WEBVTT 00:00:00.001 --> 00:00:05.920 Have you ever waited around for pip to do its thing while installing packages or syncing virtual environments, 00:00:05.920 --> 00:00:09.280 or even through some higher-level tools such as pip-tools? 00:00:09.280 --> 00:00:14.000 Then you'll be very excited to hear about what just got announced from Astral, 00:00:14.000 --> 00:00:17.480 a PIP-compatible CLI tool called uv. 00:00:17.480 --> 00:00:19.960 It's like pip, but 100 times faster. 00:00:19.960 --> 00:00:24.480 Charlie Marsh from Rough Fame and the founder of Astral is here to dive in. 00:00:24.480 --> 00:00:25.240 Let's go. 00:00:25.740 --> 00:00:31.220 This is Talk Python To Me, episode 453, recorded March 12, 2024. 00:00:31.220 --> 00:00:49.380 Welcome to Talk Python To Me, a weekly podcast on Python. 00:00:49.380 --> 00:00:51.120 This is your host, Michael Kennedy. 00:00:51.120 --> 00:00:58.620 Follow me on Mastodon, where I'm @mkennedy, and follow the podcast using @talkpython, both on fosstodon.org. 00:00:58.620 --> 00:01:03.700 Keep up with the show and listen to over seven years of past episodes at talkpython.fm. 00:01:03.700 --> 00:01:07.440 We've started streaming most of our episodes live on YouTube. 00:01:07.440 --> 00:01:15.020 Subscribe to our YouTube channel over at talkpython.fm/youtube to get notified about upcoming shows and be part of that episode. 00:01:15.500 --> 00:01:18.020 This episode is sponsored by Neo4j. 00:01:18.020 --> 00:01:26.180 It's time to stop asking relational databases to do more than they were made for and simplify complex data models with graphs. 00:01:26.180 --> 00:01:33.020 Check out the sample FastAPI project and see what Neo4j, a native graph database, can do for you. 00:01:33.020 --> 00:01:37.440 Find out more at talkpython.fm/Neo4j. 00:01:38.100 --> 00:01:41.900 And it's also brought to you by us over at Talk Python Training. 00:01:41.900 --> 00:01:46.520 Did you know that we have over 250 hours of Python courses? 00:01:46.520 --> 00:01:47.720 Yeah, that's right. 00:01:47.720 --> 00:01:50.280 Check them out at talkpython.fm/courses. 00:01:50.280 --> 00:01:52.300 Hey, y'all. 00:01:52.300 --> 00:01:55.340 I have a quick announcement before we jump into the conversation with Charlie. 00:01:55.340 --> 00:01:57.880 And sticking with the theme of the episode, I'll keep it quick. 00:01:57.880 --> 00:02:01.440 Over at Talk Python, we just released a new free course. 00:02:01.440 --> 00:02:08.340 It's all about working with audio in Python, converting it to text, applying all sorts of AI awesome sauce, and more. 00:02:08.340 --> 00:02:18.940 We end up with a dynamic web app that lets you have a ChatGPT style Q&A with a podcast, guests, and hosts starting from just the RSS feed. 00:02:18.940 --> 00:02:20.320 It's really mind-blowing. 00:02:20.860 --> 00:02:27.620 As a sweet bonus, the course shows you a bunch of hands-on examples of FastAPI, HTMX, Beanie, MongoDB, and more. 00:02:27.620 --> 00:02:28.740 I hope you love it. 00:02:28.740 --> 00:02:32.460 Give it a look as it is 100% free, no strings attached. 00:02:32.460 --> 00:02:38.320 The link is on the episode page, in your podcast player show notes, and at talkpython.fm/courses. 00:02:38.320 --> 00:02:40.840 Hey, Charlie. 00:02:40.840 --> 00:02:41.460 Hey. 00:02:41.460 --> 00:02:42.320 Welcome. 00:02:42.320 --> 00:02:44.300 Yeah, thanks so much for having me back on the show. 00:02:44.300 --> 00:02:45.260 I really appreciate it. 00:02:45.260 --> 00:02:47.880 Yes, it's really great to have you back on the show. 00:02:48.400 --> 00:02:53.720 You and your team are doing awesome work these days and making a big dent in Python. 00:02:53.720 --> 00:02:54.100 Thank you. 00:02:54.100 --> 00:02:54.840 Thank you. 00:02:54.840 --> 00:02:56.460 Hopefully a good dent, I guess. 00:02:56.460 --> 00:02:59.860 Yeah, I guess you could make a good dent or a bad dent. 00:02:59.860 --> 00:03:01.780 But no, I meant that in a good way. 00:03:01.780 --> 00:03:02.920 Definitely in a good way. 00:03:02.920 --> 00:03:03.700 I appreciate it. 00:03:03.700 --> 00:03:04.140 Yeah. 00:03:04.140 --> 00:03:05.800 Yeah, I think excellent work. 00:03:05.800 --> 00:03:12.280 It's really, it's kind of surprising how some of these tools just come in and just catch fire in a way. 00:03:12.280 --> 00:03:16.580 I don't know how you felt about it, but it seemed more than I would have expected. 00:03:16.900 --> 00:03:21.960 I mean, I did not expect your things to be popular, but they became really popular really quickly. 00:03:21.960 --> 00:03:23.340 Yeah. 00:03:23.340 --> 00:03:23.960 Yeah. 00:03:23.960 --> 00:03:30.160 I mean, it was different, you know, with Ruff and with uv, this sort of context around it has been pretty different. 00:03:30.160 --> 00:03:36.520 Because with Ruff, like when I started working on Ruff, no one knew who I was or had any reason to care about what I was doing. 00:03:37.220 --> 00:03:41.000 And the project just grew, you know, grew over time. 00:03:41.000 --> 00:03:47.320 And then I knew that when we were working on uv, like there would just be a lot more of an expectation around like anything we build. 00:03:47.320 --> 00:03:47.880 Yeah. 00:03:47.880 --> 00:03:49.600 The spotlight was already shining, right? 00:03:49.600 --> 00:03:50.160 A little bit. 00:03:50.160 --> 00:03:50.480 Yeah. 00:03:50.480 --> 00:03:50.860 Yeah. 00:03:50.860 --> 00:03:53.740 And that's awesome because it means our stuff gets traction. 00:03:53.740 --> 00:03:55.740 Also, it comes with a lot of responsibility. 00:03:56.060 --> 00:03:59.440 It means anything we release, we're, you know, we're kind of committed to maintaining. 00:03:59.440 --> 00:04:01.220 So, like, yeah. 00:04:01.220 --> 00:04:01.260 Yeah. 00:04:01.260 --> 00:04:01.560 Yeah. 00:04:01.560 --> 00:04:07.300 It's been very interesting, like this kind of release cadence versus what happened with Ruff. 00:04:07.300 --> 00:04:10.720 Well, you have more people behind the scenes helping you out, right? 00:04:10.720 --> 00:04:11.380 I do. 00:04:11.380 --> 00:04:11.760 Yeah. 00:04:11.760 --> 00:04:12.220 Thankfully. 00:04:12.220 --> 00:04:12.760 Yeah. 00:04:12.760 --> 00:04:20.360 Tell us, you know, maybe that's a good place to just kick off our conversation is, you know, we're talking a bit about Ruff for everyone. 00:04:20.360 --> 00:04:24.040 And Ruff was the thing that really launched all of this. 00:04:24.040 --> 00:04:27.000 But you started a company, Astral. 00:04:27.000 --> 00:04:27.560 Yep. 00:04:27.560 --> 00:04:28.840 Got some funding behind it. 00:04:28.840 --> 00:04:29.220 Yep. 00:04:29.220 --> 00:04:30.200 Building awesome tools. 00:04:30.200 --> 00:04:30.660 Yeah. 00:04:30.660 --> 00:04:32.900 Just give people the broad view of your world these days. 00:04:32.900 --> 00:04:33.780 Yeah, totally. 00:04:33.780 --> 00:04:42.020 So, yeah, I started, this all started when I released Ruff probably like a year and a half ago, maybe at this point. 00:04:42.020 --> 00:04:49.860 And at that point, it was really not just a side project for me, but very much an exploratory project. 00:04:49.860 --> 00:04:58.340 Like I was kind of curious about building faster Python tooling, really, and curious about getting better at Rust and all this stuff. 00:04:58.340 --> 00:05:07.460 And Ruff, when I released it, it's, you know, it was a lint or it is a linter, but it kind of brings a lot of different tools together. 00:05:07.460 --> 00:05:16.560 So if you're familiar with like Flategate or Pylent, but also Black, Ruff is designed to kind of bundle all those tools together. 00:05:16.560 --> 00:05:19.140 So it can detect problems with your code. 00:05:19.140 --> 00:05:20.560 It can fix them automatically. 00:05:20.560 --> 00:05:22.680 It can reformat your code. 00:05:23.260 --> 00:05:29.100 So often when we see people like adopting Ruff, they're replacing like 20 or 30 tools with Ruff. 00:05:29.100 --> 00:05:36.720 And, you know, it has, there's a bunch of nice things about it, but ultimately it's very, very fast and hopefully a lot simpler for users. 00:05:36.720 --> 00:05:38.120 Yeah. 00:05:38.120 --> 00:05:39.440 But, go ahead. 00:05:39.440 --> 00:05:44.060 I was going to say, it's really nice that it's just one tool that brings that all together. 00:05:44.060 --> 00:05:45.240 You know, it's not a patchwork. 00:05:45.240 --> 00:05:50.700 Well, here's how we use ISORT and here's how we use Flategate and here's how we use this and here's how we use that, you know? 00:05:50.940 --> 00:05:51.140 Right. 00:05:51.140 --> 00:06:00.480 I think ISORT's a really good example because instead of having an ISORT's a great tool, but instead of having a separate tool to do your import sorting, Ruff Frames import sorting is just a lint rule. 00:06:00.480 --> 00:06:05.840 And so you don't have to learn a bunch of different tools and figure out how to sort of merge them together. 00:06:06.760 --> 00:06:14.300 And, you know, Ruff, again, when I released it, it was just a side project and it sort of took off. 00:06:14.300 --> 00:06:25.080 You know, there were, it was originally, you know, I released it in part because I had this thesis that like Python tooling could be much faster and Ruff was meant to evidence that. 00:06:25.360 --> 00:06:30.840 It was meant to be an example of what I saw as like this potential to build really different Python tooling. 00:06:30.840 --> 00:06:40.940 And took off a couple of big projects like FastAPI and such were kind of early adopters and it just grew and grew and grew from there. 00:06:41.560 --> 00:06:47.080 And I started a company, Astral, you know, shortly after based on a lot of these ideas. 00:06:47.080 --> 00:06:54.640 So again, Ruff in this context was meant to be, you know, the goal wasn't necessarily to start a company just based around Ruff. 00:06:54.800 --> 00:07:07.460 The goal was to start a company based around, you know, the ideas behind Ruff, which is, can we build really different Python tooling by, well, it's all written in Rust. 00:07:07.460 --> 00:07:09.440 It's sort of an implementation detail, right? 00:07:09.440 --> 00:07:16.900 But ultimately what we're trying to do is build like really high performance Python tooling, but also tooling that works together really well and simplifies things. 00:07:16.900 --> 00:07:19.020 So tackle, it's not just about performance. 00:07:19.020 --> 00:07:23.840 It's also about tackling a lot of, I guess, what I see as the complexity in Python tooling today. 00:07:24.580 --> 00:07:36.640 And really trying to, I guess, challenge certain status quos and respect others, but challenge some status quos in what it feels like to work with and use Python. 00:07:36.640 --> 00:07:47.500 And so, you know, our focus as a company, you know, right now we're very, very focused on our open source roadmap and our open source tooling. 00:07:47.500 --> 00:07:51.340 So that's Ruff now, uv, which we'll talk about in a bit. 00:07:52.080 --> 00:07:58.200 And these are, you know, MIT licensed, you know, fully open source tools that we build. 00:07:58.200 --> 00:08:08.160 And ultimately our goal is to build and sell paid services that integrate really well with the tooling and are kind of like the natural next thing you need when you're building with Python. 00:08:08.620 --> 00:08:12.400 So, you know, we'll never charge money for like the tools themselves. 00:08:12.400 --> 00:08:14.240 Those remain free, permissively licensed. 00:08:14.240 --> 00:08:19.700 But our goal is to kind of build, you know, imagine you're building with Python, you need to deploy a web server. 00:08:19.700 --> 00:08:22.300 There's a lot of things that you need naturally to do that. 00:08:22.380 --> 00:08:30.420 So for people who are already using our tools, what are the natural next things they need and how can we build really good services on top of the tooling that just integrate with it really well? 00:08:30.420 --> 00:08:32.460 So that's kind of the premise of the company. 00:08:32.460 --> 00:08:34.760 There's like really two focuses to the roadmap. 00:08:34.760 --> 00:08:36.540 It's like the open source and the commercial. 00:08:37.400 --> 00:08:46.020 And our goal is to have an incentive structure where we're just continuously aligned to like invest a ton in the open source and grow it as much as possible in a really natural way. 00:08:46.400 --> 00:08:49.080 So you're not going to do like per line fix pricing. 00:08:49.080 --> 00:08:55.160 You get a million, million line fixes a month free and then no, I think it's an awesome model. 00:08:55.160 --> 00:09:07.540 And, you know, open source 10 years ago, it really looked like it was already super vibrant and super making huge dents in the space. 00:09:07.540 --> 00:09:07.800 Right. 00:09:07.800 --> 00:09:09.920 Really a lot of people adopting it. 00:09:10.040 --> 00:09:14.480 But I think there was kind of a funding side that was a little bit missing. 00:09:14.480 --> 00:09:18.880 I still remember like, you know, buy me a coffee type of links, like send me 10 bucks. 00:09:18.880 --> 00:09:19.580 Yeah. 00:09:19.580 --> 00:09:20.840 If you like this. 00:09:20.840 --> 00:09:22.600 And that's just it's not a job. 00:09:22.600 --> 00:09:26.580 That's hardly a hobby type of thing to get stuff done. 00:09:26.580 --> 00:09:28.840 You can't have a team of people focused on it. 00:09:28.840 --> 00:09:33.260 And it had traditionally really been around kind of a beneficial. 00:09:33.260 --> 00:09:39.920 I need to find a big tech company that's willing to give me sufficient amount of time to work on this project. 00:09:39.920 --> 00:09:42.140 To sort of fund its existence. 00:09:42.140 --> 00:09:47.400 Or maybe I'm a consultant and I'm I create Flask, but I consult on Flask, that type of thing. 00:09:47.400 --> 00:09:47.880 Right. 00:09:47.880 --> 00:09:54.460 And I think it's just super positive to see more direct ways people are making open source successful. 00:09:54.460 --> 00:09:55.600 You know, I think. 00:09:55.600 --> 00:09:56.380 Yeah, I appreciate that. 00:09:56.380 --> 00:09:56.980 I wish. 00:09:56.980 --> 00:09:57.420 Yeah, I think. 00:09:57.420 --> 00:10:02.980 You know, I wish I sort of wish that there were more more good examples of that working. 00:10:02.980 --> 00:10:04.360 There are some for sure. 00:10:05.180 --> 00:10:06.880 You know, for me, it was pretty clear. 00:10:06.880 --> 00:10:08.460 Like it was immediately clear to me. 00:10:08.460 --> 00:10:11.560 I needed to work on a rough full time if I wanted the project to succeed. 00:10:11.560 --> 00:10:20.780 And that's kind of a microcosm of the bigger the bigger challenge, which is like we wanted to build like we really wanted to like professionalize the building of this. 00:10:20.780 --> 00:10:24.360 You know, the work on this project and bring in more people to help build it. 00:10:24.360 --> 00:10:28.680 And so, you know, ultimately, it's not obviously it's not it's not charity. 