WEBVTT 00:00:00.020 --> 00:00:05.300 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to earbuds. 00:00:05.860 --> 00:00:12.960 This is episode 466, recorded Monday, January 19th, 2026. 00:00:13.130 --> 00:00:13.740 I'm Michael Kennedy. 00:00:14.060 --> 00:00:14.840 And I'm Brian Okken. 00:00:15.120 --> 00:00:17.380 This episode is brought to you by us, all of our things. 00:00:17.570 --> 00:00:22.440 I know Brian is very lean these days with his TDD and would like to tell you about it more on that later, 00:00:22.680 --> 00:00:25.740 but he's got his upcoming book and course. 00:00:26.440 --> 00:00:30.440 I've got my Talk Python in production book, which is still going strong. 00:00:30.570 --> 00:00:31.100 I love that. 00:00:31.720 --> 00:00:34.300 Of course, all of our courses, things like that. 00:00:34.860 --> 00:00:36.460 Also, connect with us on social media. 00:00:36.900 --> 00:00:42.260 Links for most of the places that you might do so in the episode show notes to your podcast player and Sean. 00:00:43.040 --> 00:00:45.000 And finally, subscribe to the newsletter. 00:00:45.560 --> 00:00:47.100 We're putting awesome things in there. 00:00:47.400 --> 00:00:49.400 Lots more than just the show notes. 00:00:50.040 --> 00:00:54.560 So it's kind of a nice reinforcement, a few extra resources if you like them. 00:00:54.820 --> 00:00:57.440 Yeah. Mostly put together by Brian. So we appreciate that, Brian. 00:00:57.880 --> 00:00:57.980 Thanks. 00:00:58.580 --> 00:01:05.540 Yeah, you bet. I also appreciate a good web framework with a solid community. What do you got in that realm? 00:01:06.560 --> 00:01:22.320 Well, we're going to talk about Django. So put this up here. Django, I ran across this article about that's six months old from the revsys. It's from June of 2025. But I had it in my queue and I'm going to cover it. 00:01:22.460 --> 00:01:26.800 So better Django management commands with Django click and Django typer. 00:01:27.100 --> 00:01:29.380 And this is from Lacey Henschel. 00:01:30.080 --> 00:01:33.960 And so let's look at the management commands. 00:01:34.300 --> 00:01:37.000 If anybody that's used Django, you know, there's management commands. 00:01:37.140 --> 00:01:37.900 There's a bunch of them. 00:01:38.080 --> 00:01:44.500 So there's, there's, you just say like, man, you've got, if you're in the Django thing, 00:01:44.600 --> 00:01:50.540 you can go manage.py command, or you can do Django admin command or Python dash M Django, 00:01:51.180 --> 00:01:53.700 then the command, there's a bunch of built-in ones. 00:01:54.380 --> 00:02:00.480 And since they're things like just the maintenance around your website stuff and other things, 00:02:00.740 --> 00:02:04.560 but there's a lot of stuff there. 00:02:04.860 --> 00:02:08.020 And it's so convenient that it'd be cool if you could add your own. 00:02:08.280 --> 00:02:10.100 And there is built-in stuff. 00:02:10.140 --> 00:02:14.720 You can create custom commands, but you do it by deriving. 00:02:14.760 --> 00:02:21.260 you have to derive from a base command class and then have self everywhere. 00:02:21.590 --> 00:02:26.140 And there's the object-oriented part of that that's weird. 00:02:26.680 --> 00:02:30.140 But there's better ways to do, there are at least different ways. 00:02:30.320 --> 00:02:32.500 If this doesn't jive with you, there's better ways. 00:02:32.750 --> 00:02:33.780 And that's what we're talking about. 00:02:34.640 --> 00:02:36.840 This is with Django Click and Django Typer. 00:02:37.140 --> 00:02:41.120 So first off, one of the cool things is she runs down 00:02:41.340 --> 00:02:43.160 why you would want to add your own management commands. 00:02:43.600 --> 00:02:47.400 And there's a lot of great reasons like data operations, importing. 00:02:47.850 --> 00:02:52.560 Like she has an example of importing CSV file from clients, exporting data for reports. 00:02:53.320 --> 00:02:54.120 Those are great things. 00:02:56.099 --> 00:03:03.400 You're doing it on the system, but you can just go use the same internals of Django to be able to pull things out. 00:03:03.580 --> 00:03:03.900 That's great. 