Python Bytes - #455 Gilded Python and Beyond

Starting point is 00:00:00 Hello and welcome to Python Bites, where we deliver Python news and headlines directly to your earbuds. This is episode 455, recorded October 27, 2025, and I'm Brian Ockin. I'm Michael Kennedy. And this episode is sponsored by the wonderful people at Python Bites, actually both and everybody here. So please check out the work we do that we do for a little bit of money. We've got Talk Python training with tons of wonderful courses. his new book of course michael's new book and then um also if you if you want to check out pie test there's the complete pie test course or you can take the pie test course at talk python

Starting point is 00:00:40 training and as always thank you to our patreon supporters um you guys rock uh we don't shout you out all the time but we always appreciate you if you'd like to submit a topic idea or some feedback for us or just say hi um check out the show notes uh at python bytes.fm. There's links to our socials. We've got the show and both Michael and I are on Blue Sky and Faustanon,

Starting point is 00:01:10 but that's really Massina. So wherever you're at there. If you're listening to this and you're like, sounds like they're recording live somewhere we can see it. Yes, we are recording and streaming, but you can watch all the back episodes or participate in the chat if you want to.

Starting point is 00:01:28 Check out Pythonbytes.fm. Live with all the details there and a link to when we're going to record next. Before we get started, one last thing. We do send out a newsletter that's getting better and better as we go along, thanks to both Michael and I working on a little bit. So there's a newsletter we send out with all the links. So you don't have to take notes. You will send you out an email.

Starting point is 00:01:55 So sign up for the newsletter and we'll send you those links. and extra details too. It's a great resource. Yeah, I think people really appreciate it, Brian. You know, if you look at when you send out the email, then book back in the campaign, and how many people actually open to the thing that we send out or click on it?

Starting point is 00:02:13 Those are really close to 100%, which is ridiculous. So that's a really good sign that other people might also like it. Yeah, I think that people are using it in place of taking notes. And also, if you're in a car or something, you don't want to, like, try to remember. You don't have to. So awesome. All right, Michael, what kind of, what do you got for us right away?

Starting point is 00:02:33 Well, let me see what I can do here. Let's see what the Cyclops says. A cyclops. It's like a little programming kitty, but it's a cyclops. It's a cute. Like opts, O-PT, operations, cyclop. I get it. So we've heard of Click.

Starting point is 00:02:51 You and I were both just actually singing the praises of Argpars, which is pretty interesting. I am such a fan of ARGPARs these days because it's just built in and it's simple and it's good enough. But if you were building something where the CLI API or interface or dev user experience was super important, you might choose something other than ARGPARs. So what are your options? You could choose Click. You could choose Typer. Typer is actually built on Click.

Starting point is 00:03:20 So in a sense, it's like Starlit and Fast API. You choose one, you get the other in a sense. So there's a blend there. And there's others. I know I'm not, there's a bunch of different CLI building options. But the Cyclops is one that sort of has the second mover advantage over Typer. And there's sort of tagline, I remember where I saw it, but somewhere in here, it's not actually right here on this page. But, you know, no shade at Typer, but it says what you thought Typer was going to be. And by second mover advantage, it's like, all right, we've seen different interface things built based around type information.

Starting point is 00:03:55 And now that we've seen the pluses and the minuses, could we do a slightly different version that has more pluses and fewer minuses? And so there's a whole page of the Typer comparison here. And I'm not a super expert on Typer. My CLI is not that great. So there's 13 different things, 13 different components or features or aspects of the API that they chose to address to say, type is great. We're basically inspired by Typer. That's what we're doing.

Starting point is 00:04:26 However, there's a couple of these 13 areas I think we should make better, right? So you could actually go through and look at all of them. Probably the biggest one is that Typer was made with Python 3738, where Python's types were at an earlier stage in their life. You had to say things like, this is an argument of int, or it is an int when it's actually an argument of int or something like that. And this Cyclops thing was built once annotated was introduced, which allows you to have sort of a dual role, like it really is one of these, but its type could also be like its native value, right, allowing the removal of proxy defaults. Why does that matter? Because if you build a CLI with Cyclops, you can still, and you use types, you can still use it as regular functions as if they just had an integer where it said it took an integer and so on.

Starting point is 00:05:20 So, you know, things like that. It's got about, I think, a thousand GitHub stars or so. It's not crazy, crazy popular. But if you're really loving that concept of having types for your TLI, you know, give a thing a look. Yeah, actually, I got a shout out to them for doing things like comparing to Typer. Because, like you said, a lot of us, I use Typer as well. And I've used Typer.

Starting point is 00:05:44 I've used Click and ArcPars. and so if you're you're probably not going to grab too many of the arc parse people because they're just using it because it's built in maybe but like the unit test people Brian what can you do you talk pie test for 10 years I yell louder that's what I do but the but I think it's a good a fair thing because I'd be thinking well if I'm not going to use arc parse I'm probably going to reach for type or click lately typeer But having this comparison here, and then even a migration from Typer,

Starting point is 00:06:22 I think that's cool to show that. And there's always room for making CLI tools better. We love CLI tools. Yeah, there definitely are. But yeah, so cool. I ran across this and I built something with it recently, and it was nice. So we're checking out.