00:10:28.680 --> 00:10:29.720 It is venture funding. 00:10:29.720 --> 00:10:36.600 And ultimately, we need to build a business, you know, around the tooling to make it, you know, to make it sustainable. 00:10:38.060 --> 00:10:50.660 But, you know, again, for me, I want the incentive structure to be such that like the open source is obviously free, open source, extremely permissively licensed and truly open source. 00:10:50.660 --> 00:10:52.640 Like we have tons and tons of contributors. 00:10:52.640 --> 00:10:54.100 We have a huge community around it. 00:10:55.540 --> 00:11:02.160 And I'm just glad that we've been able to, you know, bring in more people to kind of supercharge the development of it all. 00:11:02.160 --> 00:11:02.740 So, yeah. 00:11:02.740 --> 00:11:07.580 So I guess you talked on you touched on this before, but we're now eight people. 00:11:07.580 --> 00:11:14.840 So we've grown, you know, it was it was I guess a year ago it was just me in March. 00:11:14.840 --> 00:11:20.420 So over the past year, our first two team members joined actually in March of last year. 00:11:20.420 --> 00:11:21.580 So almost exactly a year. 00:11:21.580 --> 00:11:24.760 And, yeah, we're completely distributed. 00:11:25.020 --> 00:11:30.640 So we're, you know, we have people in U.S. Pacific time all the way to Bangalore. 00:11:30.640 --> 00:11:35.200 So we're just, you know, around the like open source. 00:11:35.200 --> 00:11:41.200 We're kind of all over the world and and just building, you know, building all this stuff out in the open. 00:11:41.200 --> 00:11:54.260 It's such a good fit to be a distributed company as an open source with open source roots, because the tools of open source in some degree for software in general, but especially for open source is asynchronous. 00:11:54.500 --> 00:11:55.180 Yeah. 00:11:55.180 --> 00:11:59.520 Semi-connected, you know, kind of the zen of get, but for the way you work. 00:11:59.520 --> 00:11:59.860 Right. 00:12:00.120 --> 00:12:04.240 And it makes it super easy to hire just the most engaged people, regardless. 00:12:04.240 --> 00:12:04.860 Yeah, exactly. 00:12:04.860 --> 00:12:05.240 Yeah. 00:12:05.240 --> 00:12:08.460 The people who are willing to commute to my office in the office park. 00:12:08.460 --> 00:12:09.060 Yeah. 00:12:09.060 --> 00:12:18.400 It was I would be pretty surprised if a single person I've hired would have been willing to move to New York, which is around based. 00:12:19.180 --> 00:12:25.300 So from that perspective, it was a no brainer, which is and actually, you know, multiple people we've hired. 00:12:25.300 --> 00:12:28.500 Our relationship with them started on the repo. 00:12:28.500 --> 00:12:29.620 Yeah. 00:12:29.860 --> 00:12:31.020 InRuff itself. 00:12:31.740 --> 00:12:40.760 And then almost everyone we've hired, we discovered through open source in some way or had some relationship through open source in some way. 00:12:40.760 --> 00:12:41.020 Yeah. 00:12:41.380 --> 00:12:51.320 So, yeah, it's been a really it's very much in it's also very much in the DNA of the company, which is, you know, we actually also intentionally try to hire people who have been maintainers. 00:12:51.320 --> 00:12:51.520 Right. 00:12:51.520 --> 00:12:53.300 And have a lot of experience with that, too. 00:12:53.300 --> 00:12:54.920 We view that as a really important skill. 00:12:55.320 --> 00:12:56.220 Yeah, that's excellent. 00:12:56.220 --> 00:12:59.780 So we talked about Ruff. 00:12:59.780 --> 00:13:03.480 One thing I do want to give a bit of a shout out to is I don't remember. 00:13:03.480 --> 00:13:05.460 I think I found this on Mastodon. 00:13:05.460 --> 00:13:07.560 I can't remember where somebody talked about it. 00:13:07.560 --> 00:13:12.380 But one of the things that Ruff does, one, it changes code, which is awesome. 00:13:12.380 --> 00:13:14.400 I love your Ruff format story. 00:13:14.400 --> 00:13:16.580 But it also just tells you, hey, this could be better. 00:13:16.580 --> 00:13:17.820 You're violating PEP 8. 00:13:17.820 --> 00:13:21.620 You're violating this other convention or even a security thing. 00:13:21.620 --> 00:13:21.840 Right. 00:13:21.980 --> 00:13:26.720 But the why is the why of that, I think, requires a little bit of experience. 00:13:26.720 --> 00:13:32.820 Like, why should I use this kind of loop or that kind of comprehension over another? 00:13:32.820 --> 00:13:33.400 Right. 00:13:33.400 --> 00:13:37.300 It's people told me that who are much better have been doing this longer. 00:13:37.300 --> 00:13:38.100 I shouldn't do it. 00:13:38.100 --> 00:13:38.820 Why? 00:13:38.820 --> 00:13:42.680 When you have when you have 700 rules, there's a lot of why. 00:13:42.680 --> 00:13:52.940 And so you guys came up with this thing at docs.astral.sh slash rules, put in links that literally for each one of these has like a little example. 00:13:52.940 --> 00:13:57.440 And what is good and what is bad and when you should use it and so on. 00:13:57.440 --> 00:13:59.300 You want to tell people about this a bit? 00:13:59.300 --> 00:13:59.900 I think it's a mess. 00:13:59.900 --> 00:14:00.540 Yeah, totally. 00:14:00.760 --> 00:14:04.020 So we didn't have, I mean, we obviously didn't have this when we started the project. 00:14:04.020 --> 00:14:08.060 And then we accumulated a lot of rules. 00:14:08.060 --> 00:14:12.240 Like, when we decided to add this, I think we already had like hundreds of rules. 00:14:12.240 --> 00:14:20.060 And the motivation for the format really comes from Clippy, which is the linter used in Rust. 00:14:20.580 --> 00:14:21.020 Okay. 00:14:21.020 --> 00:14:24.580 And they have really, Rust in general has a really good documentation culture. 00:14:24.580 --> 00:14:27.720 And Clippy has this nice format, very similar to ours. 00:14:27.720 --> 00:14:34.580 We've made some adjustments, but it's pretty similar around how you document lint rules, like what's the motivation and giving examples. 00:14:34.580 --> 00:14:37.300 And so we were like, all right, let's, we want to do this. 00:14:37.300 --> 00:14:46.580 And we just over a very long period of time and a lot of contributors contributed on this documentation, but it took like six months or something probably to document it. 00:14:46.580 --> 00:14:49.660 And now, yeah, now we can, now we require it for all rules. 00:14:49.660 --> 00:14:53.520 So it's much easier to add going forward, but it was a huge effort to add this. 00:14:53.520 --> 00:14:59.980 And, you know, some of the explanations too are very extensive and detailed with lots of references. 00:14:59.980 --> 00:15:01.200 Yeah. 00:15:01.200 --> 00:15:01.840 Yeah. 00:15:01.840 --> 00:15:02.100 Yeah. 00:15:02.100 --> 00:15:07.040 I pulled the one up for iSort, unsorted imports, which is error i001. 00:15:07.040 --> 00:15:13.420 And it also talks about whether it can automatically fix it, what this does, why is it bad? 00:15:13.420 --> 00:15:20.140 So it says things like consistency is good, use a common convention for imports to make your code more readable and idiomatic or Pythonic. 00:15:20.140 --> 00:15:20.380 Yep. 00:15:20.380 --> 00:15:22.760 And it just gives an example of bad, good. 00:15:22.760 --> 00:15:24.040 Right, right. 00:15:24.040 --> 00:15:25.200 I think that's nice. 00:15:25.200 --> 00:15:25.820 Yeah. 00:15:25.820 --> 00:15:28.660 And, you know, we've started to put more stuff in here over time too. 00:15:28.660 --> 00:15:36.640 Like whether, like in some cases we let users apply automatic fixes that aren't completely safe, like by opting in. 00:15:36.640 --> 00:15:39.600 Like it might change the meaning of your code and you kind of need to be careful. 00:15:39.800 --> 00:15:43.760 And so over time we've started to document that here too, like rules that have unsafe fixes. 00:15:43.760 --> 00:15:44.260 Why? 00:15:44.260 --> 00:15:47.460 Like in what case might it break your code and what should you look out for? 00:15:47.460 --> 00:15:55.300 So, yeah, we want it to be a tool with really good, like we're kind of inspired by Rust in this way. 00:15:55.300 --> 00:16:01.320 Because Rust has a really good documentation culture and also the Rust like compiler itself has really good error messages. 00:16:02.020 --> 00:16:06.280 It's sort of famous for that, like when your code doesn't work, it's really good at telling you why. 00:16:06.280 --> 00:16:09.660 It doesn't always get it right, but it has famously good error messages. 00:16:09.660 --> 00:16:19.400 And that's something that we try and channel too, to like when we make changes to configuration, like just putting extra effort into trying to make like those, those like error messages and those hints helpful. 00:16:19.400 --> 00:16:20.680 But yeah, it's a lot of work. 00:16:20.680 --> 00:16:21.480 So I appreciate you. 00:16:21.480 --> 00:16:22.760 I appreciate that you called it out. 00:16:23.860 --> 00:16:29.720 I'm sure it looks like a lot of work, but it's one of those things that's, where else are you going to get this knowledge, right? 00:16:29.720 --> 00:16:34.540 I mean, I know you can go search for one at a time, but in this aggregates is really good. 00:16:34.540 --> 00:16:35.220 Right, right, right. 00:16:37.220 --> 00:16:40.560 This portion of Talk Python To Me is brought to you by Neo4j. 00:16:40.560 --> 00:16:42.500 Do you know Neo4j? 00:16:42.500 --> 00:16:45.240 Neo4j is a native graph database. 00:16:45.240 --> 00:16:50.100 And if the slowest part of your data access patterns involves computing relationships, 00:16:50.100 --> 00:16:57.340 why not use a database that stores those relationships directly in the database, unlike your typical relational one? 00:16:57.340 --> 00:17:04.020 A graph database lets you model the data the way it looks in the real world, instead of forcing it into rows and columns. 00:17:04.460 --> 00:17:11.640 It's time to stop asking a relational database to do more than they were made for and simplify complex data models with graphs. 00:17:11.640 --> 00:17:16.720 If you haven't used a graph database before, you might be wondering about common use cases. 00:17:16.720 --> 00:17:17.600 You know, what's it for? 00:17:17.600 --> 00:17:19.000 Here are just a few. 00:17:19.000 --> 00:17:20.120 Detecting fraud. 00:17:20.120 --> 00:17:21.640 Enhancing AI. 00:17:21.640 --> 00:17:23.440 Managing supply chains. 00:17:23.440 --> 00:17:26.320 Gaining a 360-degree view of your data. 00:17:26.320 --> 00:17:29.680 And anywhere else you have highly connected data. 00:17:29.680 --> 00:17:34.380 To use Neo4j from Python, it's a simple pip install. 00:17:34.380 --> 00:17:35.140 Neo4j. 00:17:35.140 --> 00:17:42.260 And to help you get started, their docs include a sample web app demonstrating how to use it both from Flask and FastAPI. 00:17:42.260 --> 00:17:47.280 Find it in their docs or search GitHub for Neo4j Movies Application Quick Start. 00:17:47.280 --> 00:17:51.460 Developers are solving some of the world's biggest problems with graphs. 00:17:51.460 --> 00:17:52.660 Now it's your turn. 00:17:52.660 --> 00:17:56.960 Visit talkpython.fm/Neo4j to get started. 00:17:57.180 --> 00:18:00.900 That's talkpython.fm/Neo4j. 00:18:00.900 --> 00:18:05.000 Thank you to Neo4j for supporting Talk Python To Me. 00:18:06.320 --> 00:18:09.300 So Ruff will fix a lot of these things for you. 00:18:09.300 --> 00:18:14.820 And one of the things that I like to do is I just have it integrated into my editor these days. 00:18:14.820 --> 00:18:21.000 So both with PyCharm and with VS Code, if I just say format this document, that's a Ruff format. 00:18:21.000 --> 00:18:21.660 Nice. 00:18:21.660 --> 00:18:22.920 Yeah, yeah. 00:18:23.000 --> 00:18:25.460 So we have a VS Code extension. 00:18:25.460 --> 00:18:30.660 And it supports formatting, autofixing, all that kind of stuff. 00:18:30.660 --> 00:18:35.380 The PyCharm one's a third-party one, right? 00:18:35.380 --> 00:18:36.680 Someone else decided to make that. 00:18:36.840 --> 00:18:39.640 Yeah, the JetBrains one is by Huda. 00:18:39.640 --> 00:18:45.160 And it's, yeah, that's like a community extension. 00:18:45.160 --> 00:18:53.260 We're considering doing our own at some point because there's a lot of demand for, obviously, we have a lot of users who use the JetBrains stuff. 00:18:53.260 --> 00:18:54.400 I mean, I use it. 00:18:54.400 --> 00:18:58.060 So we may do our own at some point, but there's a community-maintained one right now. 00:18:58.060 --> 00:19:02.040 And that's based on, we have sort of like a language server. 00:19:02.040 --> 00:19:08.320 So there's kind of like an underlying piece of technology that wraps rough, that powers all of these editor integrations. 00:19:08.320 --> 00:19:13.220 So you can also use that from like NeoVim or Sublime Text or Emacs. 00:19:13.220 --> 00:19:15.580 All of those editors support the LSP. 00:19:15.580 --> 00:19:20.620 And so you can actually, it's the same thing that powers the VS Code extension. 00:19:20.620 --> 00:19:23.440 Just that's wrapped in some VS Code specific stuff. 00:19:23.440 --> 00:19:25.420 But it's the same piece of technology. 00:19:26.920 --> 00:19:34.320 We're actually rewriting that right now to make it sort of like more natively integrated into Ruff, which will let us do some cool things. 00:19:34.320 --> 00:19:41.860 So it should get like actually significantly better, I think, over the course of this year because it's becoming more and more of a priority for us. 00:19:41.860 --> 00:19:42.640 Yeah, cool. 00:19:42.640 --> 00:19:44.900 The whole LSP thing is pretty interesting. 00:19:44.900 --> 00:19:47.900 I think it's really opened up a lot of editors being better. 00:19:47.900 --> 00:19:49.680 Yeah, it has. 00:19:49.680 --> 00:19:55.540 And it made it, like from our perspective, it made it way easier to build like editor integrations. 00:19:55.780 --> 00:20:01.600 Because we just built an LSP and then it like works for like everyone, almost. 00:20:01.600 --> 00:20:03.260 Yeah. 00:20:04.020 --> 00:20:11.320 Like the JetBrains stuff, it does support the LSP, but I think only in the paid version, not in the community version. 00:20:11.320 --> 00:20:13.220 So like there's some limitations around it. 00:20:13.220 --> 00:20:20.600 But in general, like if you support the LSP, like it's actually very easy to build editor integrations that work with everyone, which is like super, super useful. 00:20:20.600 --> 00:20:21.260 Yeah. 00:20:21.260 --> 00:20:26.020 Instead of rebuilding your own integration for every single thing, which is not fun. 00:20:26.240 --> 00:20:27.020 Yeah, it's a lot. 00:20:27.020 --> 00:20:27.700 All right. 