00:03:04.320 --> 00:03:06.440 Also, development and debugging, one of my favorite reasons. 00:03:07.180 --> 00:03:07.340 Okay. 00:03:07.830 --> 00:03:10.240 So first off, Django Click. 00:03:11.120 --> 00:03:16.840 And Django Click is based on Click, of course. 00:03:17.220 --> 00:03:20.620 And Click is a great way to do CLI tools too. 00:03:21.100 --> 00:03:26.740 And so it's just an integration of Click and Django to get these management commands. 00:03:27.050 --> 00:03:28.700 And you end up just creating a file. 00:03:29.010 --> 00:03:34.840 And it's a lot shorter with some click options and stuff. 00:03:34.870 --> 00:03:36.620 And it's just that integration of Click. 00:03:37.260 --> 00:03:41.400 And then, again, Typer is also built on click also. 00:03:42.200 --> 00:03:44.800 But if you're used to Typer, grab Typer. 00:03:44.800 --> 00:03:50.100 But one of the cool things about Typer is the coloring and the different, the output. 00:03:50.360 --> 00:03:52.040 You can do more structured output. 00:03:52.520 --> 00:04:00.660 So they have an example where you have better things like emoji support and tables and things like that. 00:04:00.660 --> 00:04:05.139 So if you want to do cool, and the article goes in and says, 00:04:06.440 --> 00:04:08.440 hey, well, which one should you use? 00:04:08.860 --> 00:04:11.060 I'm not going to scroll to the bottom, but which one should you use? 00:04:11.120 --> 00:04:16.140 And it's basically, she says, she usually reaches for click, Django click, 00:04:16.440 --> 00:04:21.200 but when she wants to do nice reports and stuff, do Django type her. 00:04:21.600 --> 00:04:23.620 And they're both, I wrote this down, 00:04:24.420 --> 00:04:28.720 they're both supported under, what is it? 00:04:28.760 --> 00:04:37.680 The Django Commons group on GitHub supports both of these extra plugins or whatever. 00:04:38.140 --> 00:04:39.240 So pretty cool. 00:04:39.640 --> 00:04:40.300 Yeah, it was really nice. 00:04:40.820 --> 00:04:43.920 Click and Typer are obviously super popular. 00:04:44.400 --> 00:04:48.080 And if you're like, I know one of those, but I want to add a custom Django command, 00:04:48.260 --> 00:04:49.060 like, let's just do it. 00:04:49.200 --> 00:04:52.960 You know, I imagine I say this, as I say this, I'm sure it's out there, 00:04:53.000 --> 00:04:54.960 but I have no awareness of its existence. 00:04:55.560 --> 00:05:06.940 It would also be cool to have a textual sort of UI if you could just type manage pi-ish, manage tx or whatever, and you get like a UI for all these managed Django commands. 00:05:07.340 --> 00:05:09.040 That'd be pretty neat, including your custom ones. 00:05:09.480 --> 00:05:10.100 Probably out there. 00:05:10.500 --> 00:05:11.320 I have no idea about it. 00:05:11.420 --> 00:05:13.660 So maybe someone will send us a message where you can do a follow-up. 00:05:13.880 --> 00:05:16.700 You know, it's the age-old question, Brian. 00:05:16.880 --> 00:05:19.240 What would you do if you had a million dollars? 00:05:19.350 --> 00:05:21.000 Think of a million dollars. 00:05:21.210 --> 00:05:24.920 You never have to be subservient to society again. 00:05:25.360 --> 00:05:26.480 Well, maybe it's not quite that way. 00:05:26.940 --> 00:05:28.260 But we have really... 00:05:28.360 --> 00:05:29.800 I think I'd need one and a half. 00:05:30.180 --> 00:05:31.660 You would need one and a half because inflation. 00:05:31.960 --> 00:05:33.540 You would actually need more than one and a half probably. 00:05:33.610 --> 00:05:36.280 But you could sure make a dent with one and a half million dollars. 00:05:36.580 --> 00:05:45.620 And that is really good news because Anthropic has just invested $1.5 million in the PSF and open source security. 00:05:45.900 --> 00:05:46.380 That's awesome. 00:05:46.740 --> 00:05:49.760 I want to take a minute and just, you know, thank you, Anthropic. 00:05:50.940 --> 00:05:52.760 People hate on AI companies. 00:05:53.600 --> 00:05:56.400 So I thought it was the folks that bring us Claude, right? 