WEBVTT 00:00:00.001 --> 00:00:06.380 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:06.380 --> 00:00:11.500 This is episode 404, recorded October 7th, 2024. 00:00:11.500 --> 00:00:12.860 I'm Michael Kennedy. 00:00:12.860 --> 00:00:14.040 And I'm Brian Okken. 00:00:14.040 --> 00:00:17.220 And this episode is brought to you by Scout APM. 00:00:17.220 --> 00:00:18.980 We will, of course, tell you more about them later. 00:00:18.980 --> 00:00:21.960 The link is in the podcast player show notes right at the top. 00:00:21.960 --> 00:00:27.720 Brian, I've been looking for this episode for a while and I just can't find it, man. 00:00:27.720 --> 00:00:29.500 I was like, where do I go? 00:00:30.220 --> 00:00:30.620 404. 00:00:30.620 --> 00:00:31.820 Everywhere I go, it's 404. 00:00:31.820 --> 00:00:32.780 What's going on? 00:00:32.780 --> 00:00:35.600 Yeah, I don't know. 00:00:35.600 --> 00:00:37.280 We'll find it. 00:00:37.280 --> 00:00:38.440 Let's make it. 00:00:38.440 --> 00:00:39.920 It's going to be a journey. 00:00:39.920 --> 00:00:41.940 We're going to try to get our way through it. 00:00:41.940 --> 00:00:46.240 Speaking of a journey, you want to stay in touch with us, hit by them by Scout APM. 00:00:46.240 --> 00:00:47.000 Click on newsletter. 00:00:47.000 --> 00:00:48.140 Enter your info there. 00:00:48.140 --> 00:00:49.360 We'll keep you up to date. 00:00:49.360 --> 00:00:50.720 Brian will send you some show notes. 00:00:50.720 --> 00:00:54.560 Find us on Mastodon and other social places. 00:00:54.560 --> 00:00:58.080 And even subscribe right here on YouTube if you're watching the video. 00:00:58.500 --> 00:01:03.040 And with that, were you able to find an item for your first item today? 00:01:03.040 --> 00:01:04.780 Well, I almost did it. 00:01:04.780 --> 00:01:05.900 So, okay. 00:01:05.900 --> 00:01:09.220 The first item, I really wanted to cover Python 3.13. 00:01:09.620 --> 00:01:13.580 Because Python 3.13 is released on October 7th. 00:01:13.580 --> 00:01:16.480 And today is, text notes, October 7th. 00:01:16.480 --> 00:01:18.100 So, it's today, right? 00:01:18.100 --> 00:01:24.840 So, I went to python.org and go, well, the latest, it still says latest is 3.12. 00:01:24.840 --> 00:01:26.180 So, they're working on it. 00:01:26.180 --> 00:01:27.240 Thomas Wooters and others. 00:01:27.620 --> 00:01:29.200 It's a process to get it out. 00:01:29.200 --> 00:01:30.480 But it's really out already. 00:01:30.480 --> 00:01:31.640 So, you just can't see it. 00:01:31.640 --> 00:01:33.920 If you go to downloads, it still says 3.12. 00:01:33.920 --> 00:01:35.480 I'm sure it'll be updated soon. 00:01:35.480 --> 00:01:39.120 However, we've got the link for 3.13.0. 00:01:39.120 --> 00:01:40.700 It's really not that complicated. 00:01:40.700 --> 00:01:44.560 Downloads, release, python-3.1.3.0. 00:01:45.320 --> 00:01:46.820 And then you got it. 00:01:46.820 --> 00:01:49.820 So, it's got a stable release of 3.13.0. 00:01:49.820 --> 00:01:53.140 I've been playing with it this morning and having a lot of fun. 00:01:53.140 --> 00:01:57.640 If you go down to the bottom of this link, there's where the installers are. 00:01:57.640 --> 00:02:00.480 The pre-built installers, if you want to grab those. 00:02:00.480 --> 00:02:02.480 Or you can install it yourself if you want. 00:02:02.480 --> 00:02:03.620 Or build it yourself. 00:02:03.620 --> 00:02:05.580 But I'm on a Mac right now. 00:02:05.580 --> 00:02:07.060 So, I grabbed the Mac installer. 00:02:07.060 --> 00:02:11.160 And I want to cover some of the new stuff in 3.13. 00:02:11.160 --> 00:02:14.840 But I really wanted to play with the multi-threading stuff, which I haven't yet. 00:02:14.960 --> 00:02:16.940 I just wanted to make sure it was enabled. 00:02:16.940 --> 00:02:18.220 And there's a bit of a trick. 00:02:18.220 --> 00:02:20.320 So, if you look in... 00:02:20.320 --> 00:02:22.300 Maybe it's not in here. 00:02:22.300 --> 00:02:23.540 It's in the What's New page. 00:02:23.540 --> 00:02:30.640 So, if you go to the What's New, it does talk about the Python, the macOS, and Windows installers. 00:02:30.640 --> 00:02:32.700 And there's a link somewhere in here. 00:02:32.700 --> 00:02:35.160 But we'll include the link also. 00:02:35.160 --> 00:02:41.820 That when you're installing it, and on the Mac, when I did it, you go through this install process. 00:02:41.820 --> 00:02:46.180 You have to hit customize and click the free-threaded. 00:02:46.180 --> 00:02:49.340 Because free-threading is off by default. 00:02:49.340 --> 00:02:51.740 So, when you install it, you have to turn it on. 00:02:51.740 --> 00:02:53.720 So, just remember that. 00:02:53.720 --> 00:02:56.740 And I will play with the free-threading later. 00:02:56.740 --> 00:02:59.040 So, what do we have new in 3.13? 00:02:59.040 --> 00:03:00.620 I think we've covered this before. 00:03:00.620 --> 00:03:02.600 But now we actually have 3.13. 00:03:02.600 --> 00:03:03.480 So, we can play with it. 00:03:04.160 --> 00:03:08.100 We have the interactive interpreter changes, which... 00:03:08.100 --> 00:03:09.680 Actually, let's back up a little bit. 00:03:09.680 --> 00:03:11.280 This isn't even called out here. 00:03:11.280 --> 00:03:14.600 My favorite feature of 3.13 is exit. 00:03:14.600 --> 00:03:17.340 Oh, I guess it is here. 00:03:17.340 --> 00:03:21.860 Direct REPL support for help, exit, and quit. 00:03:21.860 --> 00:03:23.580 And thank you. 00:03:23.580 --> 00:03:23.980 Thank you. 00:03:23.980 --> 00:03:24.360 Thank you. 00:03:24.360 --> 00:03:24.360 Thank you. 00:03:24.360 --> 00:03:25.240 Whoever worked on this. 00:03:25.240 --> 00:03:28.180 Because it used to say, oh, did you mean to exit out? 00:03:28.180 --> 00:03:31.860 You have to type exit with the parans or quit with the parans. 00:03:31.860 --> 00:03:35.120 And if you know what I want to do, just do it. 00:03:35.120 --> 00:03:36.500 So, you know what it does. 00:03:36.500 --> 00:03:36.960 It's really cool. 00:03:36.960 --> 00:03:41.920 I can imagine somebody was like, well, what if we have a variable named exit and you want 00:03:41.920 --> 00:03:42.720 to see its value? 00:03:42.720 --> 00:03:45.180 But if there's no variable defined, just quit. 00:03:45.180 --> 00:03:45.880 I love it. 00:03:46.840 --> 00:03:49.380 So, exit and quit now work. 00:03:49.380 --> 00:03:50.700 I didn't know quit worked. 