00:20:27.700 --> 00:20:31.400 Well, let's talk about the main project here. 00:20:31.400 --> 00:20:35.360 The main reason for being here today is uv. 00:20:35.360 --> 00:20:37.260 Tell people what uv is. 00:20:37.260 --> 00:20:40.460 This is your next big project, next big tool. 00:20:40.460 --> 00:20:40.840 Yes. 00:20:40.840 --> 00:20:41.440 Yeah. 00:20:41.440 --> 00:20:46.180 UV, we released uv a little under a month ago. 00:20:47.180 --> 00:20:59.860 And this is something that I've wanted to do, like basically since we started the company, I wanted to start, I thought there was an opportunity to build some really interesting tooling in Python packaging. 00:21:00.420 --> 00:21:03.220 And uv is kind of the first milestone in that. 00:21:03.220 --> 00:21:17.660 So uv, in its current form, it's designed as a sort of drop-in alternative to pip, pip-tools, like if you use pip compile, pip sync, and virtualenv. 00:21:17.660 --> 00:21:21.220 So it takes those three tools and tries to bundle them into one. 00:21:22.820 --> 00:21:32.720 And so it can do things like, you know, given a set of input requirements, generate a requirements.txt file similar to what you'd get from pip compile. 00:21:32.720 --> 00:21:39.820 It can, you know, you can do uv pip install-r requirements.txt, uv pip install black. 00:21:40.820 --> 00:21:53.380 So it's really intended right now to be sort of, you know, it won't always, we don't support absolutely 100% of what pip does, but it's, we try to be pretty close. 00:21:53.380 --> 00:22:07.620 And so in many cases, the intention is that people can basically drop in uv in their project that uses pip today and get, in many cases, like very, very significant performance improvements. 00:22:07.720 --> 00:22:14.700 So similar to Ruff, like some of the goals we had for this project were one, it should be very, very fast. 00:22:14.700 --> 00:22:18.200 And uv, it's really fast. 00:22:18.200 --> 00:22:25.940 In particular, there's like a benchmarks.md in here, actually, if you want to click on that. 00:22:25.940 --> 00:22:27.540 Now scroll up. 00:22:27.540 --> 00:22:28.380 Does that do it? 00:22:28.380 --> 00:22:29.980 Oh yeah, that works too, I guess. 00:22:29.980 --> 00:22:31.360 That was smart of me. 00:22:31.360 --> 00:22:41.800 Yeah, so it's really, really fast, especially if you've already have, if you're, if you have things cached, which is pretty common. 00:22:41.800 --> 00:22:51.340 So like often on your machine, when you're doing like pip install-r, whatever, a lot of those packages, you've probably installed them at least like once before on your machine. 00:22:52.400 --> 00:22:57.920 Or maybe you deleted your virtual environment and you're recreating it with the exact same dependencies or something like that. 00:22:57.920 --> 00:23:02.160 UV is like very, very well optimized for that case. 00:23:02.160 --> 00:23:08.200 So if you have a package that you're installing multiple times, like the installation is effectively free. 00:23:08.200 --> 00:23:12.340 And it's also a lot more like disk space efficient. 00:23:12.540 --> 00:23:20.000 We use kind of like, we store all of them in one place and kind of copy them into your virtual environment so that you don't have like many, many different copies of the package too. 00:23:20.000 --> 00:23:36.440 So this was all kind of taking techniques and inspiration from like other ecosystems and other package managers and kind of molding them to fit Python's model for how packaging and dependencies work and all the specifications. 00:23:37.320 --> 00:23:44.640 And so again, the end result is like you, you know, installing into a virtual environment, a package that you've already installed before is basically free. 00:23:44.640 --> 00:23:49.600 And yeah, a lot of these things are just like way, way, way faster. 00:23:49.600 --> 00:23:52.640 Similar to rough, like dramatically faster. 00:23:52.640 --> 00:23:53.100 Yeah. 00:23:53.100 --> 00:23:53.440 Yeah. 00:23:53.440 --> 00:23:54.280 And similar to radically. 00:23:54.280 --> 00:24:01.180 And I think, you know, it's worth pointing out that like on one hand, we all have an extra 10 seconds, right? 00:24:01.180 --> 00:24:03.220 Or of course, whatever. 00:24:04.220 --> 00:24:09.920 But in the aggregate, all of these things are just like little paper cuts over your day, you know? 00:24:09.920 --> 00:24:11.000 And so, yeah, sure. 00:24:11.000 --> 00:24:15.340 I can format the code and I could use something nice and sure if it takes 20 seconds. 00:24:15.340 --> 00:24:15.700 Okay. 00:24:15.700 --> 00:24:19.660 But then you stop running it because it's kind of a hassle and you're just, you're in your flow. 00:24:19.660 --> 00:24:21.940 And this is kind of the same way. 00:24:21.940 --> 00:24:28.260 It's like, oh, I want to maybe see if there's a new, some new packages for this project as I'm sitting down to work on it. 00:24:28.260 --> 00:24:37.660 Like, okay, well, wait for a pip compile to run and you wait and you wait and then like, okay, now it's done the onto the development version and you wait and you wait. 00:24:37.660 --> 00:24:40.720 And like, you know, it's just some of these things that are instant. 00:24:40.720 --> 00:24:45.400 You just ask the question more and it just doesn't break your train of thought, you know? 00:24:45.400 --> 00:24:45.880 Totally. 00:24:46.320 --> 00:24:50.860 I think there's like maybe like three different things that come to mind here when we think about performance. 00:24:50.860 --> 00:24:52.180 Because like, yeah, okay. 00:24:52.180 --> 00:24:58.740 In this case, like it may not seem to change your life if you're going from one second to, you know, 50 milliseconds. 00:24:58.740 --> 00:25:00.860 But I think three things come to mind. 00:25:00.860 --> 00:25:06.380 So one, like for companies and large projects, this can actually be like a really big difference. 00:25:07.380 --> 00:25:15.620 And so if you're at a company and you have like a big monorepo with like a bunch of different sub projects, they all have requirements files and you want to bump a dependency. 00:25:15.620 --> 00:25:22.080 That can be like, at that point you're talking like 15 plus minutes for a command. 00:25:22.080 --> 00:25:28.320 And so like for the larger the project gets, obviously the more performance matters and the more it helps out. 00:25:28.320 --> 00:25:30.360 And we're trying to build tooling that can basically scale. 00:25:30.540 --> 00:25:35.580 I don't know about like to arbitrarily large projects, but like two large projects and useful for companies too. 00:25:35.580 --> 00:25:47.980 You know, the other thing I'd say is I think there's something that happens like when tools, when you make something like way faster, it just changes a lot of the ergonomics around what it's like to do things. 00:25:47.980 --> 00:25:57.540 And so, you know, in this case, like, okay, so installing into a virtual environment, a package that you've already installed before is free now. 00:25:57.540 --> 00:25:59.280 And so how does that change things? 00:25:59.280 --> 00:26:00.360 It's nearly instant. 00:26:00.360 --> 00:26:01.480 So how does that change things? 00:26:01.480 --> 00:26:01.640 Right. 00:26:01.640 --> 00:26:07.360 So that means like if you mess up your virtual environment and have to delete it and recreate it, it doesn't, it doesn't cost you anything. 00:26:07.360 --> 00:26:09.060 It's totally, it's totally ephemeral. 00:26:09.060 --> 00:26:11.020 You can throw it away and recreate it at any time. 00:26:11.020 --> 00:26:18.640 And like, that's just like a different, like, it sounds small, but that's kind of like a different relationship to a virtual environments than before where it's like, okay, I created the virtual environment now. 00:26:18.640 --> 00:26:19.740 Oh no, my virtual environment. 00:26:19.740 --> 00:26:20.260 So I messed up. 00:26:20.260 --> 00:26:21.000 I got to recreate it. 00:26:21.000 --> 00:26:25.320 Like we want to just change a lot of the dynamics around some of those like abstractions. 00:26:25.320 --> 00:26:32.600 And then the other thing is this performance budget or this, if you think of performance, like a budget, this buys us a lot of room. 00:26:33.000 --> 00:26:41.320 And so like the fact that we are so fast on some of these things means that we can do things that otherwise might be like prohibitively slow. 00:26:41.320 --> 00:26:44.100 And this is more sort of forward looking and for the future. 00:26:44.100 --> 00:27:01.900 Like one thing that we're thinking about a lot right now is this idea of like resolving Python dependencies for many platforms because it's a little bit complicated, but you know, pip compile and pip and uv, they're really designed to only work on like the current Python platform and the current Python that you're using. 00:27:01.900 --> 00:27:22.240 So if you're on like macOS and you run pip compile, it will give you dependencies that are correct for like macOS and your current Python version, but they might be wrong on Windows because Python and Python, you can have dependencies that vary based on things like the current Python version or the current operating system. 00:27:23.240 --> 00:27:26.320 And so one example would be uvloop, right? 00:27:26.320 --> 00:27:27.560 For people trying to speed up. 00:27:27.560 --> 00:27:28.760 I think it doesn't work on Windows. 00:27:28.760 --> 00:27:30.580 It doesn't work at all on Windows. 00:27:30.580 --> 00:27:31.020 That's right. 00:27:31.020 --> 00:27:31.840 And so, yeah. 00:27:31.840 --> 00:27:32.940 So that's a very common one. 00:27:32.940 --> 00:27:33.620 Yeah. 00:27:34.360 --> 00:27:48.600 And it is, I think, very common that this is a very common thing, like on like lots of really, really popular packages have these kinds of markers, these what are called markers, which basically makes the dependencies conditional based on the current Python. 00:27:49.600 --> 00:28:05.040 And so, you know, like again, like in pip and pip-tools, uv right now just locks for the current Python poetry and I think PDM, but poetry for sure does this very interesting thing where they actually try to lock. 00:28:05.040 --> 00:28:07.980 They try to resolve for like all possible platforms. 00:28:09.260 --> 00:28:17.920 And that's something, it's really convenient because it means that like you kind of do one lock and then any user anywhere can like use that and it works. 00:28:17.920 --> 00:28:24.460 And it's kind of like guaranteed to be right as opposed to maybe it doesn't work on Windows or whatever else. 00:28:24.460 --> 00:28:29.920 And so we're pretty interested in like that, adding that behavior and that's going to be like more expensive. 00:28:30.440 --> 00:28:41.160 And so if we have something that's really fast right now, it kind of like opens up a design space of interesting things we can do to solve that problem while still being, you know, very fast. 00:28:41.160 --> 00:28:44.160 So that's kind of how I think about performance. 00:28:44.160 --> 00:28:50.880 Like I think it's, I think people like a little bit underrate how much of an impact it can have on even small interactions. 00:28:50.880 --> 00:28:55.360 Because once you try something that's much faster, it's often hard to go back to the other thing. 00:28:55.360 --> 00:28:59.340 But I'm glad you've had that experience at least. 00:28:59.340 --> 00:29:00.220 Yeah, absolutely. 00:29:00.580 --> 00:29:02.180 A couple of interesting comments. 00:29:02.180 --> 00:29:07.980 Danny says, if it takes 10 seconds, I go to triage that issue on GitHub and come back five minutes later, right? 00:29:07.980 --> 00:29:10.360 Like that's kind of where I was getting at with the flow stuff. 00:29:10.360 --> 00:29:11.320 It's like, okay, I'm ready. 00:29:11.320 --> 00:29:15.000 Mind wonders and you're kind of like starting over. 00:29:15.000 --> 00:29:16.160 Yeah. 00:29:16.160 --> 00:29:19.360 And then Henry out in the audience also says, hey, Henry. 00:29:19.360 --> 00:29:29.440 Also, this doing things so quickly, it encourages good environment usage because if doing things the right way is just as easy and as fast as doing it the wrong way, you might as well do it the right way. 00:29:29.440 --> 00:29:30.120 That's a nice thought. 00:29:30.120 --> 00:29:30.940 That's a nice thought. 00:29:30.940 --> 00:29:31.120 Yeah. 00:29:31.120 --> 00:29:31.620 Yeah. 00:29:31.620 --> 00:29:32.220 It is. 00:29:32.280 --> 00:29:32.720 Yeah. 00:29:32.720 --> 00:29:39.320 I think there are certain things though where we aren't any faster. 00:29:39.900 --> 00:29:44.360 And I'm interested in thinking about those and what we can do about them. 00:29:44.360 --> 00:29:53.600 The most obvious one is in Python, sometimes when you install a dependency, you might have to build it from source. 00:29:53.600 --> 00:30:00.820 And in Python, you often are installing things like NumPy, which includes some native code. 00:30:01.360 --> 00:30:05.600 And so actually turning that into something that your system can run is kind of expensive. 00:30:05.600 --> 00:30:14.620 And you typically don't have to think about this with NumPy because they ship what are called wheels, which are kind of like these pre-compiled artifacts. 00:30:16.000 --> 00:30:19.180 But not all packages ship wheels. 00:30:19.180 --> 00:30:21.640 You could be on a platform that doesn't have a wheel. 00:30:21.640 --> 00:30:33.000 For a variety of reasons, you might get the raw NumPy source code and then have to compile it down into a build distribution. 00:30:33.740 --> 00:30:41.460 And for us, if we get into that situation, it's not going to be any faster because we still have to build the thing from source. 00:30:41.460 --> 00:30:41.800 Yeah. 00:30:41.800 --> 00:30:42.500 Just like this. 00:30:42.500 --> 00:30:44.320 That's the slowest part of the whole process. 00:30:44.320 --> 00:30:44.660 Yeah. 00:30:44.660 --> 00:30:51.920 And so there are some cases and some bottlenecks where I'm kind of thinking about how could we do even better there. 00:30:51.920 --> 00:30:55.340 And there are actually things that are changing in the standards that have made this a little easier. 00:30:55.340 --> 00:30:58.280 There's this new thing called metadata 2.2. 00:30:58.280 --> 00:30:59.180 Sounds cool. 00:30:59.180 --> 00:30:59.680 I know. 00:31:00.460 --> 00:31:03.920 It basically lets you read the metadata. 00:31:03.920 --> 00:31:05.400 You thought you liked metadata 2.1. 00:31:05.400 --> 00:31:05.760 Yeah. 00:31:05.760 --> 00:31:06.860 Yeah. 00:31:06.860 --> 00:31:09.340 We're actually on 2.3 now. 00:31:09.340 --> 00:31:11.600 I'm just kidding. 00:31:11.600 --> 00:31:18.880 But the point of that is it's a standard that makes it so you don't have to build the thing as often. 