00:05:56.660 --> 00:05:58.920 Yeah, Claude, Opus, Claude Sonnet, Claude Code. 00:05:59.480 --> 00:06:03.820 And honestly, I've been using Claude, I think it's Claude AI, 00:06:04.360 --> 00:06:06.180 as like the ChatGPT alternative. 00:06:06.440 --> 00:06:09.260 Chat GPT has gotten really less good lately, and I don't understand. 00:06:09.680 --> 00:06:10.060 It's weird. 00:06:10.640 --> 00:06:13.100 Like I was just writing a blog post, which I'll talk about later, 00:06:13.540 --> 00:06:16.500 and it's just doing super weird things, and I found like Claude is doing a lot. 00:06:16.580 --> 00:06:19.200 So I think it's honestly one of the better AIs. 00:06:19.620 --> 00:06:21.559 And if all these companies are going to make all this money 00:06:21.560 --> 00:06:23.080 and exchange it around. 00:06:23.230 --> 00:06:25.480 I really like to see Anthropic 00:06:26.080 --> 00:06:28.560 putting this money back into Python. 00:06:29.180 --> 00:06:29.580 You know, they're, 00:06:29.700 --> 00:06:31.380 they're, it's not just Python too. 00:06:31.480 --> 00:06:32.660 They just bought Bun, 00:06:33.140 --> 00:06:34.660 which is like, as far as I understand, 00:06:34.730 --> 00:06:36.040 it kind of like a node alternative, 00:06:36.620 --> 00:06:38.660 certainly a JavaScript runtime thing 00:06:38.820 --> 00:06:39.800 for a ton of money as well. 00:06:39.980 --> 00:06:41.140 So they're, they're doing, 00:06:41.260 --> 00:06:41.600 I mean, they're not, 00:06:41.690 --> 00:06:42.680 they didn't buy the PSF. 00:06:43.010 --> 00:06:44.160 They just gave money. 00:06:44.190 --> 00:06:45.860 So this is over two years. 00:06:46.280 --> 00:06:47.040 So what is that? 00:06:47.560 --> 00:06:48.680 PSF isn't for sale. 00:06:48.840 --> 00:06:50.000 It's not like Greenland or anything. 00:06:51.560 --> 00:07:02.200 Well, I mean, I think we know that because they're turning down of that, probably not a coincidence, $1.5 million offered by the recent U.S. government with strings attached. 00:07:03.020 --> 00:07:05.040 This one comes without strings, more or less. 00:07:05.120 --> 00:07:06.840 I'm sure there are strings, but not to the same degree. 00:07:07.640 --> 00:07:09.520 But it's really $750,000 a year. 00:07:10.180 --> 00:07:15.280 And it has, let me see, I took some notes on exactly the goals of this. 00:07:15.380 --> 00:07:20.140 but it's basically mostly for Python security from YRSTM, 00:07:20.260 --> 00:07:25.040 but also to fund things like the developer in residence initiative, 00:07:25.760 --> 00:07:28.500 community grants, which remember there was a big kerfuffle 00:07:28.860 --> 00:07:32.720 about grants being paused or canceled because they ran out of money, right? 00:07:32.780 --> 00:07:35.620 I mean, it sucks, and maybe they promised it and then didn't deliver it. 00:07:35.620 --> 00:07:37.620 I don't remember. I feel like there was something like that. 00:07:37.800 --> 00:07:41.160 But even if you promised it, if the money's gone, it's not like, 00:07:41.240 --> 00:07:42.440 well, you work for the PSF. 00:07:42.600 --> 00:07:45.540 we have your 401k so we can keep our promise. 00:07:45.780 --> 00:07:47.440 You know, at some point, if the money's not there, 00:07:47.460 --> 00:07:48.040 the money's not there. 00:07:48.220 --> 00:07:50.240 So this is really awesome that this is back. 00:07:51.060 --> 00:07:54.360 Infrastructure like PyPI, like I say this every now and then, 00:07:54.360 --> 00:07:56.320 but I'm not sure how many people understand 00:07:56.800 --> 00:07:58.900 how expensive PyPI is to run. 00:07:59.200 --> 00:08:03.340 If it were not for fast to leave the CDN giving free bandwidth, 00:08:04.380 --> 00:08:07.320 it's $100,000 or more a month. 00:08:07.460 --> 00:08:09.980 I mean, it's a lot of money, right? 00:08:10.160 --> 00:08:11.700 Compared to the overall budget. 00:08:12.300 --> 00:08:14.220 And this is a lot of money compared to the overall budget. 00:08:14.250 --> 00:08:16.840 I feel like the budget of the PSF is around $7 million. 