00:03:50.700 --> 00:03:51.280 That's neat. 00:03:51.280 --> 00:03:57.160 So, the other couple things in interactive, there's a lot of stuff in the interactive REPL 00:03:57.160 --> 00:03:58.300 that's changed. 00:03:58.300 --> 00:04:00.240 Multi-line editing, amazing. 00:04:00.240 --> 00:04:05.520 So, you type in a multi-line thing and you get the three dots, which is great. 00:04:05.520 --> 00:04:10.880 But when you want to go and type something wrong and you want to go change that, now you 00:04:10.880 --> 00:04:14.900 can go back and you can just hit the up arrow and it works. 00:04:14.900 --> 00:04:16.160 It's great. 00:04:16.520 --> 00:04:17.640 So, I'm really enjoying that. 00:04:17.640 --> 00:04:20.580 There's F1 for command history. 00:04:20.580 --> 00:04:26.440 History is, we've had history before, but now if you exit and then start the REPL again, 00:04:26.440 --> 00:04:27.740 history is still there. 00:04:27.740 --> 00:04:29.920 It preserves all of your history from the session. 00:04:29.920 --> 00:04:30.320 Oh, that's nice. 00:04:30.320 --> 00:04:33.660 Super nice for when you're debugging something, trying it out. 00:04:33.660 --> 00:04:35.620 And the colors are amazing. 00:04:35.620 --> 00:04:37.820 So, they've turned on colors by default. 00:04:37.820 --> 00:04:40.160 So, in prompts and tracebacks. 00:04:40.160 --> 00:04:43.680 So, even just having the little prompt be in a color, it's nice. 00:04:43.680 --> 00:04:44.480 It helps. 00:04:44.480 --> 00:04:45.600 So, love that. 00:04:45.840 --> 00:04:47.460 We've also got improved error messages. 00:04:47.460 --> 00:04:53.480 Like, before, if you named, like, random is built in, one of the built-in things. 00:04:53.480 --> 00:04:56.540 But if you named a module named random, it would just mess up. 00:04:56.540 --> 00:05:01.100 And it would, like, it would say stuff like, you know, module random has no attribute. 00:05:01.620 --> 00:05:08.200 But now it says, hey, you've already considered renaming your own module something else because you're messing up. 00:05:08.200 --> 00:05:09.900 So, if you rename them. 00:05:09.900 --> 00:05:20.300 If you name a module or a script, the same thing as a built-in or now as one of your installed third-party libraries, Python will tell you that you've messed up, which is nice. 00:05:20.720 --> 00:05:22.440 There's other improvements, too. 00:05:22.440 --> 00:05:30.800 Like, if you misspell a keyword argument, it will suggest that maybe you had the correct spelling. 00:05:30.800 --> 00:05:31.800 It will suggest that. 00:05:31.800 --> 00:05:32.820 That's pretty cool. 00:05:32.820 --> 00:05:35.460 And then, of course, there's lots of changes in 3.13. 00:05:35.460 --> 00:05:37.920 But the REPL, I'm excited about. 00:05:37.920 --> 00:05:41.800 I'm also excited about the error messages and free-threaded Python. 00:05:42.080 --> 00:05:43.480 So, there's a lot of other stuff, too. 00:05:43.480 --> 00:05:45.240 So, check out the What's New. 00:05:45.240 --> 00:05:48.840 But I really am looking forward to playing with the free-threading. 00:05:48.840 --> 00:05:50.120 Yeah, this is amazing. 00:05:50.120 --> 00:05:51.600 I'm really excited to see this here. 00:05:51.600 --> 00:05:54.520 I actually can't believe that free-threaded Python is a thing. 00:05:54.520 --> 00:05:55.220 I know. 00:05:55.220 --> 00:05:56.040 It's like. 00:05:56.040 --> 00:05:56.920 It's been so long. 00:05:56.920 --> 00:05:58.060 The journey's been so long. 00:05:58.060 --> 00:06:01.300 I mean, I was walking this morning and I saw a pig flying through the sky. 00:06:01.300 --> 00:06:07.800 It had imported free-threaded Python, rather than imported anti-gravity. 00:06:07.800 --> 00:06:08.380 That's what it did. 00:06:08.380 --> 00:06:09.020 Yes. 00:06:09.020 --> 00:06:09.420 Yeah. 00:06:09.700 --> 00:06:14.740 So, I know that a lot of people have worked a lot, long and hard, to make that work. 00:06:14.740 --> 00:06:16.000 So, I'm looking forward to it. 00:06:16.000 --> 00:06:19.480 Now, free-threading is still officially experimental. 00:06:19.480 --> 00:06:22.660 So, I don't know what that means. 00:06:22.660 --> 00:06:23.860 If they're going to take it back from. 00:06:23.860 --> 00:06:25.780 Please don't take it away from us. 00:06:25.780 --> 00:06:26.820 But we'll see. 00:06:26.820 --> 00:06:27.260 Yeah. 00:06:27.260 --> 00:06:31.180 The original PEP that got it accepted, 703 or something like that, said, 00:06:31.180 --> 00:06:32.800 we are allowing this in. 00:06:32.800 --> 00:06:35.140 But if it causes too much trouble, we're taking it back out. 00:06:35.140 --> 00:06:36.700 Which I've never seen in a PEP. 00:06:36.700 --> 00:06:37.580 Maybe it's happened before. 00:06:37.580 --> 00:06:38.420 But we approve it. 00:06:38.600 --> 00:06:40.020 But we might unapprove it. 00:06:40.020 --> 00:06:41.500 Might unapprove it. 00:06:41.500 --> 00:06:43.180 We'll turn this car around. 00:06:43.180 --> 00:06:45.380 If you keep fighting, we'll turn it around. 00:06:45.380 --> 00:06:45.900 Yeah. 00:06:45.900 --> 00:06:46.740 That's right. 00:06:46.740 --> 00:06:47.840 Don't make me turn this car. 00:06:47.840 --> 00:06:48.740 Don't make me pull over. 00:06:48.740 --> 00:06:54.300 Speaking of pulling over, this was supposed to be out a week ago, Brian. 00:06:54.300 --> 00:06:54.800 Yeah. 00:06:54.800 --> 00:06:55.240 Yeah. 00:06:55.420 --> 00:07:03.300 And apparently, it was pushed back due to performance issues or some kind of issue with the incremental GC. 00:07:03.300 --> 00:07:03.980 Right. 00:07:04.160 --> 00:07:08.580 And I'll put the Python, the discuss.python.org link in my extras. 00:07:08.580 --> 00:07:10.760 But I'm not going to cover it separately. 00:07:10.760 --> 00:07:12.380 But it's kind of irrelevant. 00:07:12.380 --> 00:07:13.160 Things are out now. 00:07:13.160 --> 00:07:14.160 But yeah. 00:07:14.160 --> 00:07:14.700 Very cool. 00:07:14.700 --> 00:07:15.180 All right. 00:07:15.180 --> 00:07:16.800 Well, what do you got for us next, Python? 00:07:16.800 --> 00:07:17.240 Michael? 00:07:17.360 --> 00:07:22.860 Well, let's stick with some core language topics and stuff like that. 00:07:22.860 --> 00:07:30.400 I want to talk about PyPI bandwidth and PEP 759 external wheel hosting. 