00:31:18.880 --> 00:31:24.280 If you just want to know the metadata of NumPy, for example, you may no longer need to build it. 00:31:24.280 --> 00:31:25.160 And before you did. 00:31:25.160 --> 00:31:27.920 And so the standards are moving in a helpful direction here. 00:31:28.040 --> 00:31:33.320 But I just thought I'd call it out because like the source distribution, there are cases where it won't be faster. 00:31:33.320 --> 00:31:35.420 And I'm pretty interested in those. 00:31:35.420 --> 00:31:37.300 But people should know that they also exist. 00:31:37.300 --> 00:31:40.520 There's also, it's not just time. 00:31:40.520 --> 00:31:42.980 There's also security issues around that, right? 00:31:42.980 --> 00:31:50.940 If things are still based on building source from source and then running the setup pie to see what the metadata is. 00:31:50.940 --> 00:31:57.520 It's both slow and also potentially running arbitrary code just to install a thing, which can make you, I mean, I don't know. 00:31:57.520 --> 00:32:01.860 Rarely has anything gone wrong with just installing running arbitrary code off the internet. 00:32:01.860 --> 00:32:03.700 But sometimes it could go wrong. 00:32:03.700 --> 00:32:04.280 Yeah. 00:32:04.280 --> 00:32:11.420 It would be nice to get to a world where that's not required or at least for dependency resolution. 00:32:11.420 --> 00:32:13.660 And like the standards are moving in that direction. 00:32:13.660 --> 00:32:14.540 Yeah, exactly. 00:32:14.540 --> 00:32:16.560 I mean, you might, if you're going to use it, you're going to have to get it. 00:32:16.560 --> 00:32:20.320 But if you just want to know, well, what versions do I need to put these two things together? 00:32:20.320 --> 00:32:21.360 You shouldn't have to do that. 00:32:21.360 --> 00:32:21.900 Yeah, exactly. 00:32:22.080 --> 00:32:28.680 Because today for NumPy, for example, if you didn't have a wheel, you would have to run some arbitrary code to ask it for its dependencies. 00:32:28.680 --> 00:32:29.760 Yeah, exactly. 00:32:29.760 --> 00:32:32.740 Well, let's talk about some of the usages here. 00:32:32.740 --> 00:32:35.020 Let's see some more interesting comments in the chat. 00:32:35.020 --> 00:32:35.920 I'll get back to them. 00:32:35.920 --> 00:32:40.640 First of all, you know, is it like Highlander the movie? 00:32:40.640 --> 00:32:42.220 Should there just be one? 00:32:42.220 --> 00:32:47.360 Is UPy a thing that I put into, say, my requirements file if I want to use it? 00:32:47.360 --> 00:32:49.840 Or do I install it just once for my machine? 00:32:49.840 --> 00:32:51.660 Like, what's the scale? 00:32:51.940 --> 00:32:52.520 UV. 00:32:52.520 --> 00:32:54.240 So there's a lot of different ways. 00:32:54.240 --> 00:32:56.080 You can, there's a lot of different ways to install it. 00:32:56.080 --> 00:32:58.280 You definitely can install it into a virtual environment. 00:32:58.280 --> 00:33:03.040 My general recommendation would be to install it on your machine once. 00:33:03.040 --> 00:33:09.780 Because an interesting thing about uv is you don't need Python to install it. 00:33:09.780 --> 00:33:13.080 And it doesn't have, you need to have Python on your machine. 00:33:13.080 --> 00:33:19.180 Because we need to be able to, for example, build source distributions or know where to install the thing. 00:33:19.180 --> 00:33:19.500 Right? 00:33:19.540 --> 00:33:22.000 We have to install into a virtual environment somewhere. 00:33:22.000 --> 00:33:26.180 And so you have to have Python installed on your machine. 00:33:26.180 --> 00:33:27.780 But uv does not depend on Python. 00:33:27.780 --> 00:33:33.840 So like it can install into arbitrary virtual environments or into arbitrary other Python interpreters. 00:33:33.840 --> 00:33:41.800 So that's why we generally recommend using like these standalone installers that we have at the top of this snippet on the screen. 00:33:41.800 --> 00:33:45.080 They're like the curl and the PowerShell invocations. 00:33:45.520 --> 00:33:50.260 Those will install a single binary on your machine that's uv. 00:33:52.160 --> 00:34:00.120 And from there, you can use uv to install into virtual environments that it creates, to install into your system Python, to do whatever. 00:34:00.840 --> 00:34:11.040 So like I would say most people, based on the statistics, based on our download statistics, most people still install uv with pip, which makes sense, which is not that surprising to me. 00:34:11.040 --> 00:34:15.880 But I would generally recommend having like one uv install on your machine. 00:34:15.880 --> 00:34:19.260 Although if you do install into a virtual environment, like that's also totally fine. 00:34:19.360 --> 00:34:20.560 Like nothing's going to go wrong. 00:34:20.560 --> 00:34:26.060 It's just not completely necessary because a single uv can already install into that virtual environment. 00:34:26.060 --> 00:34:26.720 Right. 00:34:26.720 --> 00:34:29.040 You're never going to import uv. 00:34:29.040 --> 00:34:29.580 No. 00:34:29.580 --> 00:34:32.700 I mean, you might, but in general, people are not going to. 00:34:32.700 --> 00:34:34.600 I think you can. 00:34:34.600 --> 00:34:36.580 I'm trying to think what would happen. 00:34:36.580 --> 00:34:42.220 It's not like things like pie tests and other things. 00:34:42.220 --> 00:34:48.880 A lot of times there could be just one of those, but then you want to use fixtures and you need to import something out. 00:34:48.880 --> 00:34:52.380 So then it really needs to be in the virtual environment and accessible like a library. 00:34:52.380 --> 00:34:57.300 But this is just something you run maybe even before your codes, before your virtual environment exists. 00:34:57.300 --> 00:34:57.580 Right. 00:34:57.580 --> 00:35:01.580 Like it is, it really kind of does ideally exist outside of it, I think. 00:35:01.580 --> 00:35:01.960 Yeah. 00:35:01.960 --> 00:35:03.140 Ideally it exists outside of it. 00:35:03.140 --> 00:35:07.620 And there are some cases where you'd want to install it within the virtual environment, but ideally it exists outside of it. 00:35:07.620 --> 00:35:15.820 The other thing that's nice about using the standalone installers, like at the top here, is soon we're going to ship like a self-update command. 00:35:16.120 --> 00:35:20.400 So you can do like uv self-update and it will update uv to the latest version. 00:35:20.920 --> 00:35:25.640 And we can't really support that if you install it with a different package manager, like pip or brew. 00:35:25.640 --> 00:35:27.760 Those have their own upgrade commands, obviously. 00:35:27.760 --> 00:35:35.260 But like if you install it with our installer, then like we can control the upgrades and stuff like that and provide some other features. 00:35:36.180 --> 00:35:41.300 So that's generally what we recommend, but obviously we're always going to support like installing with pip and stuff too. 00:35:41.300 --> 00:35:41.700 Yeah. 00:35:41.700 --> 00:35:44.520 I'm all about pipX these days. 00:35:44.520 --> 00:35:45.520 I really like... 00:35:45.520 --> 00:35:46.200 Yeah, we're pipX. 00:35:46.200 --> 00:35:46.540 That's mine. 00:35:46.540 --> 00:35:48.020 Yeah, yeah. 00:35:48.020 --> 00:35:53.900 PipX really, you know, you can just say, kind of like brew, you can say upgrade all my Python CLI tools. 00:35:53.900 --> 00:35:55.360 PipX upgrade all. 00:35:55.580 --> 00:35:55.980 Yeah, yeah. 00:35:55.980 --> 00:35:57.420 Yeah, yeah, yeah. 00:35:57.420 --> 00:36:02.900 I find rough and uv is just, you guys are moving super fast and there's frequently an update. 00:36:02.900 --> 00:36:04.760 There are often updates, yeah. 00:36:04.760 --> 00:36:05.880 Yeah. 00:36:06.580 --> 00:36:18.480 Yeah, and then, you know, the uv interface, like it's meant to be pretty familiar to, or it's meant to be, I guess, hopefully very familiar to people who have used, who are already using pip and other tools. 00:36:18.480 --> 00:36:22.220 And so like UVVM creates a virtual environment. 00:36:22.220 --> 00:36:27.180 And then there's, we have this sort of pip subcommand. 00:36:27.180 --> 00:36:29.840 So like uv pip install flask. 00:36:29.840 --> 00:36:33.300 And that doesn't call or use pip in any way. 00:36:33.300 --> 00:36:37.180 We're just using pip here to convey what the interface looks like. 00:36:37.180 --> 00:36:49.500 And part of the motivation there is at some point in the future, we probably want to add a new interface to uv that's a little bit more high level. 00:36:49.500 --> 00:36:55.380 So you could think of something like Poetry, where they have like Poetry install. 00:36:55.380 --> 00:37:00.200 You know, they have this sort of higher level interface for interfacing with packages. 00:37:00.200 --> 00:37:02.380 We want to do something like that. 00:37:03.060 --> 00:37:13.880 And so that's why we left the top level of the interface clear, because we might kind of like integrate and, sorry, innovate and like ship a bunch of stuff that would otherwise break this interface. 00:37:13.880 --> 00:37:23.780 So like, by putting this stuff under pip, we basically created this like isolated space where we can make sure that those commands like keep working no matter what we choose to change in the future. 00:37:25.000 --> 00:37:27.620 And it's very similar to the pip's interface, right? 00:37:27.620 --> 00:37:29.580 It's like pip install, then it could be a package name. 00:37:29.580 --> 00:37:31.960 It could be a requirement, -r requirements file. 00:37:31.960 --> 00:37:33.660 We support editable installs. 00:37:33.660 --> 00:37:38.760 We support like URL dependencies, get dependencies, all that kind of stuff. 00:37:38.760 --> 00:37:39.460 Right. 00:37:39.460 --> 00:37:44.060 All the stuff you might normally do pip install something for the most part, that's the same. 00:37:44.060 --> 00:37:44.520 Yep. 00:37:44.520 --> 00:37:45.160 Yeah, exactly. 00:37:45.500 --> 00:37:48.320 And we support a lot of the same flags, too, that like pip supports. 00:37:48.320 --> 00:37:54.700 So like --index URL, --extra index URL, like --no binary, like all that stuff. 00:37:54.700 --> 00:38:00.260 Like we've put a lot of, we don't support every flag, but we've put a lot of effort into supporting a lot of them. 00:38:00.260 --> 00:38:01.500 And we've added more over time. 00:38:01.680 --> 00:38:08.800 Like now we support like --no build isolation and stuff like that, that maybe most listeners have never even had to look at. 00:38:08.800 --> 00:38:11.700 But like, you know, these things matter in some circumstances. 00:38:11.700 --> 00:38:15.140 And so like over time, our goal is to add like more of the interface. 00:38:15.140 --> 00:38:22.500 But, you know, I would say for common use cases, I would expect that it just works. 00:38:22.500 --> 00:38:25.620 Like just adding uv at front, like it should just work. 00:38:25.620 --> 00:38:28.900 Now, as you become more complex, it won't. 00:38:29.180 --> 00:38:33.260 And you'll run into some, we have a whole document of like subtle ways that we deviate. 00:38:33.260 --> 00:38:34.800 Some are intentional. 00:38:34.800 --> 00:38:36.320 Some I consider bugs, right? 00:38:36.320 --> 00:38:37.780 Like some we want to fix over time. 00:38:37.780 --> 00:38:41.180 But the intention is that for most people, we want it to just work. 00:38:41.180 --> 00:38:44.460 Once we fix the first round of bugs following the release, of course. 00:38:44.460 --> 00:38:45.260 Yeah, yeah, sure. 00:38:45.260 --> 00:38:50.840 So you've got a uv venv VIMP, for creating a virtual environment. 00:38:50.840 --> 00:38:56.740 And it has a lot of the flags that Python dash MVNV would have, where you can say, you know, 00:38:57.080 --> 00:39:02.080 bring along pip or upgrade things or set the prompt name, that kind of stuff, right? 00:39:02.080 --> 00:39:03.160 Yeah, exactly. 00:39:03.160 --> 00:39:03.720 Yep. 00:39:03.720 --> 00:39:06.460 Similar for uv, pip, and then stuff. 00:39:06.460 --> 00:39:11.820 What I find I do for most of my work, at least if I'm not trying to teach or do a presentation, 00:39:11.820 --> 00:39:13.780 where people are like, what did you just do? 00:39:13.780 --> 00:39:15.560 These all have kind of aliases, right? 00:39:15.820 --> 00:39:21.180 Like uv, pip, install, dash, R, requirements.txt is just PIR, pip install requirements. 00:39:21.180 --> 00:39:22.480 You know, things like that, right? 00:39:22.480 --> 00:39:23.180 Yeah, yeah, yeah. 00:39:23.280 --> 00:39:25.940 And so try to just take the common stuff. 00:39:25.940 --> 00:39:33.780 And it doesn't matter if it's pip or uv really backing it from my CLI ergonomics, right? 00:39:33.780 --> 00:39:35.600 I'm just, I use my alias. 00:39:35.600 --> 00:39:38.400 And if I want to change that, I just go and edit it. 00:39:38.400 --> 00:39:43.360 And so for me, it was super easy to switch to adopt these things because I'm like, I edit 00:39:43.360 --> 00:39:49.920 my ZSH RC file once and now everything just, I just do the same stuff and it works, but 00:39:49.920 --> 00:39:50.880 the new way, way faster. 00:39:50.880 --> 00:39:51.440 Yeah. 00:39:51.440 --> 00:39:57.000 And, you know, the goal of like pip compatibility has been interesting because there are actually 00:39:57.000 --> 00:40:00.100 like some things that we very intentionally want to do differently. 00:40:00.100 --> 00:40:05.660 And like, I won't speak for the pip maintainers, but I think in some cases, there are probably 00:40:05.660 --> 00:40:08.160 things that they would do differently too, if they could. 00:40:08.160 --> 00:40:09.140 I think so too, yeah. 00:40:09.720 --> 00:40:16.080 Yeah, because like pip, I mean, pip is extremely important and its number one goal is like, 00:40:16.080 --> 00:40:19.200 it needs to be like robust, right? 00:40:19.200 --> 00:40:22.200 It needs to keep working and it needs to be compatible. 00:40:22.200 --> 00:40:27.660 And because it's like, it's truly like the cornerstone of like a lot of the Python ecosystem. 00:40:27.660 --> 00:40:32.500 And so pip, it's very, it's harder for them to change things, especially in a breaking 00:40:32.500 --> 00:40:32.800 way. 00:40:32.800 --> 00:40:36.400 And they have to be really thoughtful and do that over a long period of time. 00:40:36.600 --> 00:40:41.120 And so we're in kind of a privileged position, honestly, because we come in with a new tool. 00:40:41.120 --> 00:40:44.340 We can actually do a bunch of, choose to do things differently. 00:40:44.340 --> 00:40:48.920 And some people will get upset about it, obviously, because they're like, this isn't exactly how 00:40:48.920 --> 00:40:49.500 PIP does it. 00:40:49.500 --> 00:40:53.120 I'm trying to hold firm on some of those things and also be open-minded on others. 00:40:53.120 --> 00:40:53.420 Right. 