00:08:17.300 --> 00:08:18.780 I mean, I haven't checked in the last couple of years, 00:08:18.960 --> 00:08:20.360 but last time I heard it was something like that. 00:08:20.880 --> 00:08:23.040 So this is a significant portion of that. 00:08:23.440 --> 00:08:23.840 Anything else? 00:08:24.030 --> 00:08:29.320 Yeah, the current CDN cost, last article I found, is $1.5 million a month. 00:08:29.540 --> 00:08:29.800 A month? 00:08:30.380 --> 00:08:31.460 That can't be right. 00:08:32.280 --> 00:08:33.500 I feel like that's high. 00:08:33.840 --> 00:08:36.360 I mean, I know it's high, but I feel like even that's high. 00:08:36.659 --> 00:08:37.640 It might be a year. 00:08:38.870 --> 00:08:39.479 But I don't know. 00:08:39.570 --> 00:08:40.099 I could be wrong. 00:08:40.620 --> 00:08:46.980 There's actually a really nice article written up somewhere that I didn't make any part of our news about the stats of the PSF. 00:08:47.320 --> 00:08:51.120 But I just heard it indirectly, so I don't have it to pull up. 00:08:51.260 --> 00:09:01.260 Anyway, so one of the goals is to fund new automated tools for proactively reviewing all packages uploaded to PyPI, moving beyond a reactive only review process. 00:09:01.440 --> 00:09:07.600 Remember a few weeks ago, I talked about, hey, here's how you set up like Python supply chain security. 00:09:07.820 --> 00:09:14.640 And one of the main ways, honestly, is wait a little while to install it and then check if a problem has been reported. 00:09:15.340 --> 00:09:16.460 It's kind of that reactive thing. 00:09:16.480 --> 00:09:22.800 It's like, well, probably somebody, if something happens to a major package, will figure it out within a few days and react to it. 00:09:22.940 --> 00:09:26.020 So wouldn't it be nice if that didn't have to happen that way? 00:09:26.020 --> 00:09:26.360 It would. 00:09:26.940 --> 00:09:28.240 You know, that is how it happens, right? 00:09:29.879 --> 00:09:34.680 And the PSF plans to build a new data set of new malware for capabilities analysis. 00:09:35.000 --> 00:09:36.180 So all sorts of good stuff. 00:09:36.820 --> 00:09:40.420 I just wanted to shine a light on this, say thank you, Anthropic. 00:09:41.020 --> 00:09:45.020 And yeah, it's really great for PSF and everyone, I think. 00:09:45.220 --> 00:09:50.960 Well, I did notice this, but also a couple of people mentioned it to us. 00:09:51.420 --> 00:09:56.100 An article by Andrew Nesbitt, how uv got so fast. 00:09:56.480 --> 00:10:00.620 And there's a lot of reasons why I like this article. 00:10:01.620 --> 00:10:04.160 First off, just right off the top of the bat, this says, 00:10:04.320 --> 00:10:08.020 says, hey, usual explanation is it's written in Rust. 00:10:08.140 --> 00:10:09.920 And that's kind of what we think is, right? 00:10:10.520 --> 00:10:13.360 Is like you can take Python tools, rewrite them in Rust, 00:10:13.360 --> 00:10:14.020 and they'll be faster. 00:10:14.240 --> 00:10:15.580 But it's a lot more than that. 00:10:16.340 --> 00:10:18.940 The Rust part is contributing, of course. 00:10:21.840 --> 00:10:24.060 But there's some tweaks around it 00:10:24.340 --> 00:10:28.040 and also some design decisions that make it a lot faster. 00:10:28.800 --> 00:10:29.800 And it's interesting to read. 00:10:30.000 --> 00:10:32.660 So one of the things is around the standards. 00:10:33.060 --> 00:10:35.980 So there's a bunch of standards that came in to make this possible. 00:10:36.200 --> 00:10:39.720 There's 518 that created pyproject.toml. 00:10:40.300 --> 00:10:43.000 So uv reads pyproject.toml. 00:10:43.000 --> 00:10:44.260 It doesn't read setup.py. 00:10:44.290 --> 00:10:45.600 You can't do any other stuff. 00:10:45.670 --> 00:10:46.560 You have to do wheels. 00:10:46.780 --> 00:10:48.640 That's what uv looks for. 00:10:50.660 --> 00:10:52.440 Also, actually, I could be wrong. 00:10:52.700 --> 00:10:54.220 I think I'm overstating that. 00:10:55.000 --> 00:10:59.840 I think it does do backwards compatible stuff, and it just falls back to different modes. 