00:07:30.660 --> 00:07:37.700 So right now, when you pip install, uv pip install, whatever, you go to get something and you, 00:07:37.700 --> 00:07:44.020 as a publisher of a package like Flask, you send that to pypi.org, maybe as a wheel, maybe 00:07:44.020 --> 00:07:48.080 as a multi-platform variations of different wheels, if it's got to compile differently or 00:07:48.080 --> 00:07:48.320 something. 00:07:48.320 --> 00:07:51.620 And then people pip install that and it downloads from there. 00:07:51.620 --> 00:07:54.520 That results in an insane amount of traffic. 00:07:54.520 --> 00:07:57.840 I actually want to talk a little bit more about that later in general. 00:07:57.840 --> 00:07:59.660 But it's an insane amount of traffic. 00:07:59.880 --> 00:08:05.920 Over 66 petabytes a month, which is six-figure level of bandwidth costs. 00:08:05.920 --> 00:08:13.880 And luckily, Fastly, the CDN company, is basically donating all that bandwidth at reasonably high 00:08:13.880 --> 00:08:14.640 expense for them. 00:08:14.640 --> 00:08:18.900 Probably not the same as retail, but whatever wholesale that is in bandwidth, I'm sure it's 00:08:18.900 --> 00:08:19.320 still a lot. 00:08:19.320 --> 00:08:25.560 And they announced a five-year plan or five-year commitment to the PSF to do that at PyCon. 00:08:25.560 --> 00:08:26.100 All right. 00:08:26.100 --> 00:08:27.380 They're on the big keynote stage. 00:08:27.380 --> 00:08:28.040 That was nice. 00:08:28.040 --> 00:08:29.040 Thank you, Fastly. 00:08:29.100 --> 00:08:33.560 However, there's still problems with this, even though they're covering it. 00:08:33.560 --> 00:08:35.640 What if Fastly changes their mind eventually? 00:08:35.640 --> 00:08:43.840 But more importantly, there's hard limits on how large your project's overall usage on 00:08:43.840 --> 00:08:45.100 pypi.org can be. 00:08:45.100 --> 00:08:49.340 It gets no individual package can be bigger than 100 megs. 00:08:49.400 --> 00:08:55.220 You can't have 10 gigs total across all platforms, across all releases, things like that, right? 00:08:55.220 --> 00:08:57.760 You can't apply for exceptions, but it's not great. 00:08:57.760 --> 00:09:02.120 What if you could just say, here's my wheel, but when you actually download it, download it 00:09:02.120 --> 00:09:02.680 from over there. 00:09:02.680 --> 00:09:06.620 Download it off of Microsoft.com if it's a Microsoft package. 00:09:06.860 --> 00:09:13.080 Or download it off of explosion.ai if it's some sort of spaCy package. 00:09:13.080 --> 00:09:14.580 Those things have to be quite large. 00:09:14.580 --> 00:09:16.280 A lot of their models that you pip install. 00:09:16.280 --> 00:09:19.720 I think they have to get an exception to be even uploaded in the first place. 00:09:19.720 --> 00:09:20.080 All right. 00:09:20.080 --> 00:09:20.960 So that's what this is about. 00:09:20.960 --> 00:09:21.740 It's by Barry Warsaw. 00:09:21.740 --> 00:09:22.680 Ethan Smith. 00:09:22.680 --> 00:09:29.500 Delegate is Donald Stufft, who is a person largely behind the rewrite of pypi.org. 00:09:29.500 --> 00:09:31.140 So here's the deal. 00:09:31.140 --> 00:09:38.620 Brian, if you had a wheel, which is the metadata about a package and all of its binary details 00:09:38.620 --> 00:09:44.500 and its source code and all that, and you ripped off a lot of the wheel stuff, like the binary 00:09:44.500 --> 00:09:46.220 bits, what would you be left with? 00:09:46.220 --> 00:09:46.720 I don't know. 00:09:46.720 --> 00:09:47.320 A rim. 00:09:47.320 --> 00:09:49.620 A wheel rim, you know? 00:09:49.620 --> 00:09:50.200 Oh. 00:09:50.200 --> 00:09:57.640 So this PEP literally defines a new package format called a rim file. 00:09:58.020 --> 00:10:00.280 And the rim file is basically the wheel without the binary. 00:10:00.280 --> 00:10:02.800 Instead, it's got a URL to where the binary lives. 00:10:02.800 --> 00:10:07.960 And a checksum of the binary, which may already be there anyway, but it's left in the rim file, 00:10:07.960 --> 00:10:09.280 if not added to it. 00:10:09.280 --> 00:10:09.560 Okay. 00:10:09.560 --> 00:10:10.160 Interesting. 00:10:10.160 --> 00:10:15.200 So basically, you upload just the metadata if you want to participate in this. 00:10:15.200 --> 00:10:17.100 And then you point somewhere else. 00:10:17.100 --> 00:10:22.620 And there's a lot of motivation, things they tried, things that didn't work, limits. 00:10:22.620 --> 00:10:25.740 You can see in this pep, they list the limits that I talked about. 00:10:26.060 --> 00:10:35.040 And all the details, but most importantly, there's a security section way down here. 00:10:35.040 --> 00:10:36.120 I was worried about that. 00:10:36.120 --> 00:10:36.840 So, yeah. 00:10:36.840 --> 00:10:37.320 Yes. 00:10:37.320 --> 00:10:37.780 Here. 00:10:37.780 --> 00:10:39.560 As well as stability constraints. 00:10:39.560 --> 00:10:44.020 But so for example, the wheel checksum must be included in rim files. 00:10:44.020 --> 00:10:47.420 Once it's uploaded, just like current wheels, it can't change. 00:10:47.420 --> 00:10:47.860 Yeah. 00:10:48.120 --> 00:10:49.480 So here's my metadata. 00:10:49.480 --> 00:10:50.700 That's where the thing lives. 00:10:50.700 --> 00:10:52.300 Here's the hash of it. 00:10:52.300 --> 00:10:53.860 So it can't be tampered with. 00:10:53.860 --> 00:10:56.340 It must be served over HTTPS. 00:10:56.340 --> 00:10:57.000 Okay. 00:10:57.260 --> 00:11:01.840 And you're only allowed to do this if you get pre-approved by the PyPI admin. 00:11:01.840 --> 00:11:04.540 So it's not just all sorts of chaos. 00:11:04.540 --> 00:11:06.480 People just upload and stuff everywhere. 00:11:06.480 --> 00:11:07.040 All right. 00:11:07.040 --> 00:11:10.100 But a lot of consideration about like malware and things like that. 00:11:10.100 --> 00:11:12.360 And also a bunch of stuff about stability. 00:11:12.360 --> 00:11:16.020 Like external wheels must be hosted over SSL. 00:11:16.020 --> 00:11:21.240 And not just that, but it actually must be signed by Mozilla's root certificate store, which 00:11:21.240 --> 00:11:22.180 is interesting. 00:11:22.180 --> 00:11:24.240 And they should use CDNs. 00:11:24.240 --> 00:11:26.100 They must be stable, et cetera. 00:11:26.100 --> 00:11:26.720 What do you think? 00:11:26.720 --> 00:11:28.840 So who would do this? 00:11:28.840 --> 00:11:34.180 I mean, because then the person with the package doing this, they'd have to pay the hosting fees. 00:11:34.180 --> 00:11:37.780 Or get it on GitHub and point to it as a raw or something, you know? 00:11:37.780 --> 00:11:38.900 Oh, yeah. 00:11:38.900 --> 00:11:39.920 Something like that. 00:11:39.920 --> 00:11:40.340 I don't know. 00:11:40.340 --> 00:11:41.120 Oh, okay. 00:11:41.120 --> 00:11:47.820 Or if you're a big company and you want to ship ML models and you'd rather have more control 00:11:47.820 --> 00:11:50.040 over that and how it works, right? 