00:40:53.420 --> 00:40:58.700 But like, you know, like one thing that we do that's different, that's, I guess maybe it's 00:40:58.700 --> 00:41:01.880 not evident from this exact, from this screen exactly, or I guess it sort of is. 00:41:01.880 --> 00:41:05.400 So when you do UVVM, we default to the name .vm. 00:41:05.400 --> 00:41:08.400 So like, if you don't provide a name, we just default to .vm. 00:41:08.400 --> 00:41:13.100 And there was a PEP to make that a standard that got pulled back. 00:41:13.100 --> 00:41:16.220 But like, it's not like we're trying to make it a standard, but I just think it's a good 00:41:16.220 --> 00:41:16.600 default. 00:41:16.600 --> 00:41:18.880 Like we want to abstract away some of this stuff. 00:41:18.880 --> 00:41:21.900 And so you can just like have a virtual, you can obviously pass a name if you want, 00:41:21.980 --> 00:41:23.060 but we default to .vm. 00:41:23.060 --> 00:41:24.400 That's different from other tools. 00:41:24.400 --> 00:41:29.740 And then, you know, the other thing is when you do UVPIP install, by default, we require 00:41:29.740 --> 00:41:30.620 a virtual environment. 00:41:30.620 --> 00:41:37.060 And so if you do UVPIP install Flask and there's no virtual environment, like we'll throw, we'll 00:41:37.060 --> 00:41:37.380 error. 00:41:37.380 --> 00:41:39.440 And it doesn't have to be active. 00:41:39.440 --> 00:41:42.660 We'll look for it in the current directory, even if it's not active. 00:41:42.660 --> 00:41:46.580 So like if you do UVVM, then UVPIP install Flask, like it will work. 00:41:46.580 --> 00:41:50.040 And in my opinion, it will do the intuitive thing, which is it will install into the environment 00:41:50.040 --> 00:41:51.160 in the current directory. 00:41:51.340 --> 00:41:53.000 So it's like, it's trying to provide... 00:41:53.000 --> 00:41:53.360 It will traverse up. 00:41:53.360 --> 00:41:58.080 Like if I'm a little farther down, it'll go up and say, oh, just, oh, that's amazing. 00:41:58.080 --> 00:41:58.440 Exactly. 00:41:58.440 --> 00:42:01.620 So it'll find the virtual environment in the current directory or any parent directory. 00:42:01.620 --> 00:42:10.240 So like for me, that's just like a nice default workflow that encourages virtual environments 00:42:10.240 --> 00:42:11.800 and like doesn't get in your way about it. 00:42:11.800 --> 00:42:15.940 And we do have a flag for installing outside virtual environments, but the difference is in 00:42:15.940 --> 00:42:20.640 UV, you have to opt in if you want to do something like install into the system Python. 00:42:20.700 --> 00:42:26.660 It's a very virtual environment first while abstracting away a lot of the details of virtual environments. 00:42:26.660 --> 00:42:30.380 So there are things like that where we've made intentional decisions. 00:42:30.380 --> 00:42:31.560 In my opinion. 00:42:31.560 --> 00:42:31.900 Yeah. 00:42:31.900 --> 00:42:32.580 Yeah. 00:42:32.580 --> 00:42:40.240 And I think, and it's nice because we can start with that behavior and we're in a privileged 00:42:40.240 --> 00:42:42.320 position because we can start with different defaults. 00:42:42.320 --> 00:42:43.320 Yeah, for sure. 00:42:43.320 --> 00:42:53.700 So if I do, excuse me, if I do a uv, VNV for a virtual environment, then maybe I uv pip install 00:42:53.700 --> 00:42:58.360 something, but then I forget and I pip install something else. 00:42:58.360 --> 00:42:59.200 Does it break? 00:42:59.200 --> 00:42:59.840 Is it okay? 00:42:59.840 --> 00:43:02.100 As in the third time there's no UV? 00:43:02.460 --> 00:43:03.160 Well, yeah, exactly. 00:43:03.160 --> 00:43:07.420 Like, so I'm in uv pip install, something uv pip install that. 00:43:07.420 --> 00:43:08.580 And I was, oh, I forgot this thing. 00:43:08.580 --> 00:43:09.420 It was pip install. 00:43:09.420 --> 00:43:11.140 Oh, I forgot the uv. 00:43:11.140 --> 00:43:11.960 To this day. 00:43:11.960 --> 00:43:13.000 No, it shouldn't. 00:43:13.000 --> 00:43:13.720 It shouldn't break. 00:43:13.720 --> 00:43:14.040 Assuming. 00:43:14.040 --> 00:43:16.120 So let's say you activate the virtual environment. 00:43:16.120 --> 00:43:16.660 Yeah. 00:43:16.660 --> 00:43:16.840 Yeah. 00:43:16.840 --> 00:43:17.400 It's already activated. 00:43:17.400 --> 00:43:17.920 Okay, cool. 00:43:18.000 --> 00:43:22.380 So you do uv VM, you activate the virtual environment, you do uv pip install flask, and then you 00:43:22.380 --> 00:43:24.440 do pip install black or whatever. 00:43:24.440 --> 00:43:24.980 Yeah. 00:43:24.980 --> 00:43:25.420 Something, yeah. 00:43:25.420 --> 00:43:30.880 That should work just fine because like the thing that we ultimately create in the virtual 00:43:30.880 --> 00:43:33.040 environment, like that's all based on standards. 00:43:33.040 --> 00:43:38.220 So like the stuff that we do in the virtual environment to install packages, it should look 00:43:38.220 --> 00:43:41.060 like roughly indistinguishable from what pip would do. 00:43:41.060 --> 00:43:42.940 And it's totally interoperable with pip. 00:43:43.300 --> 00:43:48.920 So like similarly, if you did pip install and then like pip install flask and then uv pip 00:43:48.920 --> 00:43:51.280 uninstall flask, like that would work correctly. 00:43:51.280 --> 00:43:55.680 Like pip would install flask and we would remove it because everything that pip and uv are doing 00:43:55.680 --> 00:43:58.100 in the virtual environment is based on standards. 00:43:58.100 --> 00:44:03.200 So the way that we add and remove packages, all that kind of stuff, it should be totally interoperable 00:44:03.200 --> 00:44:03.920 with other tools. 00:44:03.920 --> 00:44:04.860 Okay, cool. 00:44:04.860 --> 00:44:11.540 PyPI is pretty good at having analytics around what packages are being downloaded, what's popular, 00:44:11.540 --> 00:44:12.840 what platforms. 00:44:12.840 --> 00:44:18.100 Do you do anything like send a user agent or something so that people could answer the 00:44:18.100 --> 00:44:18.400 question? 00:44:18.400 --> 00:44:23.320 Or how many people are using straight pip and how many people are using uv in a year? 00:44:23.320 --> 00:44:29.000 Yeah, we sent, we set a uv user agent and it's right now it's just uv in the uv version. 00:44:29.000 --> 00:44:37.580 There's an open issue in the repo from Pradyan who works on pip that outlines all this stuff 00:44:37.580 --> 00:44:43.280 that pip includes in its own user agent, which we want to add, which is pretty interesting. 00:44:43.280 --> 00:44:44.120 I actually didn't know this. 00:44:44.120 --> 00:44:49.700 So pip in its user agent includes a lot of things about the current Python platform. 00:44:49.700 --> 00:44:55.640 Data that they use in PyPI basically to inform decisions. 00:44:56.260 --> 00:44:58.100 Like, oh, no one's using Python. 00:44:58.100 --> 00:44:58.100 Right. 00:44:58.100 --> 00:45:01.080 This is just, I don't think this is actually how it's used, but just as an example. 00:45:01.080 --> 00:45:03.680 Oh, no one's using Python 3.7 anymore. 00:45:03.680 --> 00:45:08.120 Let's lower priority on that because no one's installing Python with Python 3.7. 00:45:08.120 --> 00:45:11.360 And so we want to add, we don't have that right now. 00:45:11.360 --> 00:45:12.400 We're planning to add it. 00:45:13.740 --> 00:45:16.260 But yeah, it will be interesting to see. 00:45:16.260 --> 00:45:21.160 We can get statistics over time too on like uv and pip and how common they are in different 00:45:21.160 --> 00:45:21.560 dimensions. 00:45:21.560 --> 00:45:22.200 Yeah. 00:45:22.200 --> 00:45:26.880 You know, one thing that would be kind of fun to have is, I'm not sure if it's at all 00:45:26.880 --> 00:45:33.160 interesting to you, but to have some kind of cron job or background thing, even if you 00:45:33.160 --> 00:45:37.040 just interact with uv, if it kind of kicked off sort of a background process that did its 00:45:37.040 --> 00:45:43.040 thing, but, you know, take the history of maybe the top 500 packages plus the ones you've 00:45:43.040 --> 00:45:46.620 used and just kind of keep a cache up to date on your system. 00:45:46.620 --> 00:45:51.320 So if you're on an airplane or you're in a coffee shop or for some reason, the internet's 00:45:51.320 --> 00:45:51.900 not great. 00:45:51.900 --> 00:45:55.760 Like it's kind of pre-filled the cache and everything's there. 00:45:55.760 --> 00:45:55.960 Yeah. 00:45:55.960 --> 00:45:58.960 And that kind of, that might be fun. 00:45:58.960 --> 00:45:59.600 Yeah. 00:45:59.600 --> 00:46:05.280 We do have a dash track offline flag, which will like, it forbids network requests, obviously, 00:46:05.580 --> 00:46:09.740 but it, so like it'll, it'll try and do the full install without accessing the network. 00:46:09.740 --> 00:46:10.200 Yeah. 00:46:10.200 --> 00:46:10.500 That's cool. 00:46:10.500 --> 00:46:12.160 It would be cool to pre-fill the cache. 00:46:12.160 --> 00:46:12.740 Yeah. 00:46:12.740 --> 00:46:16.600 But it's just like, you know, if you're a uv user and you sit down on your computer and 00:46:16.600 --> 00:46:19.840 for some reason the internet's out, I'm at a conference, I'm doing a talk and it's got 00:46:19.840 --> 00:46:21.620 the dreaded, terrible conference wifi. 00:46:21.620 --> 00:46:22.820 It's not the end of your demo. 00:46:22.820 --> 00:46:24.260 It's just like, yeah, see how fast this is? 00:46:24.260 --> 00:46:24.760 It's amazing. 00:46:24.760 --> 00:46:25.580 Yeah. 00:46:25.580 --> 00:46:26.000 Yeah. 00:46:26.000 --> 00:46:26.120 Yeah. 00:46:26.120 --> 00:46:26.820 Yeah. 00:46:26.820 --> 00:46:30.860 It was actually, it's actually funny because like the benchmarking, like when we do benchmarks, 00:46:30.860 --> 00:46:36.560 benchmarks can just vary so dramatically based on like internet connectivity, right? 00:46:36.560 --> 00:46:38.500 Like that's often, that is often the bottle. 00:46:38.500 --> 00:46:43.120 I mean, it's not, it's clearly not always the bottleneck because we're able to be faster 00:46:43.120 --> 00:46:43.900 than other tools. 00:46:43.900 --> 00:46:48.560 So like if it was always the bottleneck, I don't think that wouldn't really be true, but, but, 00:46:48.560 --> 00:46:49.340 but it matters a lot. 00:46:49.340 --> 00:46:49.380 Yeah. 00:46:49.380 --> 00:46:54.040 And so sometimes I'll throw on, on macOS, they have a tool called network link conditioner. 00:46:54.040 --> 00:46:56.500 So you can actually force. 00:46:56.500 --> 00:46:57.900 Like drag your network down. 00:46:57.900 --> 00:46:58.580 Yeah. 00:46:58.580 --> 00:46:59.820 You can force it down. 00:46:59.820 --> 00:47:04.880 And the categories are like, it's not literally called like bad mobile phone or something, 00:47:04.880 --> 00:47:06.620 but it's, that's like basically what it is. 00:47:06.620 --> 00:47:06.760 Yeah. 00:47:06.760 --> 00:47:09.340 It's like, it's like, it's like very lossy edge network. 00:47:09.340 --> 00:47:11.180 Like those are like the pre-built category. 00:47:11.180 --> 00:47:14.960 It's pretty funny, but like it is actually, that's a very helpful thing for benchmarking. 00:47:14.960 --> 00:47:15.320 Yeah. 00:47:15.320 --> 00:47:18.740 And sometimes actually, if you have a really good network connection, sometimes it's actually helpful 00:47:18.740 --> 00:47:25.540 to, to set that anyway, to a pretty good level, because then at least it'll be like very consistent 00:47:25.540 --> 00:47:27.320 across your executions. 00:47:27.320 --> 00:47:28.760 Anyway, it's something I've learned over time. 00:47:28.760 --> 00:47:29.380 Yeah. 00:47:29.380 --> 00:47:31.380 That is super interesting. 00:47:31.380 --> 00:47:36.960 So one of the things I think it's interesting to consider here is obviously this is way faster. 00:47:36.960 --> 00:47:43.260 I saw some of the announcements and some of the discussion around the announcements and the 00:47:43.260 --> 00:47:45.140 people were really surprised how fast this is. 00:47:45.140 --> 00:47:48.380 And I think, I don't think Rust is the full answer. 00:47:48.540 --> 00:47:52.860 I mean, you tell me if I'm wrong, but like one part, obviously making the code run ultra 00:47:52.860 --> 00:47:55.020 fast, that is an important part of it. 00:47:55.020 --> 00:48:00.640 But it also seems like you all have rethought some of the internal algorithms and some of 00:48:00.640 --> 00:48:04.800 the caching and some of the ways things work, maybe with a fresh take on it, or you're not 00:48:04.800 --> 00:48:07.440 constrained by the way it's been done for many, many years. 00:48:07.600 --> 00:48:11.900 So what elements are at play to make it as fast as it is? 00:48:11.900 --> 00:48:13.700 Is it just Rust or is it something else? 00:48:13.700 --> 00:48:20.900 So Rust is important, but the way I typically talk about frame it is that Rust is kind of like an 00:48:20.900 --> 00:48:23.320 enabler to writing really fast programs. 00:48:23.780 --> 00:48:32.980 It forces you to slash lets you care about certain things that matter a lot for performance that like 00:48:32.980 --> 00:48:37.120 in Python, there's actually just no way to care about them. 00:48:37.120 --> 00:48:39.860 I'm talking specifically here about memory allocation. 00:48:39.860 --> 00:48:42.200 Like in Python, that all just kind of happens. 00:48:42.200 --> 00:48:45.940 In Rust, you're forced to think about, am I going to allocate memory here or not? 00:48:46.000 --> 00:48:48.920 And like, when is it going to be deallocated and all that kind of stuff? 00:48:48.920 --> 00:48:54.700 And so that unlocks the ability to care and like be really careful in how you manage it. 00:48:54.700 --> 00:48:58.800 And so Rust is like a really important, Rust is an important part of the success story here. 00:48:58.800 --> 00:49:05.240 But I think it's like really, I think it would be really like very incorrect to say that it's 00:49:05.240 --> 00:49:08.380 like all about Rust and it's not a Rust versus Python thing. 00:49:08.380 --> 00:49:09.000 Right. 00:49:09.000 --> 00:49:11.600 Like if we could just recompile pip with. 00:49:11.600 --> 00:49:13.500 Yeah. 00:49:13.500 --> 00:49:14.140 So if you took. 00:49:14.140 --> 00:49:14.640 If you took. 00:49:14.640 --> 00:49:15.540 Cython or something. 00:49:15.540 --> 00:49:15.860 Boom. 00:49:15.860 --> 00:49:16.800 It would just be. 00:49:16.800 --> 00:49:18.980 No, it wouldn't be as big of a deal. 00:49:18.980 --> 00:49:19.180 Right. 00:49:19.180 --> 00:49:24.060 If you took pip and just like rewrote it line by line as close as you could in Rust, like 00:49:24.060 --> 00:49:28.200 it would probably be faster than pip, but it would not be nearly as fast as uv. 00:49:28.200 --> 00:49:30.820 Like we just do a lot of things differently. 00:49:30.820 --> 00:49:35.740 And part of like, I guess evidence for this too, is like throughout the development of uv, 00:49:35.740 --> 00:49:41.400 there were like multiple pull requests that sped up like uv as a whole by like 30 to 50%. 