Starting point is 00:06:41 A little, I'm starting to add. I've been kind of, I've drawn the Kool-Aid now that maybe having an actual CLI, even from my own tools, might be a little bit better than just, here's a script I run, and maybe I'll pass it something and just look at sys.orgv and see if it's in there. Oh, man. Let's do a little bit better. And now that I've started to do that, I've been playing with these different ones, which is,

Starting point is 00:07:04 I think, where I kind of got inspired to go down this path and talk about it today. Yeah, awesome. Cool. I want to talk about the present, the future, a little bit of both. So there was this interesting article. Of course, in Python 314, well, we had it in 313, is that the gill was optional with the, what they call it, the free threading version, right? Free threaded Python. Not the gilless, not the gillectomy, free threaded.

Starting point is 00:07:38 I think it's where we landed from the PEP. Okay, but is it kind of the same thing? Yeah. Okay. Oh, right. The gallectomy was a different project. So, yeah, I was, um, gosh. Anyway. Larry Hastings was doing that.

Starting point is 00:07:53 But I believe, I mean, the free threaded stuff, there's a lot of things that had to happen. And actually, my next topic is also on free threaded Python. And it's not just let's take out the Gill, but they had to rework memory management, um, the, the structure of object. There's like a lot of moving parts in order to make this possible. Okay, cool. Um, well, okay. So with, with the, uh, free threaded Python.

Starting point is 00:08:15 So now we have in three. 313, we had two versions of Python release. So if you go to download Python, you get a choice. And with UV2, you can add a T to it, 314T, and you get the free-threater version. See, but like, and also definitely encourage everybody to test and publish that you've test. In CI, you go ahead and test it because the CI tools support both. Why do we care about this? Well, because some people are migrating to it.

Starting point is 00:08:46 Because as with 314, it's no longer just an experimental thing. So here, I'm going to cover an article by Giovanni. Burr really? Sorry, Giovanni. I can pronounce that one. I'm pretty sure this is the same Giovanni that makes Grannion. Oh, okay. I think the future of Python web surfaces looks gill-free.

Starting point is 00:09:10 So starts out with two. major changes when comparing free-threaded variant of 314 versus 313. First, free-threaded support now reaches phase two, meaning it is no longer considered experimental. So everybody's on board with, this is what we're going to at least try to do for a while. Not just try to do, but the Python core team is supporting this, and we're going to do it moving forward. What does that mean? It means that people can depend on a free-threaded version and change their code according

Starting point is 00:09:43 if necessary, which is great, especially for a lot of async stuff, like web services. So secondly, the implementation is now completed, meaning that the workarounds introduced in 313 make code sound. To make code sound without the Gill are now gone, and free-threaded implementation now uses the adaptive interpreter as the Gill-enabled variant. What does that mean? It means that the additional optimizations, oh, okay. Also, those facts plus additional optimizations make the performance penalty that we had in 313 way better or way less annoying. So in 313, the free thread, if you're doing non-Async code, just straight code, you had to face it like a 35% time penalty, which sucks. Now it's like a 5 to 10% difference and I think it's going to shrink even more.

Starting point is 00:10:40 but so if you're if you don't need the free threaded don't use it unless or or use it whatever but if you if it's time critical don't right now but I think it's going to get better so and especially if you do are using async and stuff it's it's way faster to use the free threaded of course so this this article does a shout out to Miguel Greenberg's article about performance I think we covered it last we did you did okay no maybe I did But we did. I can't remember who started it. So this has a whole, this article has a whole bunch more benchmarks, and I'm not going to go walk through all of them, but talks about doing a service with Jason responses and both

Starting point is 00:11:26 measuring with whiskey and ASCII implementations. And yes, talking about using Grinian, so that makes sense. So I'm going to jump down. So there's a bunch of metrics. Basically, the end result is if you're using ASCII or ASGI, the free-threaded one is obviously the way to go. And you don't get really much of a hit because of free-threaded implementation at all. You get speed-ups and memory usage is reasonable. So what I want people to do, if you're going to check this out, is to jump down to the final thoughts

Starting point is 00:12:04 because there's some great stuff here. essentially so what do I want to pick out on asynchronous protocols like ASGII despite the fact that concurrency model doesn't change that much it's a shift from one event loop per process to one event loop per thread I think that's a big change actually just the fact that we no longer need to scale memory allocations just to use more CPU is a massive improvement

Starting point is 00:12:33 that's cool didn't know that that's neat So for everybody out there coding a web application of Python, simplifying the concurrency paradigms and the deployment process of such applications is a good thing, obviously, and the conclusion being, for me, the future of Python web services looks like the gil-free. So if you're doing web services, try out the free-threaded version. At least one developer there is completely happy with what we have now. So, yeah, super nice. A bit of real-time follow-up.

Starting point is 00:13:05 Yes, indeed. It is the same Giovanni. All right. Who creates a granion? So with any of these web production app servers where you run your Python code when it's running, whether it's Django, Flask, whatever, there's a process that runs your code. Very often, the way that we've got them to work around the Gill, like in the web, the gill is a much smaller problem. It's still a problem, but it's less. Because what we do often is we create a web garden.