00:11:50.040 --> 00:11:51.040 You could set that. 00:11:51.120 --> 00:11:55.940 I don't think small individuals are going to be like, hey, I'd love to cover a terabyte 00:11:55.940 --> 00:11:57.800 of bandwidth charges a month for no reason. 00:11:57.800 --> 00:11:58.360 Yeah. 00:11:58.360 --> 00:11:58.800 Right. 00:11:58.800 --> 00:12:04.760 Well, also, like you said, it's somebody that's got a, it's going to be a large one. 00:12:04.760 --> 00:12:08.040 You're not going to do this for just a little Python only package. 00:12:08.040 --> 00:12:09.000 Yeah, exactly. 00:12:09.000 --> 00:12:09.860 It's very unlikely. 00:12:09.860 --> 00:12:10.440 Right. 00:12:10.440 --> 00:12:11.420 It's very unlikely. 00:12:11.420 --> 00:12:18.160 So they said this is only about hosting the binary downloads, not packages, package indexes, 00:12:18.160 --> 00:12:20.660 package listings, none of that stuff. 00:12:20.880 --> 00:12:25.100 Because that's already supported through having just a separate private white label. 00:12:25.100 --> 00:12:26.300 You run it yourself. 00:12:26.300 --> 00:12:29.020 PyPI, mirror, something like that, right? 00:12:29.020 --> 00:12:29.520 Yeah. 00:12:29.520 --> 00:12:30.820 I think this is cool, actually. 00:12:30.820 --> 00:12:31.600 Neat. 00:12:31.600 --> 00:12:31.960 Yeah. 00:12:31.960 --> 00:12:32.780 Flexibility is good. 00:12:32.780 --> 00:12:34.500 I think it's, I think it's all right. 00:12:34.500 --> 00:12:35.780 So anyway, very cool. 00:12:35.780 --> 00:12:37.180 That's, that's what I wanted to cover. 00:12:37.180 --> 00:12:41.300 And before we move on, how about a quick shout out to our sponsor? 00:12:41.300 --> 00:12:41.820 Yep. 00:12:41.820 --> 00:12:45.440 Let me tell you real quick about Scout APM. 00:12:45.440 --> 00:12:47.580 They're big supporters of Python Bytes. 00:12:47.580 --> 00:12:49.080 So we appreciate that very much. 00:12:49.080 --> 00:12:55.260 So if you are tired of spending hours trying to find the root cause of issues impacting your performance, 00:12:55.260 --> 00:12:58.020 then you owe it to yourself to check out Scout APM. 00:12:58.020 --> 00:13:07.140 They're a leading Python application performance monitoring tool, APM, that helps you identify and solve performance abnormalities faster and easier. 00:13:07.360 --> 00:13:17.920 Scout APM ties bottlenecks such as memory leaks, slow database queries, background jobs, and the dreaded N plus one queries that you can end up if you do lazy loading in your ORM. 00:13:17.920 --> 00:13:20.320 And then you say, oh no, why is it so slow? 00:13:20.320 --> 00:13:22.660 Why are you doing 200 database queries for what should be one? 00:13:22.660 --> 00:13:24.040 So you can find out things like that. 00:13:24.040 --> 00:13:31.260 And it links it back directly to source code so you can spend less time in the debugger and healing logs and just finding the problems and moving on. 00:13:31.260 --> 00:13:34.320 And you'll love it because it's built for developers by developers. 00:13:34.320 --> 00:13:35.980 It makes it easy to get set up. 00:13:36.300 --> 00:13:38.500 Seriously, you can do it in less than four minutes. 00:13:38.500 --> 00:13:39.340 So that's awesome. 00:13:39.340 --> 00:13:42.380 And the best part is the pricing is straightforward. 00:13:42.380 --> 00:13:47.760 You only pay for the data that you use with no hidden overage fees or per seat pricing. 00:13:47.760 --> 00:13:50.120 And I just learned this, Brian. 00:13:50.120 --> 00:13:55.240 They also have, they provide the pro version for free to all open source projects. 00:13:55.240 --> 00:14:02.580 So if you're an open source maintainer and you want to have Scout APM for that project, just shoot them a message or something on their pricing page about that. 00:14:02.780 --> 00:14:06.440 So you can start your free trial and get instant insights today. 00:14:06.440 --> 00:14:09.040 Visit pythonbytes.fm/scout. 00:14:09.040 --> 00:14:11.300 The link is in your podcast player show notes as well. 00:14:11.300 --> 00:14:12.720 And please use that link. 00:14:12.720 --> 00:14:16.580 Don't just search for them because otherwise they don't think you came from us. 00:14:16.580 --> 00:14:18.180 And then they'd stop supporting the show. 00:14:18.240 --> 00:14:20.740 So please use our link pythonbytes.fm/scout. 00:14:20.740 --> 00:14:21.660 Check them out. 00:14:21.660 --> 00:14:23.480 How about the next one for you? 00:14:23.480 --> 00:14:25.560 I feel like, just give people a little heads up. 00:14:25.560 --> 00:14:27.500 We've got some, that's a pie test coming up here. 00:14:27.500 --> 00:14:28.020 What do we got? 00:14:28.500 --> 00:14:30.840 If you didn't know it, Python 3.13 is out. 00:14:30.840 --> 00:14:34.140 So we already covered this. 00:14:34.140 --> 00:14:38.140 However, it's got the free threading thing going on. 00:14:38.140 --> 00:14:40.960 How do you know if your project supports it? 00:14:40.960 --> 00:14:46.140 So that's the question that Anthony Shaw and others at, where was this at again? 00:14:46.140 --> 00:14:48.780 Was that PyCon JP? 00:14:48.780 --> 00:14:50.460 I think that's PyCon Japan. 00:14:50.460 --> 00:14:51.040 Is that right? 00:14:51.040 --> 00:14:52.220 I think so. 00:14:52.500 --> 00:14:52.720 Okay. 00:14:52.720 --> 00:14:54.320 So it was one of their sprints. 00:14:54.320 --> 00:15:00.420 Looks like seven or eight people, Anthony, Mike, a bunch of people worked on this. 00:15:00.420 --> 00:15:08.160 And the idea is that you'd like to be able to test that your project and your code works on the free threaded, right? 00:15:08.160 --> 00:15:13.580 So if you're like, especially support libraries and packages. 00:15:13.580 --> 00:15:15.540 So what, how do you do that? 00:15:15.540 --> 00:15:19.900 And one of the thoughts was to just test, just run it. 00:15:19.900 --> 00:15:23.500 But with the free threading turned on. 00:15:23.500 --> 00:15:26.920 But pytest normally runs in single thread mode by default. 00:15:26.920 --> 00:15:28.420 So how about XTest? 00:15:28.420 --> 00:15:30.000 Because that's multiprocessing, right? 