00:11:00.660 --> 00:11:02.560 Well, I'll have to double check that, but I think so. 00:11:03.100 --> 00:11:11.700 Anyway, PEP 517 came in in 2017, which is a separate build front ends and back ends. 00:11:12.940 --> 00:11:16.280 So now even PEP doesn't need to understand set of tools internals. 00:11:16.900 --> 00:11:24.020 There's PEP 621 that standardized the bracket project table within the TOML file and reading 00:11:24.140 --> 00:11:25.320 tables, TOMLs. 00:11:26.020 --> 00:11:34.200 And then there's the PEP 658, which was a package metadata directly in the simple repository API. 00:11:35.160 --> 00:11:41.980 Essentially, you don't the tools on the back when you do install something, they don't have to grab a whole bunch of data. 00:11:42.070 --> 00:11:46.760 They just have to grab like this, this dependency information, which they used to have to. 00:11:47.200 --> 00:11:56.280 There's more of a story here about how pip used to have to go out and just really download, try stuff, and then try to run it. 00:11:56.280 --> 00:11:57.880 If it didn't work, try something else. 00:11:58.220 --> 00:11:59.120 There's a bunch of stuff there. 00:11:59.460 --> 00:12:00.860 There's more than that, though. 00:12:01.460 --> 00:12:08.900 Having these in place, basically it says uv could not have shipped in 2020 because all these weren't in place. 00:12:09.300 --> 00:12:11.280 So it relied on everything. 00:12:12.980 --> 00:12:16.500 PIP-658 went live in PIP-I in May of 2023. 00:12:17.080 --> 00:12:19.960 And uv launched February 2024. 00:12:20.360 --> 00:12:22.560 So it's building on top of all those things. 00:12:23.640 --> 00:12:25.160 Also, uv drops a bunch of stuff. 00:12:25.900 --> 00:12:31.460 PIP still supports a lot of old stuff, which uv just doesn't, like egg support. 00:12:31.800 --> 00:12:34.100 PIPConf doesn't support that. 00:12:34.460 --> 00:12:36.300 There's no bytecode compilation by default. 00:12:36.460 --> 00:12:37.400 I didn't realize this. 00:12:37.560 --> 00:12:42.320 So yeah, when you pip install something, it by default goes through and does PYC, 00:12:42.800 --> 00:12:48.120 It converts code to PYC bytecode objects. 00:12:48.820 --> 00:12:49.800 UV doesn't do this. 00:12:50.220 --> 00:12:52.960 Apparently, you can turn it on, but it doesn't do it by default. 00:12:54.220 --> 00:12:55.320 Requires virtual environments. 00:12:55.720 --> 00:12:58.060 It seems like it could be cached, you know? 00:12:58.740 --> 00:12:58.920 Yeah. 00:12:59.140 --> 00:13:03.440 But I guess I don't know how uv pairs that to Python versions, 00:13:03.600 --> 00:13:05.140 and maybe it gets cached differently. 00:13:05.480 --> 00:13:08.920 You run like Python 3.13 to compile them. 00:13:09.080 --> 00:13:10.560 Maybe it's just like not worth it, you know? 00:13:10.780 --> 00:13:11.000 Yeah. 00:13:12.160 --> 00:13:14.980 I'll just run through these quick stricter spec enforcement, 00:13:17.020 --> 00:13:20.400 ignoring the upper bound on Python because projects say, you know, 00:13:20.410 --> 00:13:22.960 I haven't tested on 4.0, so don't do that. 00:13:23.400 --> 00:13:24.560 But it usually works anyway. 00:13:25.240 --> 00:13:27.400 So I didn't realize they ignored that. 00:13:27.660 --> 00:13:29.000 Also, first index wins. 00:13:29.090 --> 00:13:33.160 I don't know why pip checks all of the indices before it grabs one, 00:13:33.360 --> 00:13:35.220 but uv stops. 00:13:36.980 --> 00:13:41.360 So at the top, it talks about speed comes from elimination. 00:13:41.580 --> 00:13:46.920 basically every code path you don't have to go down is code you don't have to wait for and also 00:13:47.100 --> 00:13:52.920 it's a new project so you're not breaking backwards compatibility if uv doesn't work use pip um like 00:13:53.200 --> 00:13:58.480 sempos mentioned uh there are many things a new project can afford to drop or work around when 00:13:58.660 --> 00:14:04.360 building from ground up which is good it shows as it shows that we can uh what we can have yeah 00:14:04.520 --> 00:14:10.779 basically um you you you could start from scratch which is great uh the other interesting things 00:14:10.800 --> 00:14:16.480 is there's a bunch of optimizations that don't rely on Rust, 00:14:17.780 --> 00:14:18.440 which are interesting. 