00:49:41.400 --> 00:49:42.480 So what does that say? 00:49:42.480 --> 00:49:45.540 That says that like, there's lots of different, those were all Rust programs. 00:49:45.540 --> 00:49:46.160 Right. 00:49:46.160 --> 00:49:51.280 So like we were able to write the same Rust program many times, optimizing it more and 00:49:51.280 --> 00:49:51.780 more and more. 00:49:51.780 --> 00:49:52.280 Right. 00:49:52.280 --> 00:49:52.620 Right. 00:49:52.620 --> 00:49:57.080 And so there's like so much engineering that went into making it fast. 00:49:57.080 --> 00:49:58.920 Some of it's like how the cache is designed. 00:49:58.920 --> 00:50:00.540 That was like a really important piece. 00:50:00.540 --> 00:50:03.680 And that's actually something that like pip could also do. 00:50:04.340 --> 00:50:10.420 And I think like it's possible a good outcome here actually is that like pip is able to take 00:50:10.420 --> 00:50:13.220 some of these changes and incorporate them over time. 00:50:13.220 --> 00:50:17.220 And we are kind of the guinea pig for that. 00:50:17.220 --> 00:50:21.660 Like we're going to kind of change user expectations a little bit about how the cache is set up. 00:50:21.660 --> 00:50:25.880 Like we're going to run into bugs about platform compatibility. 00:50:26.080 --> 00:50:28.960 Like we're going to run into bugs based on this cache design. 00:50:28.960 --> 00:50:29.200 Right. 00:50:29.200 --> 00:50:33.360 And like hopefully that could at some point help inform pip if they choose to redesign their 00:50:33.360 --> 00:50:34.020 cache this way. 00:50:34.020 --> 00:50:35.600 The cache design is really big thing. 00:50:35.600 --> 00:50:42.700 And then just a lot of like profiling the program, figuring out what the bottleneck is, and then 00:50:42.700 --> 00:50:47.140 like solving the bottleneck and just like really intense engineering work that was not me, but 00:50:47.140 --> 00:50:49.180 was like other people on the team to optimize some of that. 00:50:49.180 --> 00:50:55.260 Like a good example is it's like it sounds silly, but like we found that a bottleneck when 00:50:55.260 --> 00:50:57.580 all the data is cached and you're trying to do a resolution. 00:50:57.580 --> 00:50:59.020 So there's no network requests. 00:50:59.020 --> 00:51:00.140 Everything's local. 00:51:00.760 --> 00:51:06.760 One of the bottlenecks that we kept running into was parsing and comparing versions, version 00:51:06.760 --> 00:51:07.380 specifiers. 00:51:07.380 --> 00:51:13.580 So like in Python, when you have a requirements file, you have like FastAPI is greater than 00:51:13.580 --> 00:51:14.920 version one or whatever. 00:51:14.920 --> 00:51:19.780 And then in the FastAPI file, you have like FastAPI version 1.0.0, blah, blah, blah. 00:51:19.780 --> 00:51:24.660 So like those, like parsing those, and then we were comparing those. 00:51:24.660 --> 00:51:30.340 We were doing comparisons between version markers like so many times. 00:51:30.340 --> 00:51:32.740 Because it was just a constant thing in dependency resolution. 00:51:32.740 --> 00:51:34.080 It's like, can I use this dependency? 00:51:34.080 --> 00:51:35.520 Well, does it fit to this range? 00:51:35.520 --> 00:51:44.460 So Andrew Gallant from our team, like rewrote that parser and the version comparison specifier. 00:51:44.460 --> 00:51:47.340 And it's like, it's like one of the most incredible pull requests I've ever seen. 00:51:47.340 --> 00:51:50.560 Like it's like so, like every, every commitment is like so interesting. 00:51:50.900 --> 00:51:55.560 And he like optimized it so massively that it was like, he's basically representing the version 00:51:55.560 --> 00:52:02.640 in the end by like, like a 64 bit, like a U64 integer and comparisons is just like compare 00:52:02.640 --> 00:52:03.280 the two integers. 00:52:03.280 --> 00:52:05.740 And like, there was just so much engineering that went into it. 00:52:05.800 --> 00:52:07.560 And it sped up the cache case by like 30%. 00:52:07.560 --> 00:52:12.260 And so there's a lot of, there's just a lot of engineering that went into it. 00:52:12.260 --> 00:52:17.900 And, you know, again, it's harder for pip to do some of those things because they, the code 00:52:17.900 --> 00:52:22.260 bases, they have a lot, they have, they have actual users to support, right? 00:52:22.260 --> 00:52:24.860 Like we were able to do this before we launched publicly. 00:52:24.860 --> 00:52:30.040 Like we could break things, like we can do things, you know, we can make massive changes 00:52:30.040 --> 00:52:31.660 to the code base in short periods of time. 00:52:31.660 --> 00:52:35.420 So, you know, we're in kind of a privileged position to be able to really optimize the 00:52:35.420 --> 00:52:36.300 performance like that. 00:52:36.300 --> 00:52:41.100 But, you know, it's a big part culturally of what we wanted to do too, is like, we're going 00:52:41.100 --> 00:52:42.260 to make this thing extremely fast. 00:52:42.260 --> 00:52:43.140 Yeah. 00:52:43.140 --> 00:52:49.160 I'm seeing also interesting secondary effects, you know, talks and not, not, I think that's 00:52:49.160 --> 00:52:51.800 talks, talking about using it, making it faster. 00:52:52.140 --> 00:52:57.600 Just now here, Henry says build 1.1 is faster from looking at uv and the next version will 00:52:57.600 --> 00:53:00.980 also add uv as an installer choice, which is pretty impressive. 00:53:00.980 --> 00:53:01.680 Yeah. 00:53:01.680 --> 00:53:05.940 That's been a really cool, that's actually something I didn't really anticipate, but I'm 00:53:05.940 --> 00:53:11.100 really glad has happened, which is that there's a little bit of like, we ended up building 00:53:11.100 --> 00:53:12.040 something pretty modular. 00:53:12.040 --> 00:53:17.740 And so there's been a little bit of a, or a lot of like integrations with uv that people 00:53:17.740 --> 00:53:18.340 have been building. 00:53:18.340 --> 00:53:23.800 Like, like, Bernard had the, had the talks uv thing that was out like within a week, I 00:53:23.800 --> 00:53:28.280 think, which was like use uv to power talks in different ways. 00:53:28.280 --> 00:53:28.720 Yeah. 00:53:28.720 --> 00:53:28.720 Yeah. 00:53:28.720 --> 00:53:32.520 Knox, I think supports it bill pipe a slash build. 00:53:32.520 --> 00:53:39.520 You know, I've talked, I've talked to the hatch maintainer, you know, like he's pretty interested 00:53:39.520 --> 00:53:43.420 in making uv like an optional backend for hatch. 00:53:43.420 --> 00:53:47.880 So like that's really cool to me because it's all just like magnifying the impact. 00:53:48.740 --> 00:53:49.240 Yeah. 00:53:49.240 --> 00:53:55.200 And again, I don't know how much of that I anticipated, but it's really cool to see that 00:53:55.200 --> 00:53:58.680 like you could plug it into these other tools and they can just like go much faster. 00:53:58.680 --> 00:53:58.680 Yeah. 00:53:58.680 --> 00:53:59.680 So I don't know. 00:53:59.680 --> 00:54:04.100 I mean, that's been like, that's actually been like a big highlight from the release, honestly, 00:54:04.100 --> 00:54:08.200 has been the reception from like other maintainers of other tools who have been excited to integrate 00:54:08.200 --> 00:54:09.480 it in different ways. 00:54:09.480 --> 00:54:11.700 Cause it's just not, it just didn't happen. 00:54:11.700 --> 00:54:15.260 Like rough wasn't nearly as much of a fit for that kind of thing. 00:54:15.260 --> 00:54:19.740 Like there just aren't as nearly, there are some, but it's just not as natural to have like 00:54:19.740 --> 00:54:21.360 integrations with rough in that way. 00:54:21.360 --> 00:54:24.480 But for uv, it makes a lot of sense and it's been really cool to see. 00:54:24.480 --> 00:54:24.880 Yeah. 00:54:24.880 --> 00:54:29.880 It's more, uv is more of a building block than rough is rough fits in editors and CI and 00:54:29.880 --> 00:54:30.300 stuff. 00:54:30.300 --> 00:54:31.020 Yeah. 00:54:31.020 --> 00:54:31.280 Okay. 00:54:31.280 --> 00:54:31.540 Yeah. 00:54:31.540 --> 00:54:36.180 And there were, there were other tools, but yeah, it was, it's, it's mostly an end user thing 00:54:36.180 --> 00:54:37.820 and not a library, I think. 00:54:37.940 --> 00:54:38.200 All right. 00:54:38.200 --> 00:54:38.660 All right. 00:54:38.660 --> 00:54:40.900 So a couple of interesting things out in the audience. 00:54:40.900 --> 00:54:42.380 Let's, let's knock them out. 00:54:42.380 --> 00:54:42.620 Okay. 00:54:42.620 --> 00:54:49.900 So Tushar asks, and this is, I think a big tension that's building in the Python packaging space 00:54:49.900 --> 00:54:53.900 is packaging tools use Python to do stuff. 00:54:53.900 --> 00:54:56.520 What are the packaging tools control the Python? 00:54:56.520 --> 00:54:59.200 Who, who is controlling whom here? 00:54:59.200 --> 00:55:02.680 And so Tushar asks, will uv also install Python sometime soon? 00:55:02.680 --> 00:55:06.720 Can I just express, Hey, I want to use 312 on this and these dependencies. 00:55:07.420 --> 00:55:07.920 Yeah. 00:55:07.920 --> 00:55:10.580 We almost, we almost certainly want to do this. 00:55:10.580 --> 00:55:13.440 I don't think it will be, it won't be required. 00:55:13.440 --> 00:55:19.460 Like you can still use your own Python, but we almost certainly want to add Python bootstrapping. 00:55:19.460 --> 00:55:25.600 So like, that's another reason why the dream workflow is like you use the standalone uv installer, 00:55:25.600 --> 00:55:29.680 because then it's like, it will actually install Python for you too. 00:55:29.680 --> 00:55:36.580 And so like the reason, so then it's like, you don't have to have Python on your machine basically to start being productive with uv. 00:55:36.580 --> 00:55:38.640 You can just bootstrap the environment for you. 00:55:38.640 --> 00:55:50.020 And we can also do, you know, yeah, it would just be nice if you did, you know, pip compile, like uv pip compile, --Python 3.13 or something. 00:55:50.020 --> 00:55:53.920 And then we just like bootstrapped Python 3.13 and did the resolution on your machine. 00:55:53.920 --> 00:55:56.220 Like that's, that's kind of the workflow that we're trying to build towards. 00:55:56.220 --> 00:55:56.760 Yeah. 00:55:56.760 --> 00:56:02.180 Since you're so uv centric as well, it could be uv, V and V, you know, dash that version. 00:56:02.180 --> 00:56:02.540 Yeah. 00:56:02.540 --> 00:56:03.040 A hundred percent. 00:56:03.460 --> 00:56:04.680 And then we bootstrap it. 00:56:04.680 --> 00:56:05.000 Yeah. 00:56:05.000 --> 00:56:07.380 If you don't have, if you've got a cache, you just give it. 00:56:07.380 --> 00:56:09.360 And if you don't, then you, you get it on the machine. 00:56:09.360 --> 00:56:09.760 Exactly. 00:56:09.760 --> 00:56:10.500 Yeah. 00:56:10.500 --> 00:56:11.100 Yeah. 00:56:11.100 --> 00:56:12.700 So we'll definitely, we'll definitely do this. 00:56:12.700 --> 00:56:17.560 again, I do think it's important that it's not like required to use like the uv pythons. 00:56:17.560 --> 00:56:17.900 Yeah. 00:56:17.900 --> 00:56:25.440 we actually kind of had to do like a minimal version of this for, for our CI and testing. 00:56:25.440 --> 00:56:28.820 so, but it was a bash script. 00:56:28.820 --> 00:56:32.040 And then I think it became a Python script, but eventually it would be like built into uv. 00:56:32.040 --> 00:56:32.580 Yeah. 00:56:32.580 --> 00:56:32.940 Yeah. 00:56:33.240 --> 00:56:33.440 Yeah. 00:56:33.440 --> 00:56:34.560 I think that would be really. 00:56:34.560 --> 00:56:35.020 Yeah. 00:56:35.020 --> 00:56:36.320 I think it'll be super, really quite awesome. 00:56:36.320 --> 00:56:36.860 Yeah. 00:56:36.860 --> 00:56:37.760 I think it will be too. 00:56:37.760 --> 00:56:39.380 let's see. 00:56:39.380 --> 00:56:43.000 Tushar also says loves that the --require V and V. 00:56:43.000 --> 00:56:43.580 Yep. 00:56:43.580 --> 00:56:44.340 As a thing. 00:56:44.340 --> 00:56:49.280 he thinks it'll become the default and also points out for those who don't uv that 00:56:49.280 --> 00:56:53.560 pip require virtual uv as an environment flag is in your RC files. 00:56:53.560 --> 00:56:54.220 It's pretty nice. 00:56:54.220 --> 00:56:55.100 I need to do that. 00:56:55.100 --> 00:56:55.200 Yes. 00:56:55.200 --> 00:56:55.840 The other day. 00:56:55.840 --> 00:56:56.040 Yeah. 00:56:56.040 --> 00:56:56.920 You can, you can. 00:56:56.920 --> 00:57:00.540 Oh, I spelled that to the, to my, to the wrong spot. 00:57:00.540 --> 00:57:00.560 Yeah. 00:57:00.560 --> 00:57:03.180 So pip can be configured to require virtual ones. 00:57:03.180 --> 00:57:03.780 Yes. 00:57:03.780 --> 00:57:04.180 Yeah. 00:57:04.180 --> 00:57:05.180 I do something weird. 00:57:05.180 --> 00:57:09.320 I'm actually looking at my terminal behind our, our shared screen meeting here. 00:57:09.700 --> 00:57:12.340 And my prompt says global V and V. 00:57:12.340 --> 00:57:16.560 And so one of the things that I do is I have a, just, if I log it, if I open up my terminal, 00:57:16.560 --> 00:57:18.340 it already has an activated virtual environment. 00:57:18.340 --> 00:57:19.640 That's just for whatever. 00:57:19.640 --> 00:57:20.580 Oh, interesting. 00:57:20.580 --> 00:57:22.620 Just to avoid messing up your system Python. 00:57:22.720 --> 00:57:23.960 Just to avoid messing up. 00:57:23.960 --> 00:57:27.880 Or so I can have a, I can just make that virtual environment version of Python. 00:57:27.880 --> 00:57:31.760 I want it to be, even if it's not the system one in general that I want to set, you know, 00:57:31.760 --> 00:57:33.060 just like it's its own thing. 00:57:33.060 --> 00:57:35.480 And so it wouldn't actually help me, but I still like that. 00:57:35.480 --> 00:57:36.260 Cause that's smart. 00:57:36.440 --> 00:57:37.440 Yeah. 00:57:37.440 --> 00:57:37.440 That's smart. 00:57:37.440 --> 00:57:38.440 Thanks. 00:57:38.440 --> 00:57:39.440 Yeah. 00:57:39.440 --> 00:57:41.860 I've learned a lot about like, oh my gosh. 00:57:41.860 --> 00:57:43.460 I mean, I've learned a lot about packaging. 00:57:43.460 --> 00:57:47.180 Like I didn't know like anything about packaging, you know, like six months ago. 00:57:47.180 --> 00:57:49.840 I shouldn't, I should actually shouldn't admit that. 00:57:49.840 --> 00:57:53.940 Cause it's really bad for my credibility, but like I've learned a lot about packaging. 00:57:53.940 --> 00:58:00.320 And in the past, like two weeks, I've also learned a lot about like system pythons, like 00:58:00.320 --> 00:58:06.700 how Python is installed and like how on like older versions of like Debian, like Python, 00:58:06.700 --> 00:58:10.820 they do like a lot of customization to Python that makes it like really different from like 00:58:10.820 --> 00:58:14.060 I learned like way more about this than I ever thought I would. 