00:15:30.000 --> 00:15:31.640 Yes, it's multiprocessing. 00:15:31.640 --> 00:15:32.920 It's not multithreading. 00:15:32.920 --> 00:15:36.520 So Anthony and others made this plugin. 00:15:36.520 --> 00:15:39.800 It's called pytest-FreeThreaded. 00:15:39.800 --> 00:15:41.880 And link in the show notes, of course. 00:15:41.880 --> 00:15:52.120 But they made the plugin to be able to run the tests in a thread pool with the guild disabled and help identify that your tests are thread safe. 00:15:52.120 --> 00:15:59.180 So, you know, as long as your tests possibly are testing your code in such a way to test that, great. 00:15:59.180 --> 00:16:03.640 But at least your tests will be free thread safe compatible. 00:16:03.640 --> 00:16:10.640 Anyway, they tested the first library that was marked as compatible with free threading. 00:16:10.640 --> 00:16:11.740 I don't know how to do that. 00:16:11.740 --> 00:16:14.320 So I guess that's a to-do item to look that up. 00:16:14.320 --> 00:16:15.140 Okay. 00:16:15.140 --> 00:16:18.560 So you think it might go in the classifiers of the package metadata? 00:16:18.560 --> 00:16:19.100 I'm not sure. 00:16:19.100 --> 00:16:20.380 Oh, okay. 00:16:20.380 --> 00:16:20.940 All right. 00:16:20.940 --> 00:16:22.180 Well, we'll take a look at that. 00:16:22.180 --> 00:16:27.260 Anyway, they tested one and it caused a segfault in CPython. 00:16:27.260 --> 00:16:30.200 So not really compatible if you're segfaulting. 00:16:30.540 --> 00:16:36.540 So I haven't tried this out yet, but I'm going to try it on a few projects anyway. 00:16:36.540 --> 00:16:39.400 So Python free threaded will help test this. 00:16:39.400 --> 00:16:41.020 Well, I think this is awesome. 00:16:41.020 --> 00:16:45.200 And some people might perceive this like, oh, it just means a performance boost. 00:16:45.200 --> 00:16:52.360 But if you're writing code that's meant to be supported for free threaded Python, you should test it in that style as well. 00:16:52.400 --> 00:16:54.920 Not just in some limited way, right? 00:16:54.920 --> 00:16:55.720 Yeah. 00:16:56.420 --> 00:17:13.640 And even if your tests pass, I think it would make sense to look at the different parts of your library to make sure that it really makes sense that your tests are actually testing your library in such a way that you have fairly good confidence. 00:17:13.640 --> 00:17:15.460 But, you know, mileage may vary. 00:17:15.460 --> 00:17:18.340 But this is a good step in the right direction. 00:17:18.340 --> 00:17:21.040 So thanks to Anthony and others for doing this. 00:17:21.040 --> 00:17:21.780 Yeah, that's awesome. 00:17:22.260 --> 00:17:33.040 Speaking of Anthony, I was not going to point this out, but now that you mentioned it, just published an episode on Talk Python with him designing effective load tests for your Python app. 00:17:33.040 --> 00:17:33.620 Nice. 00:17:33.620 --> 00:17:35.060 Yeah, that's a really fun one. 00:17:35.060 --> 00:17:36.140 So people can check that out. 00:17:36.140 --> 00:17:40.680 But what I'm actually talking about for this section is more pytests. 00:17:40.680 --> 00:17:42.480 Ryan, how did you not get this one? 00:17:42.480 --> 00:17:43.560 pytest edit. 00:17:43.560 --> 00:17:44.480 Do you know this? 00:17:44.480 --> 00:17:44.920 No. 00:17:44.920 --> 00:17:46.200 Okay, so here's the deal. 00:17:46.200 --> 00:17:52.240 You're on the terminal CLI sitting there running something and you run into an error and you're like, ah. 00:17:52.240 --> 00:17:54.680 My test crashed or my test failed rather. 00:17:54.680 --> 00:17:56.260 I just want to see what's up. 00:17:56.260 --> 00:17:59.260 So one way to do that is to see what test failed. 00:17:59.260 --> 00:18:03.280 Go open your editor, find your project, load it up, search around until you find it. 00:18:03.280 --> 00:18:08.260 Or the other way is to install pytest-edit from Mr. Minnow. 00:18:08.260 --> 00:18:09.280 That's the GitHub name. 00:18:09.280 --> 00:18:17.400 And then just type pytest-edit and it'll open a pre-configured editor to the line of the failing test. 00:18:17.400 --> 00:18:18.120 Oh, wow. 00:18:18.120 --> 00:18:21.420 Yeah, so you can say pytest-edit and it'll show you the last one. 00:18:21.420 --> 00:18:23.100 Or you can say some index. 00:18:23.100 --> 00:18:25.580 So for example, you can say edit equals two. 00:18:25.580 --> 00:18:28.820 And it'll open the third failing test because, of course, it's zero-based. 00:18:28.820 --> 00:18:31.120 Or negative two will open the second to the last one. 00:18:31.120 --> 00:18:34.160 So you can kind of hone in on it with this number here. 00:18:34.160 --> 00:18:35.780 And then you just set up your editor. 00:18:35.780 --> 00:18:39.820 If you don't set it up, you get VI or Vim or even Windows. 00:18:40.820 --> 00:18:42.540 On Windows, you get notepad. 00:18:42.540 --> 00:18:47.000 So I recommend picking a better editor. 00:18:47.000 --> 00:18:51.080 But anyway, you just set an environment variable to what you want your editor to be. 00:18:51.080 --> 00:18:54.680 And I got a ton of extras to cover, so I don't want to go too deep on this one. 00:18:54.680 --> 00:18:55.620 But there it is. 00:18:55.620 --> 00:18:56.220 Neat. 00:18:56.220 --> 00:18:58.540 Speaking of extras, what do you got for us? 00:18:58.540 --> 00:18:58.900 Any? 00:18:58.900 --> 00:19:01.280 I've got zero extras this week. 00:19:01.280 --> 00:19:03.060 I must have stolen them all. 00:19:03.060 --> 00:19:03.780 Look at my tabs. 00:19:03.780 --> 00:19:06.000 All right. 00:19:06.000 --> 00:19:07.780 Well, let's just jump into the extras then. 00:19:07.780 --> 00:19:09.840 Extra, extra. 00:19:09.840 --> 00:19:11.980 So I got a bunch here, but they're all pretty quick. 00:19:11.980 --> 00:19:21.820 Last week, you spoke about Jade, D-J-A-D-E, for formatting Django templates, right? 00:19:21.820 --> 00:19:26.500 Kind of like black or rough, but instead of being for Python, it's for Django templates and 00:19:26.500 --> 00:19:27.880 HTML and so on, which is awesome. 00:19:27.880 --> 00:19:32.600 So I'm linking to something from Adam Chains, Django pre-commit hook. 00:19:32.600 --> 00:19:33.980 Or sorry, Jade pre-commit hook. 00:19:33.980 --> 00:19:36.660 So if you want this to just happen automatically, there you go. 00:19:36.660 --> 00:19:37.440 Super easy, right? 