00:14:18.840 --> 00:14:23.740 Like there's using, because it's only looking at wheels, 00:14:24.860 --> 00:14:27.240 it can, wheels or zip archives, 00:14:27.320 --> 00:14:28.860 it can look at the end of the wheel first 00:14:29.020 --> 00:14:31.720 to grab the directory list, things like that. 00:14:31.940 --> 00:14:34.040 Parallel downloads, this is, you know, 00:14:34.120 --> 00:14:34.980 any language can do that. 00:14:35.200 --> 00:14:37.040 Pip could do parallel downloads, but it doesn't. 00:14:37.240 --> 00:14:40.320 The global clash, we've seen this with uv a lot, 00:14:40.420 --> 00:14:44.120 is when you download something or pip install something from uv with uv, 00:14:44.630 --> 00:14:45.560 it caches it somewhere. 00:14:45.670 --> 00:14:47.160 So if you create another virtual environment, 00:14:47.540 --> 00:14:48.860 it's going to be there already. 00:14:49.210 --> 00:14:51.200 It doesn't have to download again, which is great. 00:14:51.780 --> 00:14:53.320 Anyway, a bunch of cool things here. 00:14:54.420 --> 00:14:56.680 So one of the reasons why I like the article, 00:14:57.100 --> 00:15:01.400 but also I like the idea of thinking I can't speed it up 00:15:01.760 --> 00:15:03.420 because I'm going to stick with Python for now. 00:15:04.000 --> 00:15:06.040 There are trade-offs that you can do. 00:15:06.230 --> 00:15:08.580 You can make some changes in projects, 00:15:08.960 --> 00:15:10.580 especially ones you completely control, 00:15:10.920 --> 00:15:12.340 that you can say, well, you know what? 00:15:12.380 --> 00:15:14.220 I'm going to take away some of the assumptions 00:15:14.560 --> 00:15:16.680 and some of the use models that I used to support 00:15:16.980 --> 00:15:18.200 and just to make it faster. 00:15:19.160 --> 00:15:20.360 Henry points out there, 00:15:21.060 --> 00:15:25.780 makes the first import slower for not having PYC, I think. 00:15:26.440 --> 00:15:28.760 It's an actual savings if you don't use the entire package. 00:15:29.320 --> 00:15:32.440 Though, I guess it depends when this is happening, right? 00:15:32.580 --> 00:15:35.440 Like if this is happening on VNV on your machine, 00:15:35.620 --> 00:15:36.700 like, yeah, it doesn't really matter. 00:15:36.800 --> 00:15:37.900 Like after the first run, it's fast. 00:15:38.060 --> 00:15:43.660 If it's happening on startup on a Docker thing where every time it starts, it uses the new Docker container. 00:15:44.030 --> 00:15:46.500 Well, then it happens every time and it's not great. 00:15:46.790 --> 00:15:55.000 Yeah, it's making me wonder if maybe I can get better startup speed by running a script to pre-compile everything installed in the VNV if uv is not doing it. 00:15:55.220 --> 00:15:55.520 I'll let you know. 00:15:56.320 --> 00:15:56.680 Interesting. 00:15:57.060 --> 00:15:59.700 I didn't understand why pip looks at all of the indices. 00:16:00.440 --> 00:16:08.040 Henry Schreiner says, looking at all the indices is important when you have a newer version on a different index, which sometimes you need to do. 00:16:09.600 --> 00:16:12.760 Sometimes you do need to, but it is slower to look at all of them. 00:16:13.280 --> 00:16:15.020 So, yeah, you can when you have multiple. 00:16:15.720 --> 00:16:20.420 Yeah, this is sort of like inside baseball, but you can set up your your pip configuration. 00:16:20.720 --> 00:16:24.720 So you have multiple places where you're grabbing things like an internal and stuff. 00:16:25.060 --> 00:16:32.440 Yeah. There are opportunities though for pip to speed up without breaking any of its historical compatibility. 00:16:33.080 --> 00:16:40.320 Like it theoretically could say, you asked me to install 10 things when you said -r requirements or high project.tomlin or whatever. 00:16:40.900 --> 00:16:44.880 Let's go get those in parallel. Let's run the install in parallel, right? 00:16:44.910 --> 00:16:49.200 Like we all have 10, 14, 20, whatever cores, right? 