it's probably time for a joke here too. 00:36:42.560 --> 00:36:44.500 A pie joke, if you will, maybe. 00:36:44.500 --> 00:36:46.740 So I think this, I'll do the first one. 00:36:46.740 --> 00:36:48.460 This one, I think came from pie jokes. 00:36:48.460 --> 00:36:48.960 We'll see. 00:36:48.960 --> 00:36:50.740 We're starting to run that well dry. 00:36:50.740 --> 00:36:52.460 So people send in your jokes, please. 00:36:52.460 --> 00:36:57.960 But a programmer walks into the bar and orders 1.38 root beers. 00:36:58.500 --> 00:37:02.720 The bartender informs her that it's a root beer float. 00:37:02.720 --> 00:37:04.240 She says, no, make it a double. 00:37:04.240 --> 00:37:05.480 All right. 00:37:05.480 --> 00:37:06.500 It's pretty bad type system. 00:37:06.500 --> 00:37:08.060 Maybe it doesn't work so well in Python. 00:37:08.060 --> 00:37:10.240 Like we don't care so much about types, but you know, still. 00:37:10.240 --> 00:37:11.000 So it's all right. 00:37:11.000 --> 00:37:11.560 I like it. 00:37:11.560 --> 00:37:11.940 All right. 00:37:11.980 --> 00:37:12.700 We got one more up here. 00:37:12.700 --> 00:37:13.720 Who put this one in? 00:37:13.720 --> 00:37:13.960 Brian? 00:37:13.960 --> 00:37:14.320 Yeah. 00:37:14.320 --> 00:37:19.760 So just last night I was researching for this podcast, writing notes for the other podcast, 00:37:19.760 --> 00:37:22.680 and working on an open source project. 00:37:22.680 --> 00:37:24.500 And I came up with this. 00:37:24.500 --> 00:37:27.140 What do you call a developer without a side project? 00:37:27.140 --> 00:37:27.540 What's that? 00:37:27.540 --> 00:37:28.160 Well rested. 00:37:28.160 --> 00:37:28.940 Yeah, that's true. 00:37:28.940 --> 00:37:29.840 That's totally true. 00:37:29.840 --> 00:37:31.740 It's almost a bit sad in that sense. 00:37:31.740 --> 00:37:35.360 Yeah. 00:37:35.360 --> 00:37:36.740 I know it's too real. 00:37:36.740 --> 00:37:37.200 Too real. 00:37:37.200 --> 00:37:40.760 I personally shouldn't be complaining. 00:37:40.760 --> 00:37:42.080 That's right. 00:37:42.080 --> 00:37:45.580 But you know, there are some comments we could make here about like, you know, the culture 00:37:45.580 --> 00:37:49.280 and like what's expected of developers these days and how that's maybe not ideal. 00:37:49.280 --> 00:37:51.820 You know, sleeping enough, stuff like that. 00:37:51.820 --> 00:37:52.480 Oh, definitely. 00:37:52.480 --> 00:37:53.660 People should be. 00:37:53.660 --> 00:37:56.040 And I was, it's mostly a self-reflection. 00:37:56.040 --> 00:37:57.580 No, I mean, it is a good joke. 00:37:57.580 --> 00:37:59.040 I'm not, I wasn't criticizing your joke. 00:37:59.040 --> 00:38:00.920 It's a totally fine joke. 00:38:00.920 --> 00:38:02.360 I'm just saying it's also, it's very real. 00:38:02.360 --> 00:38:05.300 You know how jokes can be like too real and then you're like, oh. 00:38:05.300 --> 00:38:09.080 Yeah, I'm really uncomfortable now. 00:38:09.080 --> 00:38:11.800 Exactly. 00:38:11.800 --> 00:38:12.080 All right. 00:38:12.080 --> 00:38:13.680 Well, I think that's a good place to leave it. 00:38:13.680 --> 00:38:14.780 Brian, thank you as always. 00:38:14.780 --> 00:38:15.080 Thank you. 00:38:15.080 --> 00:38:16.160 It was great to have you here. 00:38:16.160 --> 00:38:16.700 Thanks for coming. 00:38:16.700 --> 00:38:17.260 Yeah, thanks. 00:38:17.260 --> 00:38:17.620 Bye. 00:38:17.620 --> 00:38:17.860 Bye. 00:38:17.860 --> 00:38:19.640 Thank you for listening to Python Bytes. 00:38:19.640 --> 00:38:22.200 Follow the show on Twitter via at Python Bytes. 00:38:22.200 --> 00:38:25.040 That's Python Bytes as in B-Y-T-E-S. 00:38:25.040 --> 00:38:28.280 And get the full show notes at Pythonbytes.fm. 00:38:28.280 --> 00:38:32.500 If you have a news item you want featured, just visit Pythonbytes.fm and send it our way. 00:38:32.500 --> 00:38:35.200 We're always on the lookout for sharing something cool. 00:38:35.200 --> 00:38:38.300 On behalf of myself and Brian Okken, this is Michael Kennedy. 00:38:38.300 --> 00:38:41.740 Thank you for listening and sharing this podcast with your friends and colleagues.