00:19:37.440 --> 00:19:37.780 Cool. 00:19:37.780 --> 00:19:38.240 Yes. 00:19:38.240 --> 00:19:39.440 This one is really fun. 00:19:39.440 --> 00:19:45.080 And remember I talked about Polyfill, Polyfill.js and somebody took over the domain. 00:19:45.080 --> 00:19:45.880 I don't remember. 00:19:45.880 --> 00:19:50.400 It was, I think it was sold even, but if it wasn't sold, it was taken over because it lapsed 00:19:50.400 --> 00:19:50.760 or whatever. 00:19:50.760 --> 00:19:56.840 However, it got into some nefarious hands and they took over polyfill.io, boot CDN, boot 00:19:56.840 --> 00:19:58.500 CSS, static file. 00:19:58.540 --> 00:20:04.800 And they had basically rewritten a whole bunch of CDN served JavaScript to hundreds of thousands 00:20:04.800 --> 00:20:08.660 of websites with malware that would then go on to the visitors of those websites. 00:20:08.660 --> 00:20:09.240 It's bad. 00:20:09.500 --> 00:20:20.660 The only reason I'm bringing this is a quick follow-up. 00:20:20.660 --> 00:20:21.940 How do they know? 00:20:21.940 --> 00:20:28.420 That person had published, they accidentally published their Cloudflare secret keys. 00:20:30.500 --> 00:20:37.640 And so then the security researchers will, you know, they basically logged into Cloudflare 00:20:37.640 --> 00:20:41.740 and just looked at their cloud setup, which I think this is awesome. 00:20:41.740 --> 00:20:42.360 Yeah. 00:20:44.140 --> 00:20:47.740 And then, yeah, funny thing out in the audience. 00:20:47.740 --> 00:20:48.520 We'll share that in a second. 00:20:48.520 --> 00:20:48.800 All right. 00:20:48.800 --> 00:20:50.500 This one is also really cool. 00:20:50.500 --> 00:20:55.120 Over at pythonbytes.fm, created that site seven years ago. 00:20:55.120 --> 00:20:55.800 How long have we been there? 00:20:55.800 --> 00:20:56.260 Eight years? 00:20:56.260 --> 00:20:58.540 Well, when I created it, I created it with Bootstrap, okay? 00:20:58.540 --> 00:21:04.360 Because Bootstrap was probably the nicest way at that time to get a good looking site that's 00:21:04.360 --> 00:21:07.260 consistent, responsive, all those things up in Atom, right? 00:21:07.260 --> 00:21:14.440 Well, if you look at the Bootstrap CSS file, it's about 140 kilobytes, which is a lot to 00:21:14.440 --> 00:21:18.480 just be serving as one of your static resources on every web request, right? 00:21:18.480 --> 00:21:20.720 But it's pretty gnarly. 00:21:20.720 --> 00:21:21.960 You don't want to mess with it, right? 00:21:21.960 --> 00:21:22.700 Like, what are you using? 00:21:22.700 --> 00:21:23.220 I don't know. 00:21:23.220 --> 00:21:24.840 And what's the knock-on effect? 00:21:24.840 --> 00:21:30.600 So I ran across this thing called Purge CSS, which is a Node.js CLI tool, but it doesn't 00:21:30.600 --> 00:21:31.840 really matter how it's implemented, right? 00:21:31.840 --> 00:21:32.800 It's just a CLI tool. 00:21:32.800 --> 00:21:39.280 And you can run this thing against your CSS file, and you can say, hey, this CSS file, 00:21:39.280 --> 00:21:44.940 I would like you to look at all the styles that I'm using, look at the CSS file, and if 00:21:44.940 --> 00:21:48.340 there's any styles in there I'm not using, take them out of the CSS file. 00:21:48.340 --> 00:21:48.940 Oh, wow. 00:21:48.940 --> 00:21:49.320 Yeah. 00:21:49.320 --> 00:21:53.040 So, and it doesn't have to work on just HTML. 00:21:53.040 --> 00:21:57.360 I pointed at chameleon templates, or it could be Jinja templates, or we just talked about 00:21:57.360 --> 00:21:57.960 Django templates. 00:21:57.960 --> 00:22:03.680 You just give it a, like a star, star, slash, star, whatever your template extension is, 00:22:03.680 --> 00:22:06.040 like .j, or if it was Jinja, or whatever. 00:22:06.040 --> 00:22:10.020 And then it'll go through and look at all of those that are being used, and then it'll 00:22:10.020 --> 00:22:12.240 rip them out, rip out all the stuff that wasn't. 00:22:12.240 --> 00:22:20.660 So now, for example, if you go to pythonbytes.org, and you click on the CSS file, it's like 20k, 00:22:20.660 --> 00:22:22.480 25k, something like that. 00:22:22.540 --> 00:22:26.880 So it's six, seven times smaller, and all I had to do is run a command line thing against 00:22:26.880 --> 00:22:27.040 it. 00:22:27.040 --> 00:22:30.100 So maybe you're not using enough bootstrap. 00:22:30.100 --> 00:22:33.440 I feel like that's still 20kb of bootstrap. 00:22:33.440 --> 00:22:34.480 That's a lot of bootstrap. 00:22:34.480 --> 00:22:38.840 I love your attitude, but there we go. 00:22:38.840 --> 00:22:47.260 Last week, on the first go-round of 3.13 release, there's also Python 3.12.7 released. 00:22:47.780 --> 00:22:52.200 So just so you know, if you're not ready for 3.13, there's also a new version of 3.12. 00:22:52.200 --> 00:22:55.320 I want to give a quick shout-out to a video by Hynek. 00:22:55.320 --> 00:22:58.500 It says, uv is the future of Python packaging. 00:22:58.500 --> 00:23:00.400 Snake box emojis. 00:23:00.400 --> 00:23:02.500 Anyway, this is a really good video. 00:23:02.500 --> 00:23:07.500 Hynek is getting his YouTube game on quite well, and he's fun to watch. 00:23:07.500 --> 00:23:08.300 So check that out. 00:23:08.300 --> 00:23:09.360 I'll link that as well. 00:23:09.360 --> 00:23:11.760 And I still got some more. 00:23:11.760 --> 00:23:12.920 We're just tabs bumping around. 00:23:13.460 --> 00:23:18.020 Chris over on Faucet On pointed out, hey, love all this uv talk, grouping these together 00:23:18.020 --> 00:23:19.920 on 4.3 and others and so on. 00:23:19.920 --> 00:23:23.520 It's interesting to be seeing it show up just broadly now. 00:23:23.520 --> 00:23:27.420 And things like Home Assistant is now using it to reduce download and setup times. 00:23:27.420 --> 00:23:28.140 Oh, yeah. 00:23:28.140 --> 00:23:28.440 Cool. 00:23:28.440 --> 00:23:29.320 Yeah, kind of cool. 00:23:29.320 --> 00:23:34.740 Over here, Brian, do you think people are getting tired of LLMs and hearing about AI tools? 00:23:34.740 --> 00:23:35.420 Yes. 00:23:35.420 --> 00:23:38.360 I feel like there's a lot of AI fatigue out there. 00:23:38.360 --> 00:23:41.560 I'm like, ah, but is that just me projecting this onto people? 00:23:41.560 --> 00:23:42.900 Is this real? 00:23:42.900 --> 00:23:49.160 So I posted on Faucet On and on Twitter and said, hey, just thinking about topics, how 00:23:49.160 --> 00:23:52.360 do you guys feel about LLM topics and spending time talking about them? 00:23:52.360 --> 00:23:55.160 Only 4% said, more LLMs, please. 