00:58:14.060 --> 00:58:16.060 I think it's fine to say that. 00:58:16.060 --> 00:58:18.560 I'm out on system pythons, but we do support them. 00:58:18.560 --> 00:58:19.080 Yeah. 00:58:19.080 --> 00:58:19.640 Awesome. 00:58:19.640 --> 00:58:24.040 You know, someone asked me like, no, Michael, you must know all of the standard library. 00:58:24.040 --> 00:58:25.080 I'm like, what are you crazy? 00:58:25.080 --> 00:58:25.600 No. 00:58:25.600 --> 00:58:26.140 Yeah. 00:58:26.140 --> 00:58:28.080 Why don't I want to stop the standard library? 00:58:28.080 --> 00:58:29.480 Yes, exactly. 00:58:29.480 --> 00:58:31.180 And I think it's the same thing about packaging. 00:58:31.180 --> 00:58:32.920 There's like so many edge cases. 00:58:32.920 --> 00:58:36.660 If you never interact with an edge case and you don't care about that edge case, why would 00:58:36.660 --> 00:58:38.680 you take the week or whatever to study it? 00:58:38.720 --> 00:58:40.960 Like it's just, it's irrelevant to you until you need it. 00:58:40.960 --> 00:58:44.560 And now you've dove in head first here. 00:58:44.560 --> 00:58:47.400 So then you're in the deep end. 00:58:47.400 --> 00:58:48.880 We spent so much time. 00:58:48.880 --> 00:58:51.400 I shouldn't even bring this up, but we did. 00:58:51.400 --> 00:58:51.720 Okay. 00:58:51.720 --> 00:58:52.840 When we were, whatever. 00:58:53.020 --> 00:58:56.980 I want to tell the whole story because I need to not, but like we did consider like lots 00:58:56.980 --> 00:59:01.940 of different names for this tool and multiple times we considered a name and then it turned 00:59:01.940 --> 00:59:05.000 out to be a standard library module that none of us had ever used. 00:59:05.000 --> 00:59:08.860 Like we wanted to use wave W-A-V-E. 00:59:08.860 --> 00:59:09.120 Okay. 00:59:09.300 --> 00:59:10.520 But that's a standard library module. 00:59:10.520 --> 00:59:10.880 Yeah. 00:59:10.880 --> 00:59:11.140 Yeah. 00:59:11.140 --> 00:59:12.100 I was like, oh, I had no idea. 00:59:12.100 --> 00:59:14.760 Is that for working with a wave audio files? 00:59:14.760 --> 00:59:15.360 Probably it is. 00:59:15.360 --> 00:59:15.560 Yeah. 00:59:15.560 --> 00:59:16.740 Like W-A-V files. 00:59:16.740 --> 00:59:17.020 Yeah. 00:59:17.020 --> 00:59:17.620 Yeah. 00:59:17.620 --> 00:59:18.160 Interesting. 00:59:18.160 --> 00:59:18.700 Yeah. 00:59:18.700 --> 00:59:19.400 There's a bunch of stuff. 00:59:19.400 --> 00:59:23.300 There's the whole dead batteries and removing certain things that like, you know, I know it's 00:59:23.300 --> 00:59:27.040 like a library, but some of these things I had never seen. 00:59:27.900 --> 00:59:28.200 Yeah. 00:59:28.200 --> 00:59:33.220 Well, it's naming tools and packages is its own whole special deal. 00:59:33.220 --> 00:59:35.900 And we have half a million packages in PyPI. 00:59:35.900 --> 00:59:38.200 And so how are you going to not conflict with that? 00:59:38.200 --> 00:59:42.060 Have an interesting name that's not for typo squatting. 00:59:42.060 --> 00:59:42.540 Yeah. 00:59:42.540 --> 00:59:42.940 Exactly. 00:59:42.940 --> 00:59:43.240 Hard. 00:59:43.240 --> 00:59:43.560 Yeah. 00:59:43.560 --> 00:59:44.200 It's a hard problem. 00:59:44.200 --> 00:59:44.380 Yeah. 00:59:44.380 --> 00:59:46.540 I don't even get me started on the name stuff. 00:59:46.540 --> 00:59:49.460 Someone actually donated this name to us, which is very kind of them. 00:59:49.460 --> 00:59:50.880 Oh, that's actually very cool. 00:59:50.880 --> 00:59:51.540 Very cool. 00:59:51.540 --> 00:59:51.980 All right. 00:59:51.980 --> 00:59:54.260 One other thing I do want to talk about here. 00:59:55.140 --> 01:00:01.200 Padnish says, asks about, you know, is there a future where Rye, R-Y-E and uv go hand 01:00:01.200 --> 01:00:01.700 in hand. 01:00:01.700 --> 01:00:02.780 Also great work. 01:00:02.780 --> 01:00:05.020 But maybe just talk a bit about Rye. 01:00:05.020 --> 01:00:06.560 That was by Armin Roeniger. 01:00:06.560 --> 01:00:10.460 And it kind of had a similar zen as uv, but really different. 01:00:10.460 --> 01:00:11.040 Yeah. 01:00:11.040 --> 01:00:12.860 So we, yeah. 01:00:12.860 --> 01:00:15.460 So we're taking over like maintainership of Rye. 01:00:15.460 --> 01:00:23.900 And we've just been like, I think I started talking to Armin like right after Rye got released, 01:00:23.900 --> 01:00:25.500 which was PyCon last year. 01:00:25.500 --> 01:00:27.240 I think it was during PyCon it got released. 01:00:27.240 --> 01:00:29.700 And we talked like shortly after. 01:00:29.700 --> 01:00:33.220 And we just found that like we were trying to do a lot of the, we were trying to solve a 01:00:33.220 --> 01:00:33.940 lot of the same problems. 01:00:33.940 --> 01:00:37.360 And we had like very similar vision for what we wanted packaging to be. 01:00:37.360 --> 01:00:40.160 But like we were coming at it from very different angles. 01:00:40.160 --> 01:00:44.140 Like Rye kind of came at it from, let's solve the Python bootstrapping problem. 01:00:44.340 --> 01:00:45.180 So we're going to ship a Rust tool. 01:00:45.180 --> 01:00:46.160 That seems like it's main focus. 01:00:46.160 --> 01:00:46.520 Yeah. 01:00:46.520 --> 01:00:47.200 Yeah, exactly. 01:00:47.200 --> 01:00:54.100 And then internally it was actually using like pip-tools and pip to do installs and resolutions. 01:00:54.100 --> 01:00:58.740 And then the thing that, I think the thing that Armin came to realize was like, if you want 01:00:58.740 --> 01:01:03.000 to fulfill this whole vision, you kind of need to like reinvent a lot of those internals. 01:01:03.000 --> 01:01:07.080 And meanwhile, like we were building uv, which was like basically those internals. 01:01:07.080 --> 01:01:12.860 And I was like, we're going to put a lot of effort into like building this packaging stack. 01:01:13.600 --> 01:01:17.080 And so we kind of found that like, it was just a really, we're trying to get to the same 01:01:17.080 --> 01:01:17.580 place. 01:01:17.580 --> 01:01:22.300 And we were putting in a lot of the engineering investment on building a lot of kind of building 01:01:22.300 --> 01:01:22.720 blocks. 01:01:22.720 --> 01:01:28.380 And we want uv to evolve into something that could like fully replace Rye. 01:01:28.380 --> 01:01:30.880 So we want it to evolve in that direction. 01:01:30.880 --> 01:01:36.140 And in the meantime, like we're going to keep maintaining Rye and kind of using it as like 01:01:36.140 --> 01:01:39.640 a little bit of a test bed for like experimental things. 01:01:39.640 --> 01:01:42.580 Like Armin added like pie test support to it, for example. 01:01:42.760 --> 01:01:45.700 So we're like, we're kind of playing with like, you know, that's meant to be more of 01:01:45.700 --> 01:01:49.220 an experimental test bed for like the future that we want to achieve. 01:01:49.220 --> 01:01:52.900 And over time, we're going to kind of evolve uv up to the point that hopefully it can fully 01:01:52.900 --> 01:01:53.440 replace it. 01:01:53.440 --> 01:01:55.900 And then we can provide a clear migration path for it. 01:01:55.900 --> 01:01:58.260 So we're going to keep maintaining it. 01:01:58.260 --> 01:02:03.280 It is our intention for uv to like, to, you know, supplant it at some point in the future. 01:02:03.280 --> 01:02:07.340 But it'll take us like time for it to get there and be able to do all the things. 01:02:07.340 --> 01:02:11.920 But, but the visions were just so similar that it seemed like it made a lot of sense to 01:02:11.920 --> 01:02:14.560 try and consolidate and kind of team up on it. 01:02:14.560 --> 01:02:14.980 Yeah. 01:02:14.980 --> 01:02:16.680 They do seem pretty similar. 01:02:16.680 --> 01:02:19.600 And I can, I can see the direction that you're going in it. 01:02:19.600 --> 01:02:20.920 It's going to make them more similar. 01:02:20.920 --> 01:02:21.560 Yeah. 01:02:21.560 --> 01:02:22.260 I know it'll take time. 01:02:22.260 --> 01:02:26.340 And like, you know, this release of uv is like, it really is like the first release, 01:02:26.340 --> 01:02:29.860 you know, it's like, like I kind of use some of this stuff as like pretty low level. 01:02:30.860 --> 01:02:34.540 Like, like we want to build something that feels higher level and more, more automatic 01:02:34.540 --> 01:02:35.260 in the future. 01:02:35.260 --> 01:02:40.000 But this, these are kind of like the ability to install and uninstall and resolve packages. 01:02:40.000 --> 01:02:44.500 These were like the fundamental things we needed and need for anything we were going to do. 01:02:44.500 --> 01:02:48.920 And so this goal was really, let's take those and let's put them in a form that's like immediately 01:02:48.920 --> 01:02:49.860 useful to people. 01:02:50.260 --> 01:02:50.940 Yeah, for sure. 01:02:50.940 --> 01:02:53.220 One more Rye question. 01:02:53.220 --> 01:02:54.240 Then I have other questions. 01:02:54.240 --> 01:02:56.260 So uv will replace Rye. 01:02:56.260 --> 01:03:01.600 Do you recommend using Rye or uv or what's the, what's that side of the story, right? 01:03:01.600 --> 01:03:01.900 Yeah. 01:03:01.900 --> 01:03:02.180 Yeah. 01:03:02.180 --> 01:03:02.460 Yeah. 01:03:02.460 --> 01:03:03.960 It kind of depends on what you're doing. 01:03:04.080 --> 01:03:06.580 So like, like Rye uses uv under the hood. 01:03:06.580 --> 01:03:11.060 So like Rye uses uv for resolution and installation. 01:03:11.060 --> 01:03:12.480 So if you're using Rye, you're using uv. 01:03:12.480 --> 01:03:19.980 I, like I say, I consider uv to be like, especially if you're like a company or something, like I 01:03:19.980 --> 01:03:22.680 would say uv is production ready for, for you. 01:03:22.680 --> 01:03:28.680 Rye, you're buying into a little bit more of something that's a little bit more experimental. 01:03:28.680 --> 01:03:33.300 Like I think Rye is a good fit for like professional projects, hobby projects, libraries, smaller 01:03:33.300 --> 01:03:33.740 projects. 01:03:33.740 --> 01:03:39.960 I don't know that I would like, you're just buying into more change, I think. 01:03:39.960 --> 01:03:40.420 Yeah. 01:03:40.420 --> 01:03:45.160 And so if you're already in a part of your monorepo, that's got a million lines of code 01:03:45.160 --> 01:03:46.380 and a hundred people on it. 01:03:46.380 --> 01:03:46.760 Yeah. 01:03:46.760 --> 01:03:49.460 You can certainly, you can certainly try it, but you should know that you're buying into 01:03:49.460 --> 01:03:51.280 something experimental is the way I'd put it. 01:03:51.280 --> 01:03:54.900 Whereas uv, like you could also view uv as experimental if you want. 01:03:54.900 --> 01:03:56.060 Like it's very new, right? 01:03:56.060 --> 01:03:59.980 Like we've, we're like when we launched, we had, I mean, I think having issues is actually 01:03:59.980 --> 01:04:02.800 good because it means people are using your project, but we had like hundreds of 01:04:02.800 --> 01:04:03.140 issues. 01:04:03.140 --> 01:04:07.440 And we closed and we closed, we've closed like hundreds and hundreds of issues since 01:04:07.440 --> 01:04:07.720 we launched. 01:04:07.720 --> 01:04:07.900 Yeah. 01:04:07.900 --> 01:04:09.840 You guys have been super, yeah. 01:04:09.840 --> 01:04:10.900 You've been super responsive on it. 01:04:10.900 --> 01:04:12.780 Like we, you and I've had some exchanges there. 01:04:12.780 --> 01:04:13.520 Yeah, we did. 01:04:13.520 --> 01:04:14.480 And yeah. 01:04:14.480 --> 01:04:14.760 Yeah. 01:04:14.760 --> 01:04:19.120 That's some nice changes I saw going on, but like I said, almost every day. 01:04:19.120 --> 01:04:19.960 It's getting like way better. 01:04:19.960 --> 01:04:20.660 It's great. 01:04:20.660 --> 01:04:20.940 Yeah. 01:04:20.940 --> 01:04:22.600 And it's just getting way, way better. 01:04:22.600 --> 01:04:26.820 Like that's the thing about developing in private is like, like we just, once we released 01:04:26.820 --> 01:04:30.260 it, we finally, we actually had users who could use it and tell us all the things we missed, 01:04:30.260 --> 01:04:31.540 also the things we didn't think about. 01:04:31.540 --> 01:04:32.940 It's just like every week getting way better. 01:04:32.940 --> 01:04:34.940 So like it's getting more and more stable. 01:04:34.940 --> 01:04:39.280 And like, I would use it in production for my stuff for sure. 01:04:39.280 --> 01:04:44.480 And I recommend it for production, but like Rye is just a little bit of a different story. 01:04:44.480 --> 01:04:48.460 I think with Rye, you know, it's not like we're going to shut it down tomorrow or anything 01:04:48.460 --> 01:04:48.880 like that. 01:04:48.880 --> 01:04:51.960 Like it's going to be, it's going to continue to be supported for sure. 01:04:51.960 --> 01:04:57.820 but you are, and I think it's branded this way, which is you are kind of buying into 01:04:57.820 --> 01:04:59.240 a little bit of an experimental tool. 01:04:59.240 --> 01:04:59.720 Yeah. 01:04:59.720 --> 01:05:00.960 Also a bit of a workflow. 01:05:00.960 --> 01:05:01.600 Yeah. 01:05:01.600 --> 01:05:02.980 It's a very opinionated workflow. 01:05:02.980 --> 01:05:03.480 Yeah. 01:05:03.480 --> 01:05:03.980 Yeah. 01:05:03.980 --> 01:05:07.460 Whereas UVs were designed to kind of slot into existing workflows. 01:05:07.460 --> 01:05:08.220 Sure. 01:05:08.220 --> 01:05:10.280 It's a little more foundational. 01:05:10.280 --> 01:05:15.880 Speaking of which, I really like that, you know, so many of these tools have their own, 01:05:15.880 --> 01:05:17.300 their complete own way. 01:05:17.300 --> 01:05:20.620 Like, I know you're going to do that, but we're going to start up our shell and then 01:05:20.620 --> 01:05:23.640 you're going to run these three commands and then we're going to have our own lock file 01:05:23.640 --> 01:05:24.800 and we're going to have our own things. 01:05:24.800 --> 01:05:30.000 And it's, it's amazing if you drink the Kool-Aid, if you go all in on the tool, but if you don't, 01:05:30.000 --> 01:05:33.020 then it's like, ah, okay, well, what is this thing even doing? 01:05:33.020 --> 01:05:35.260 It's more in my way than it's helping me, you know? 01:05:35.260 --> 01:05:35.700 And so on. 01:05:35.700 --> 01:05:42.580 So the fact that a lot of, a lot of this is kind of baking in the pip-tools way of there's 01:05:42.580 --> 01:05:46.800 going to be this external tool that can manage some of your stuff for you, but it's not 01:05:46.800 --> 01:05:48.040 prescribing a workflow. 01:05:48.040 --> 01:05:49.720 A hundred percent props for that. 01:05:49.720 --> 01:05:50.260 Love it. 01:05:50.260 --> 01:05:50.820 Thanks. 