00:16:49.360 --> 00:16:52.060 Like don't run it on one and we probably have faster than that. 00:16:52.200 --> 00:16:54.260 Don't just run it on one connection. So on. 00:16:54.600 --> 00:17:00.420 Yeah, also the caching, I love that part of just caching it, 00:17:00.910 --> 00:17:03.040 having a machine local cache. 00:17:04.220 --> 00:17:07.780 For people like me, I create virtual environments all the time. 00:17:07.920 --> 00:17:08.280 So do I. 00:17:08.680 --> 00:17:09.400 They just seem free. 00:17:10.000 --> 00:17:11.100 Yeah, especially with uv. 00:17:11.819 --> 00:17:13.480 And finally, Henry out in the audience throws out 00:17:13.569 --> 00:17:15.100 --compile dash bytecode. 00:17:15.240 --> 00:17:17.040 I'm assuming that's a command to uv. 00:17:17.600 --> 00:17:20.400 And if it is, it's getting set on some of my installs here soon. 00:17:21.420 --> 00:17:21.640 Awesome. 00:17:21.860 --> 00:17:24.560 All right, let's switch gears here 00:17:24.579 --> 00:17:27.660 to the PyView web framework. 00:17:27.990 --> 00:17:31.380 Okay, so PyView, now there have been other things 00:17:31.800 --> 00:17:33.740 like this recently out there. 00:17:34.820 --> 00:17:37.280 Something called PyWebView, not that. 00:17:37.400 --> 00:17:39.660 PyWebView was a thing where you could embed 00:17:40.240 --> 00:17:44.520 basically a browser, kind of like Electron would. 00:17:44.780 --> 00:17:46.140 So this is not that. 00:17:46.180 --> 00:17:48.160 This is a totally separate web framework 00:17:48.380 --> 00:17:49.860 that's super interesting. 00:17:50.020 --> 00:17:52.240 I just interviewed Larry, 00:17:53.680 --> 00:17:55.800 who's behind the project over on Talk Python. 00:17:55.930 --> 00:17:57.260 It's on YouTube as a live stream. 00:17:57.310 --> 00:18:00.460 It is not yet out as a podcast episode. 00:18:01.420 --> 00:18:04.300 So coming soon to a podcast player near you. 00:18:05.040 --> 00:18:07.280 It's based on this thing called Phoenix Live View, 00:18:07.750 --> 00:18:09.480 which is written in Elixir of all things. 00:18:09.980 --> 00:18:11.860 But the idea is we wanna have 00:18:11.990 --> 00:18:14.100 very reactive front-end code, right? 00:18:14.540 --> 00:18:16.340 Kind of like React, the name, 00:18:16.900 --> 00:18:18.060 or Vue or something like that. 00:18:18.680 --> 00:18:22.840 But we would rather have Python code controlling it 00:18:22.860 --> 00:18:24.300 and have it more done on the server, 00:18:24.460 --> 00:18:26.280 a little bit like an HTMX vibe, okay? 00:18:26.620 --> 00:18:29.400 So what this does is it's based on this more popular, 00:18:29.550 --> 00:18:30.540 'cause it's been around a lot longer, 00:18:30.920 --> 00:18:32.900 Phoenix Live View, which people may have heard. 00:18:33.180 --> 00:18:37.620 So what you do is you write code on the server 00:18:38.140 --> 00:18:42.100 as if it could update the front end directly. 00:18:42.480 --> 00:18:45.580 When you launch your app, what happens is it actually, 00:18:46.020 --> 00:18:49.380 this whole framework sets up a WebSocket connection 00:18:49.490 --> 00:18:52.820 from the backend and sends events to you that happen 00:18:52.840 --> 00:18:54.760 they were JavaScript events, but they happen on the server. 00:18:55.230 --> 00:18:57.820 When those events happen, you just re-render the page, 00:18:58.220 --> 00:19:00.400 and there's a diffing engine that figures out, 00:19:00.810 --> 00:19:02.060 okay, you sent the whole page, 00:19:02.300 --> 00:19:05.460 but actually it's just this div and that div 00:19:05.620 --> 00:19:07.760 that need to go down, and it sends them over. 00:19:08.230 --> 00:19:10.020 So let me show you an example, Brian, 00:19:10.080 --> 00:19:12.460 before you're like, I don't know, Michael, this sounds weird. 00:19:12.860 --> 00:19:13.700 It does sound weird. It's cool. 00:19:14.240 --> 00:19:14.840 So check this out. 00:19:16.250 --> 00:19:19.240 I linked to these examples, examples.pyview.rocks. 00:19:19.520 --> 00:19:21.260 There's a bunch of them, but there's a maps example. 