00:23:55.160 --> 00:23:57.580 22% said, ah, they're fine. 00:23:57.580 --> 00:24:00.000 38% said, I've had enough. 00:24:00.000 --> 00:24:04.280 And 36% fully said, if you say AI one more time, I will pile drive you. 00:24:04.280 --> 00:24:07.060 Well, you just said it like four times. 00:24:07.060 --> 00:24:07.400 Oh, no. 00:24:07.400 --> 00:24:09.000 It's going to hurt so much, but I had to say it. 00:24:09.000 --> 00:24:12.540 But that's 74% of people saying, oh, man, I've had enough. 00:24:12.540 --> 00:24:12.880 Yeah. 00:24:12.880 --> 00:24:13.360 So anyway. 00:24:13.360 --> 00:24:17.660 One of the things that would have been funny is if you'd have added, do you want us to start 00:24:17.660 --> 00:24:19.160 talking about blockchain again instead? 00:24:19.160 --> 00:24:21.840 Oh, my gosh. 00:24:21.840 --> 00:24:22.100 Yes. 00:24:22.100 --> 00:24:23.640 That would definitely do it. 00:24:23.640 --> 00:24:23.980 All right. 00:24:23.980 --> 00:24:26.020 A couple real quick ones here. 00:24:26.020 --> 00:24:32.700 JetBrains and the PSF have now released or opened up the Python developer survey 2024. 00:24:33.220 --> 00:24:34.460 So check that out. 00:24:34.460 --> 00:24:36.640 I'll link to how you can go and take it. 00:24:36.640 --> 00:24:38.120 But yeah, everyone participate. 00:24:38.120 --> 00:24:39.000 This is super important. 00:24:39.000 --> 00:24:43.340 We really value all the information that we get in the spring or whenever we get the answers, 00:24:43.340 --> 00:24:44.700 the responses and the write up. 00:24:44.700 --> 00:24:47.500 But it's time to submit your notes. 00:24:47.500 --> 00:24:47.980 Okay. 00:24:47.980 --> 00:24:49.720 A couple real quick things. 00:24:49.840 --> 00:24:54.220 I added a really cool way to explore courses over at Talk Python. 00:24:54.220 --> 00:24:57.200 So there's a big tag cloud up at the top here. 00:24:57.200 --> 00:24:58.560 And we get all the courses. 00:24:58.560 --> 00:25:03.500 But now if you just click around, like if you want to find stuff about test, right, I just 00:25:03.500 --> 00:25:06.760 click on testing and it'll show like your test course. 00:25:06.760 --> 00:25:08.280 And this thing is super neat. 00:25:08.280 --> 00:25:10.320 This is all implemented with HTMX. 00:25:10.320 --> 00:25:11.100 All dynamic. 00:25:11.360 --> 00:25:13.400 No page refreshes, but no JavaScript. 00:25:13.400 --> 00:25:14.360 Love it. 00:25:14.360 --> 00:25:14.800 Cool. 00:25:14.800 --> 00:25:16.140 Anyway, go check that out. 00:25:16.140 --> 00:25:18.300 And very last extra, extra, extra. 00:25:18.300 --> 00:25:22.300 I wrote an article that says, let's go easy on PyPI, okay? 00:25:22.300 --> 00:25:25.360 And I got some nice feedback from people on Updated. 00:25:25.360 --> 00:25:31.680 But basically, it talks about a bunch of different steps that you can take to not keep that number 00:25:31.680 --> 00:25:36.700 at 66 petabytes a month hitting PyPI.org, right? 00:25:36.700 --> 00:25:37.080 Yeah. 00:25:37.080 --> 00:25:42.860 I don't know how you feel, but my assessment is fully 75% or more of that has to be Docker 00:25:42.860 --> 00:25:45.840 builds, continuous integration, all that crap, right? 00:25:45.840 --> 00:25:52.000 It just, it has to be so over the top, just like machines talking to machines like, hey, 00:25:52.000 --> 00:25:54.460 here's a requirements file with 200 things in it. 00:25:54.460 --> 00:25:56.240 We're going to build a Docker container on it. 00:25:56.240 --> 00:26:00.660 So let's start with this base one, like the Python one or whichever one, it doesn't really 00:26:00.660 --> 00:26:02.020 matter and just build from it. 00:26:02.020 --> 00:26:08.000 Every single time you want any kind of code change published or tested, it's going to download 00:26:08.000 --> 00:26:10.040 all 200 over and over, right? 00:26:10.040 --> 00:26:10.500 Yeah. 00:26:10.500 --> 00:26:13.320 And then you're testing it against multiple versions of Python. 00:26:13.320 --> 00:26:15.480 So it downloads it that many more times. 00:26:15.480 --> 00:26:16.480 That's a really good point. 00:26:16.480 --> 00:26:16.760 Yeah. 00:26:16.760 --> 00:26:22.180 So this article talks you through different steps you can take to cache it over and over 00:26:22.180 --> 00:26:22.980 faster and faster. 00:26:22.980 --> 00:26:30.420 So basically, almost, unless a requirement changes, it will not ever, ever touch PyPI.org. 00:26:30.640 --> 00:26:30.840 Ever. 00:26:30.840 --> 00:26:31.660 Right? 00:26:31.660 --> 00:26:34.740 So it starts out just saying, oh yeah, you just pip install your requirements. 00:26:34.740 --> 00:26:35.680 What's wrong with that? 00:26:35.680 --> 00:26:37.780 You know, every code change causes a download. 00:26:37.780 --> 00:26:43.720 And then maybe you could run the requirements, copy just the requirements file over and run 00:26:43.720 --> 00:26:45.160 it first and then do it. 00:26:45.160 --> 00:26:47.820 That way, only if a requirement changes, does it get them all? 00:26:48.120 --> 00:26:52.440 And then you could pre-list some of your top level requirements. 00:26:52.440 --> 00:26:54.120 That would cache it in the Docker image. 00:26:54.120 --> 00:26:58.140 And then after a lot of conversations with a lot of people, there's this mount command 00:26:58.140 --> 00:27:02.600 that you can say for cache on a per command basis. 00:27:02.740 --> 00:27:08.400 And this will cache the output to somewhere on the machine that does the building of the 00:27:08.400 --> 00:27:09.060 Docker container. 00:27:09.060 --> 00:27:12.240 So it'll cache either the uv or the pip cache. 00:27:12.480 --> 00:27:16.580 So that way, if you get it again, it'll just go using cached version, even if it's a completely 00:27:16.580 --> 00:27:20.560 fresh image, long as the machine has seen that download before, right? 00:27:20.560 --> 00:27:21.400 That dependency. 00:27:21.900 --> 00:27:26.700 And then finally, if you're on probably one of the biggest offenders in the sense of CI 00:27:26.700 --> 00:27:31.820 on GitHub, over at GitHub, they actually have this thing called, maybe you know it, this 00:27:31.820 --> 00:27:36.120 thing called a cache action or actions slash cache. 