01:05:50.820 --> 01:05:51.640 Yeah. 01:05:51.640 --> 01:05:55.580 I wanted, like, I wanted to build something that, again, I use the word like modular a 01:05:55.580 --> 01:05:58.940 lot and I think it can mean different things, but like rough was kind of like this too. 01:05:58.940 --> 01:06:03.420 Like in rough, you can use it as like just a linter or like just a formatter or both. 01:06:04.360 --> 01:06:08.140 And like with uv, like you could use uv, like just to create virtual environments or just 01:06:08.140 --> 01:06:12.640 to do dependency resolution, but use pip for your installer or something. 01:06:12.640 --> 01:06:16.720 Like it's designed to be kind of flexible in how you choose to use it. 01:06:17.600 --> 01:06:19.960 And I want to, I want to maintain that. 01:06:19.960 --> 01:06:24.600 You know, I think, I think we'll start to introduce more opinionated workflows, but I think that 01:06:24.600 --> 01:06:29.620 we'll continue to provide this kind of pick and choose model as well. 01:06:29.980 --> 01:06:30.100 Yeah. 01:06:30.100 --> 01:06:30.540 Yeah. 01:06:30.540 --> 01:06:31.160 I think it's great. 01:06:31.160 --> 01:06:33.320 It lines right up with the way I do things. 01:06:33.320 --> 01:06:38.320 And also I really like it as both someone who might do presentations either at a conference 01:06:38.320 --> 01:06:43.620 or in a course or teaching because you're not putting, you know, the whole virtual environment 01:06:43.620 --> 01:06:48.380 stuff is already such a kind of a large barrier that comes a little early in people's learning 01:06:48.380 --> 01:06:49.500 path, right? 01:06:49.500 --> 01:06:50.660 They're like, I want to run Python. 01:06:50.660 --> 01:06:51.480 I want to run this thing. 01:06:51.480 --> 01:06:53.920 Whoa, you don't just install this. 01:06:53.920 --> 01:06:54.840 You don't just use it. 01:06:54.840 --> 01:06:56.620 We're going to talk about virtual environments. 01:06:56.620 --> 01:06:57.420 Like I don't care about those. 01:06:57.420 --> 01:06:58.080 I just want to run it. 01:06:58.080 --> 01:06:59.700 Like, but you're going to need to install something. 01:06:59.700 --> 01:07:00.460 So here we go. 01:07:00.460 --> 01:07:04.480 We're adding more workflow and more specific ways of working there. 01:07:04.480 --> 01:07:08.800 I think while trying to help a lot of times they end up just adding more friction at the 01:07:08.800 --> 01:07:09.120 start. 01:07:09.120 --> 01:07:09.600 Yeah. 01:07:09.600 --> 01:07:10.200 Yeah. 01:07:10.200 --> 01:07:16.280 And like, I guess my philosophy there is like, I want us to, I want us to like embrace virtual 01:07:16.280 --> 01:07:16.940 environments. 01:07:16.940 --> 01:07:20.940 Actually, Rai has this philosophy too, which is like Rai does use virtual environments and 01:07:20.940 --> 01:07:24.480 like virtual environments actually have a lot of, there's a lot of benefits to embracing 01:07:24.480 --> 01:07:24.760 them. 01:07:24.760 --> 01:07:28.700 Like all the editors and such, they're all built around virtual, they all are all built around 01:07:28.700 --> 01:07:30.700 virtual environment detection and stuff like that. 01:07:31.540 --> 01:07:36.340 But I wanted to feel, I want to feel a little, honestly, like it has a really bad reputation, 01:07:36.340 --> 01:07:40.740 but I wanted to feel more like node modules, you know, where it's like, it just kind of 01:07:40.740 --> 01:07:43.540 works in your, it just kind of sits in your project and your project just does the right 01:07:43.540 --> 01:07:43.820 thing. 01:07:43.820 --> 01:07:47.880 and I know people make fun of node modules usually for being big. 01:07:47.880 --> 01:07:51.780 but like, but that's the developer experience I want, which is you don't think, you don't 01:07:51.780 --> 01:07:53.440 think of the virtual environment as this weird other thing. 01:07:53.440 --> 01:07:56.120 It's just kind of like the project environment and it just kind of works. 01:07:56.120 --> 01:08:01.880 And I think, I think you've got that by not like fielding people from it, but abstracting 01:08:01.880 --> 01:08:04.060 some of it away for people that don't need to think about it. 01:08:04.060 --> 01:08:04.520 Yeah. 01:08:04.520 --> 01:08:10.400 There was a, there was an amazing, it's not a joke, but it's, I guess something that's 01:08:10.400 --> 01:08:14.320 a prank that somebody, came up with here. 01:08:14.320 --> 01:08:15.100 I'll put it on the screen. 01:08:15.100 --> 01:08:20.500 It's called when everything becomes too much, the npm package chaos of 2024. 01:08:20.500 --> 01:08:26.280 And I was just, as you were talking about node modules, it's amazing. 01:08:26.280 --> 01:08:33.380 So it's more of a statement of like, well, how super interdependent or how many combinatorial 01:08:33.380 --> 01:08:37.240 or transitive dependencies does a couple of npm things have? 01:08:37.240 --> 01:08:43.160 Like, for example, I used tailwind, but there was maybe 50 things in my, my node module for 01:08:43.160 --> 01:08:45.240 just using tailwind in my project. 01:08:45.240 --> 01:08:45.500 Right. 01:08:45.500 --> 01:08:51.460 So anyway, this one is some, this, an npm username, Patrick JS launched a troll campaign 01:08:51.460 --> 01:08:55.580 where a package called everything, it depended on every other npm package. 01:08:55.580 --> 01:09:01.060 So if you, you know, npm install everything that literally tried to download the repository. 01:09:01.060 --> 01:09:02.800 That's kind of, that's kind of amazing. 01:09:02.800 --> 01:09:03.600 Wow. 01:09:03.600 --> 01:09:09.160 I know it's, it's devious, but it's also kind of like, okay, that's, that's something else. 01:09:09.160 --> 01:09:09.940 Yeah. 01:09:09.940 --> 01:09:13.080 And Python independency trees tend to be much, 01:09:13.080 --> 01:09:17.500 like shorter, I guess, and like smaller, like you would tend to have way fewer dependencies. 01:09:17.500 --> 01:09:18.300 Yeah. 01:09:18.300 --> 01:09:21.220 they are sometimes like heavier, I guess. 01:09:21.220 --> 01:09:23.640 Cause like in Python, it's very common to have lots of native code. 01:09:23.640 --> 01:09:30.300 and like, I don't know, like PyTorch, like the PyTorch wheels that you download when 01:09:30.300 --> 01:09:33.840 they're zipped are like between a hundred and 200 megabytes. 01:09:34.380 --> 01:09:34.760 Wow. 01:09:34.760 --> 01:09:35.320 Okay. 01:09:35.320 --> 01:09:35.980 Yeah. 01:09:35.980 --> 01:09:36.180 Yeah. 01:09:36.180 --> 01:09:40.220 Like all the, all the ML stuff and all the native stuff, like those are like big, but, 01:09:40.220 --> 01:09:42.900 but the dependent, but the number of dependencies tends to be a lot smaller. 01:09:42.900 --> 01:09:43.500 Yeah. 01:09:43.500 --> 01:09:44.320 I totally agree. 01:09:44.320 --> 01:09:49.240 It's also very common to ship like foundational packages with like no dependencies or very few 01:09:49.240 --> 01:09:51.940 dependencies, which is very hard to do in, in JavaScript. 01:09:52.300 --> 01:09:53.060 Yeah, it is. 01:09:53.060 --> 01:09:55.020 it definitely is. 01:09:55.020 --> 01:09:55.400 All right. 01:09:55.400 --> 01:09:56.780 I think we're getting short on time here. 01:09:56.780 --> 01:10:01.200 I think I probably should let you go back and knock out another release of uv. 01:10:01.200 --> 01:10:02.340 I do have to do a release today. 01:10:02.340 --> 01:10:02.660 Okay. 01:10:02.660 --> 01:10:03.120 Thank you. 01:10:03.120 --> 01:10:03.380 Awesome. 01:10:03.380 --> 01:10:04.020 Cool. 01:10:04.020 --> 01:10:08.500 So Tony, the audience has been using uv in production for a couple of weeks now. 01:10:08.500 --> 01:10:09.100 Awesome. 01:10:09.100 --> 01:10:10.320 Seems very excited about it. 01:10:10.320 --> 01:10:12.000 And also I have two. 01:10:12.000 --> 01:10:14.180 It's an absolute delight. 01:10:14.180 --> 01:10:14.880 It hasn't. 01:10:14.880 --> 01:10:15.800 Oh, thank you so much. 01:10:16.300 --> 01:10:16.540 Yeah. 01:10:16.540 --> 01:10:21.800 I know there's certain things that it didn't do like right away, but then it, it came 01:10:21.800 --> 01:10:23.480 out and you added them quickly. 01:10:23.480 --> 01:10:28.640 I know there was some criticism for y'all for developing this in private and then releasing 01:10:28.640 --> 01:10:28.880 it. 01:10:28.880 --> 01:10:31.300 And I just want to say like, you know, thanks for doing that. 01:10:31.300 --> 01:10:32.460 And what's the alternative? 01:10:32.460 --> 01:10:37.300 You start with just a blank GitHub repo and people start to say, well, you should do this. 01:10:37.300 --> 01:10:37.920 You should be doing that. 01:10:37.920 --> 01:10:38.760 Like we have a vision. 01:10:38.760 --> 01:10:42.580 Let us just like get it a little structure in place and then we'll open source it. 01:10:42.580 --> 01:10:44.200 Like give us a month, you know? 01:10:44.200 --> 01:10:45.320 Yeah. 01:10:45.320 --> 01:10:45.700 I don't know. 01:10:45.700 --> 01:10:47.960 I think it's, I think it's excellent work you're doing. 01:10:47.960 --> 01:10:49.200 I appreciate it. 01:10:49.200 --> 01:10:49.880 I appreciate it. 01:10:49.880 --> 01:10:50.040 Yeah. 01:10:50.040 --> 01:10:50.500 Yeah. 01:10:50.500 --> 01:10:50.800 Yeah. 01:10:50.800 --> 01:10:52.460 You know, like I said at the start, it was different. 01:10:52.460 --> 01:10:58.220 It's different this time because anything we release, like, you know, you could work 01:10:58.220 --> 01:11:01.660 on a rough for a month and nobody cares until it starts to catch some traction. 01:11:01.660 --> 01:11:01.900 Right. 01:11:01.900 --> 01:11:03.560 But this is the instant that it hits. 01:11:03.560 --> 01:11:04.240 Yeah. 01:11:04.240 --> 01:11:07.420 We knew people would at least, would at least look at it and at least try it. 01:11:07.420 --> 01:11:10.980 And I think, you know, we wanted to make sure that whatever we released, first of all 01:11:10.980 --> 01:11:11.400 was good. 01:11:11.400 --> 01:11:13.500 And second of all, that we were ready to maintain it. 01:11:13.980 --> 01:11:18.280 And then I think with packaging too, it's like, there's just a lot of, it's just such 01:11:18.280 --> 01:11:22.980 a complex space that like, I wanted to make sure that we had a lot of clarity in the messaging 01:11:22.980 --> 01:11:27.280 around what the tool is and what it's not and like what we want it to be and what it isn't 01:11:27.280 --> 01:11:27.580 yet. 01:11:28.040 --> 01:11:32.600 And so I'm really happy with how the launch went and I've really appreciated just all 01:11:32.600 --> 01:11:37.620 the excitement, activity, engagement that we've had on the repo. 01:11:37.620 --> 01:11:39.180 It's been like, I don't know. 01:11:39.180 --> 01:11:45.380 I mean, honestly, it's been a lot of work, but it's like really, really, no, but it's awesome. 01:11:45.380 --> 01:11:48.340 It's like so energizing for me and for the team. 01:11:48.340 --> 01:11:52.420 And to see all the people picking it up and making it the foundation of their projects. 01:11:52.420 --> 01:11:53.640 And yeah, it's really cool. 01:11:53.640 --> 01:11:53.920 Yeah. 01:11:53.920 --> 01:11:54.880 And it's cool. 01:11:54.880 --> 01:11:57.680 You know, I guess another, like one other thing that's a little different this time is 01:11:57.680 --> 01:12:02.700 like, I talk, I just talked to more people who work on Python tooling at companies now than 01:12:02.700 --> 01:12:07.120 I did when I released Ruff, just because over the course of the past year, I've just met 01:12:07.120 --> 01:12:10.080 a lot of people and we've just talked about how they're using Ruff, how they're not. 01:12:10.080 --> 01:12:13.920 And so just hearing some of those stories too, of how like companies are starting to use 01:12:13.920 --> 01:12:17.340 it and like what the blockers are and like how much of a speed of it's providing has been 01:12:17.340 --> 01:12:18.200 pretty cool to see too. 01:12:18.540 --> 01:12:20.440 So yeah, I appreciate, I appreciate it though. 01:12:20.440 --> 01:12:21.520 Yeah, you're welcome. 01:12:21.520 --> 01:12:21.800 All right. 01:12:21.800 --> 01:12:23.280 We'll leave with this final thought from Juan. 01:12:23.280 --> 01:12:25.620 I installed 94 libraries in around two seconds. 01:12:25.620 --> 01:12:26.200 Incredible. 01:12:26.200 --> 01:12:26.540 That's great. 01:12:26.540 --> 01:12:26.780 All right. 01:12:26.780 --> 01:12:27.160 Yeah. 01:12:27.160 --> 01:12:29.240 All right. 01:12:29.240 --> 01:12:33.000 Well, I'm excited to see where things go and, you know, we'll maybe do a follow-up when, 01:12:33.000 --> 01:12:36.880 when you've got some more, more of the ideas in place. 01:12:36.880 --> 01:12:37.600 Sounds great. 01:12:37.600 --> 01:12:38.840 No, it's always, it's always a pleasure. 01:12:38.840 --> 01:12:39.860 Thanks so much for having me on. 01:12:39.860 --> 01:12:40.560 I really appreciate it. 01:12:40.560 --> 01:12:41.020 Yeah. 01:12:41.020 --> 01:12:41.540 Thanks, Charlie. 01:12:41.540 --> 01:12:45.240 This has been another episode of Talk Python To Me. 01:12:45.240 --> 01:12:47.060 Thank you to our sponsors. 01:12:47.420 --> 01:12:48.660 Be sure to check out what they're offering. 01:12:48.660 --> 01:12:50.080 It really helps support the show. 01:12:50.080 --> 01:12:56.400 It's time to stop asking relational databases to do more than they were made for and simplify 01:12:56.400 --> 01:12:58.640 complex data models with graphs. 01:12:58.640 --> 01:13:05.080 Check out the sample FastAPI project and see what Neo4j, a native graph database, can do 01:13:05.080 --> 01:13:05.460 for you. 01:13:05.460 --> 01:13:09.900 Find out more at talkpython.fm/neo4j. 01:13:09.900 --> 01:13:11.880 Want to level up your Python? 01:13:12.280 --> 01:13:15.920 We have one of the largest catalogs of Python video courses over at Talk Python. 01:13:15.920 --> 01:13:21.100 Our content ranges from true beginners to deeply advanced topics like memory and async. 01:13:21.100 --> 01:13:23.760 And best of all, there's not a subscription in sight. 01:13:23.760 --> 01:13:26.680 Check it out for yourself at training.talkpython.fm. 01:13:26.680 --> 01:13:28.780 Be sure to subscribe to the show. 01:13:28.780 --> 01:13:31.560 Open your favorite podcast app and search for Python. 01:13:31.560 --> 01:13:32.860 We should be right at the top. 01:13:33.260 --> 01:13:38.040 You can also find the iTunes feed at /itunes, the Google Play feed at /play, 01:13:38.040 --> 01:13:42.240 and the direct RSS feed at /rss on talkpython.fm. 01:13:42.640 --> 01:13:45.200 We're live streaming most of our recordings these days. 01:13:45.200 --> 01:13:49.320 If you want to be part of the show and have your comments featured on the air, be sure to 01:13:49.320 --> 01:13:53.040 subscribe to our YouTube channel at talkpython.fm/youtube. 01:13:53.040 --> 01:13:55.080 This is your host, Michael Kennedy. 01:13:55.080 --> 01:13:56.380 Thanks so much for listening. 01:13:56.380 --> 01:13:57.540 I really appreciate it. 01:13:57.540 --> 01:13:59.440 Now get out there and write some Python code. 01:13:59.440 --> 01:14:20.320 I'll see you next time.