00:19:22.040 --> 00:19:25.480 I'm going to show you the example, then I'm going to show you the code, and I want your reaction, okay? 00:19:25.980 --> 00:19:27.040 And I'll narrate for listeners. 00:19:27.900 --> 00:19:32.400 So if I zoom out, what we have is on the left, we've got a list of U.S. national parks. 00:19:32.820 --> 00:19:36.700 And on the right, we have a OpenStreetMaps map, right? 00:19:36.920 --> 00:19:40.540 So you see on the left, it has the selected park selected. 00:19:40.900 --> 00:19:46.060 And if I click on one over here in the map, it then updates the thing on the left, right? 00:19:46.220 --> 00:19:51.160 To say like, okay, this is a JavaScript map component that's loaded here. 00:19:51.480 --> 00:19:56.400 And as you interact with it, you can see the details view, or the list view rather. 00:19:57.360 --> 00:19:58.200 Oh no, I lost my window. 00:19:59.080 --> 00:19:59.940 Is updating, right? 00:20:00.000 --> 00:20:00.480 They're staying in sync. 00:20:00.880 --> 00:20:01.360 That's cool. 00:20:01.660 --> 00:20:01.960 It's cool. 00:20:02.060 --> 00:20:04.120 Now, also, when I click on the left, it moves the map. 00:20:04.200 --> 00:20:06.120 So if I click on Yellowstone, the map jumps over there. 00:20:06.520 --> 00:20:08.540 If I click the Joshua tree, it's over there. 00:20:08.640 --> 00:20:15.640 If I bring something closer to home, like Olympic National Park, National Forest, that pops up. 00:20:15.880 --> 00:20:19.080 This is a pretty neat app, and it's clearly interactive, right? 00:20:19.120 --> 00:20:20.180 It's not refreshing the page. 00:20:20.520 --> 00:20:21.120 Feels complicated. 00:20:21.300 --> 00:20:23.340 But if I go back here and I go to the maps 00:20:23.400 --> 00:20:25.380 and I show you the code, I'll describe this for everyone. 00:20:25.980 --> 00:20:27.980 There's a CSS file, that's fine. 00:20:28.380 --> 00:20:30.060 There's a, hold on, let's do it this way. 00:20:30.320 --> 00:20:32.020 There's a parks py file, 00:20:32.140 --> 00:20:34.240 and it literally is just a list of dictionaries 00:20:34.320 --> 00:20:35.520 of like, here's the name of the parks, 00:20:35.640 --> 00:20:36.860 here's the latitude, here's the description, 00:20:37.060 --> 00:20:39.420 here's the icon, it's just like a flat data file, yeah? 00:20:39.760 --> 00:20:43.780 Go back, the map.py, it says, 00:20:44.340 --> 00:20:45.440 we're gonna create a data class. 00:20:45.800 --> 00:20:47.520 And the data class has a list of parks 00:20:47.720 --> 00:20:48.960 and a selected park, okay? 00:20:49.300 --> 00:20:56.760 When somebody clicks on it, it just does a dictionary comprehension or sort of thing to search for the park that matches. 00:20:57.170 --> 00:21:02.340 It sets the selected park name on its data class, and it says there's an event to highlight the park. 00:21:02.720 --> 00:21:03.720 That's the entire application. 00:21:04.040 --> 00:21:04.980 That's pretty amazing. 00:21:05.300 --> 00:21:05.840 Is that nuts? 00:21:06.030 --> 00:21:06.180 Right? 00:21:06.270 --> 00:21:15.480 And if you look at the HTML, it's just a bunch of Jinja with Tailwind that just says, you know, bind to this value. 00:21:16.220 --> 00:21:22.160 Instead of just saying the value equals, you say Phoenix value equals, and you just pull out the context from the name. 00:21:22.520 --> 00:21:27.240 So this is a super interesting web framework that has not got a lot of light shined on it yet. 00:21:27.400 --> 00:21:28.480 It's got 63 stars. 00:21:29.159 --> 00:21:32.320 So it's basically a reimplementation of the Phoenix live view. 00:21:32.760 --> 00:21:34.400 And I think it's pretty neat. 00:21:34.780 --> 00:21:39.500 Larry says he's been using it for some projects at work and just starting to get it out in the public. 00:21:39.640 --> 00:21:41.580 So it's probably more tested than it feels. 00:21:42.040 --> 00:21:44.440 And finally, if you want to understand, I'll link to this in the show notes. 00:21:44.720 --> 00:21:51.660 If you want to understand, like, how is this even possible, you need to read the LiveView life cycle section of the docs. 00:21:52.300 --> 00:21:55.040