00:27:36.120 --> 00:27:42.580 And you can set up a cache for your CI so that subsequent CI builds don't download and install 00:27:42.580 --> 00:27:43.660 it again as well. 00:27:43.660 --> 00:27:48.360 This will not just make it lighter on PyPI, but also every one of these things makes it faster 00:27:48.360 --> 00:27:48.780 for you. 00:27:48.780 --> 00:27:50.460 So it's like incentives are aligned. 00:27:50.460 --> 00:27:51.000 Do you know what I mean? 00:27:51.000 --> 00:27:51.460 Yeah. 00:27:51.740 --> 00:27:51.980 Yeah. 00:27:51.980 --> 00:27:52.720 That's pretty cool. 00:27:52.720 --> 00:27:58.680 It's almost too bad that like GitHub itself can't notice that like multiple packages or 00:27:58.680 --> 00:28:03.740 multiple builds on different projects are using the same version of something and caches. 00:28:03.740 --> 00:28:04.020 Right. 00:28:04.020 --> 00:28:07.940 Just have a universal pip cache or something like that. 00:28:07.940 --> 00:28:11.920 One of my questions is, and I don't know if you know the answer to this, is I know that 00:28:11.920 --> 00:28:13.280 UV does a lot of caching. 00:28:13.280 --> 00:28:18.920 Will just using uv help with reducing hitting PyPI? 00:28:18.920 --> 00:28:20.660 That's an interesting question. 00:28:20.660 --> 00:28:22.960 I don't know if it'll make it less traffic. 00:28:22.960 --> 00:28:24.400 It'll definitely make it faster. 00:28:24.400 --> 00:28:25.460 There's no doubt about that. 00:28:25.460 --> 00:28:27.500 But it might be the same traffic. 00:28:27.500 --> 00:28:27.880 Okay. 00:28:27.880 --> 00:28:28.320 Yeah. 00:28:28.320 --> 00:28:28.860 Yeah. 00:28:28.860 --> 00:28:28.940 Yeah. 00:28:28.940 --> 00:28:29.380 Cool. 00:28:29.380 --> 00:28:29.980 All right. 00:28:31.240 --> 00:28:34.600 Henry says, set up dash uv has a cache option. 00:28:34.600 --> 00:28:35.080 Excellent. 00:28:35.080 --> 00:28:35.080 Excellent. 00:28:35.080 --> 00:28:36.120 That's awesome to hear. 00:28:36.120 --> 00:28:36.680 All right. 00:28:36.680 --> 00:28:41.620 And then Jay Miller dropped in to say that he assumed he wouldn't be able to find this episode. 00:28:41.620 --> 00:28:44.020 But like us, he eventually found his way. 00:28:44.020 --> 00:28:49.540 And also, Tony out in the audience says, I was literally working on caching in CI while listening. 00:28:49.540 --> 00:28:52.220 UV does support caching in CI too. 00:28:52.360 --> 00:28:52.460 Yeah. 00:28:52.460 --> 00:28:57.460 And by the way, if you're checking out these notes on the GitHub CI stuff down here for 00:28:57.460 --> 00:29:00.760 the pip example, somewhere, pip, are you here? 00:29:00.760 --> 00:29:07.620 If you just say this patch of .cache slash pip, if you just change that to .cache slash uv, 00:29:07.620 --> 00:29:10.280 I'm pretty sure everything else will just flow from there. 00:29:10.280 --> 00:29:11.160 So awesome. 00:29:11.160 --> 00:29:11.660 Nice. 00:29:12.520 --> 00:29:12.960 Indeed. 00:29:12.960 --> 00:29:14.060 All right. 00:29:14.060 --> 00:29:16.480 Let's be ready for a joke. 00:29:16.480 --> 00:29:18.840 Well, I was hoping you didn't have a lost the joke. 00:29:18.840 --> 00:29:20.700 So let's see if I can find it. 00:29:20.700 --> 00:29:21.580 Let's see what I can find. 00:29:21.580 --> 00:29:21.940 All right. 00:29:21.940 --> 00:29:25.760 So, you know, when you go to a page, that's bad. 00:29:25.760 --> 00:29:30.120 Like if you go to Python bytes slash blah, blah, blah, and it's not there, should it just 00:29:30.120 --> 00:29:30.760 say 404? 00:29:30.760 --> 00:29:33.580 No, you should have something amusing and interesting. 00:29:33.580 --> 00:29:38.740 Like for us, we have a big empty microphone in a stage with nobody at the microphone that 00:29:38.740 --> 00:29:39.700 just says 404. 00:29:39.700 --> 00:29:41.460 Try our episodes listing instead. 00:29:41.520 --> 00:29:43.240 And you can click on it and it takes you to our episodes, right? 00:29:43.240 --> 00:29:44.060 That's all fun. 00:29:44.060 --> 00:29:49.540 So my joke, which I think I'll just let people link to, but maybe we can find one or two just 00:29:49.540 --> 00:29:53.900 give people a sense, is 28 clever and funny 404 pages. 00:29:53.900 --> 00:29:55.600 So I love this. 00:29:55.600 --> 00:29:56.560 These are so good. 00:29:56.560 --> 00:30:01.500 It almost makes me want to like mess up a website just to see what they've got, right? 00:30:01.500 --> 00:30:03.460 Like one of them is for a blue path. 00:30:03.460 --> 00:30:07.000 I don't know what that is, but there's a big map with tons of little dots of like, here's 00:30:07.000 --> 00:30:10.060 all the distribution of where people live around the city or things happening. 00:30:10.060 --> 00:30:11.380 It says, this is everyone else. 00:30:11.480 --> 00:30:15.100 And then like way over somewhere else, it has a little dot says, this is you way off 00:30:15.100 --> 00:30:15.480 the map. 00:30:15.480 --> 00:30:16.220 Please go find. 00:30:16.220 --> 00:30:21.360 Some of them are not that great, but some of them are real good. 00:30:22.160 --> 00:30:29.380 So anyway, if this kind of stuff amuses you, Pixar has one with a crying inside out person. 00:30:29.380 --> 00:30:30.680 It's sad emotion. 00:30:30.680 --> 00:30:31.120 Yeah. 00:30:31.120 --> 00:30:32.120 Yeah, exactly. 00:30:32.120 --> 00:30:32.880 Exactly. 00:30:32.880 --> 00:30:34.140 Others are just weird. 00:30:34.140 --> 00:30:38.720 One has a bear wearing like a gold 404 baller type. 00:30:38.720 --> 00:30:41.880 Anyway, check it out. 00:30:42.120 --> 00:30:45.060 Lego has a unassembled disconnected sort of thing. 00:30:45.060 --> 00:30:49.240 So I thought that would be pretty good for our episode 404. 00:30:49.240 --> 00:30:51.060 People can scroll through that and enjoy it. 00:30:51.060 --> 00:30:51.520 Yeah. 00:30:51.520 --> 00:30:54.040 And also, I love watching the, looking at these. 00:30:54.040 --> 00:30:59.240 So if anybody has a, has some good 404 pages that they, that they want to share with us, 00:30:59.240 --> 00:31:05.940 head on over to Mastodon and, or our show notes page and send us, send us a link to your favorite 00:31:05.940 --> 00:31:06.400 404. 00:31:06.400 --> 00:31:06.920 That'd be fun. 00:31:06.920 --> 00:31:07.300 Yeah. 00:31:07.300 --> 00:31:07.800 Awesome. 00:31:07.800 --> 00:31:09.040 All right, guys. 00:31:09.040 --> 00:31:09.620 Thank you. 00:31:09.620 --> 00:31:11.880 Thank you everyone for listening. 00:31:11.880 --> 00:31:12.680 Brian. 00:31:12.680 --> 00:31:13.960 Thanks for being here as always. 00:31:13.960 --> 00:31:14.440 Thanks.