Starting point is 00:13:33 So when you set up grand, you can say, and start four copies of yourself. So that, yes, they all have the gill, but we'll round robin requests between the fours. If you got four cores, four CPUs, you basically can say, well, each one of these is dedicated to a CPU and it can kind of like match, right? Yeah, so we're running on grainy, good stuff. But the thing is, when you have this free threaded option, you can actually have true concurrency in your one worker. So instead of scaling out four copies, you could have just one. and to say, let that one take 10 concurrent requests or whatever it needs to take.

Starting point is 00:14:08 Yeah, right? So that's how the free thread it gets better. And like, well, okay, why? What's like, you know, six, one, half dozen the other? No, the memory becomes a problem when you create these little web gardens. Because if normally your server would use half a gig and you create four, well, now it's two gigs.

Starting point is 00:14:23 And maybe that bumps you up in another tier and so on, right? Yeah, and if you need that for, yeah, if you need that memory for data, you don't have it. Yeah, so I think that's kind of a one of the angles of not just your writing concurrent code but the foundations of your code running can do so more efficiently because they're no longer needing to work around processes uh limitations like well we got to faint out these processes because of the kill okay back to the regular scheduled programming free threaded python let's keep going so here's an interesting article from a snake

Starting point is 00:14:56 wearing a very fast it's a very fast snake with jet engines uh that is wearing a garbage collector so So what I want to talk about is unlocking the performance in Python's free-threaded future GC optimization. So garbage collection is not something that we frequently talk about a ton in Python. You know, it's just, we don't think a lot about GC, we don't think a lot about memory. I don't know why, like, as you know Brian and C, C++, people are like always about it, right? Like, oh my gosh, what size of point, what size of integer are we going to use for this, right? And like, it's just way more sort of in the forefront, but it's nice every now and then to peel

Starting point is 00:15:33 back the covers and get a look at what's going on here. So Neil Schemannauer wrote this article over here from Quantite. And they do a bunch of data science primarily. But here, check this out. This is news to me, actually, even though I'm still interested in these things. It says, first, the most important thing to know is that free threaded Python uses a different garbage collector than the default Python, the Gill enabled. The gilded, I'd still want to call it the gilded python. The gilded python.

Starting point is 00:16:01 So we have the default GC, which people probably understand pretty well. When you create a number, like a thousand, when you create a list, you create an object from a class, all of those things have a data structure at the top of them that allows them to manage and track the memory. And people think of the Gill as being a threading thing. The Gill is really a protection mechanism for memory management in Python, right? It's protection against a race condition in the reference counting. So basically interacting with this structure that says seven things reference me and I

Starting point is 00:16:38 reference it. However, you know what I mean? Like that sort of deal. Yeah. And when the thing goes out of scope, it decrements that and that's where the gill comes in to like make that decrement safe and so on. So that's the regular one. But we don't even have that structure anymore in the free threaded one.

Starting point is 00:16:54 So, so what? So instead, when you allocate stuff, it goes into a special memory managed heap called Mimalloc, or managed by the Mimalloc, right? And this allows the GC to loop over all these objects and figure out which ones are junk and which ones are a little bit like a mark and sweep garbage collector. So there's a couple of interesting things. One is that most of the mark and sweep garbage collectors and that type of thing are, they have generations, like a generational one. So it has gen zeros where most objects come,

Starting point is 00:17:31 then it'll check that frequently and so on. That's also how the garbage collector that looks for cycles and the regular Python goes, right? But it doesn't work here because of like unmanaged sea extensions and a bunch of stuff like that. So what it does is it can mark all the objects reachable from what it calls known roots. So instead of trying to scan the entire memory space. It says, well, what are globals? What are locals? What are like a, you know, it sort of follow those and it marks those as active. And that can, they can automatically be excluded from the search, which is kind of like a generational optimization. So pretty interesting. But here's the takeaway. We were talking about making things faster by using free threaded

Starting point is 00:18:13 Python. With this one, it's the free threaded GC collection is between two to 12 times faster than the 313 version. Wow. That's pretty, yeah, it's pretty wild, right? Yeah. Yeah. So depending on how your algorithms work, do you have a super pointer-pointer-heavy allocation-heavy type of algorithm, then, you know, we'll probably see a bigger benefit than if you just allocate a few things and jam on those. So anyway, if you want to peel back the covers and see the GC, I didn't even know they were different, but apparently they're different. And presumably, maybe they could even have like multi-threaded GC stuff going on. That might be cool. Or background thread GC. I know those are a lot of different things that happen in some of these systems, these garbage-class systems.

Starting point is 00:18:52 Yeah. And finally, a little bit of real time, not quite real time, but real world follow-up, I'll say. So for some reason, I can't remember what I was doing. Something I had to do, I had to go back, go through all the examples of my async programming course at Talk Python. So I went back in there and I said, all right, well, let me just make sure everything's running, update all the dependencies. I remember what it was. But anyway, something had to be updated. So I made sure all the dependencies got updated and penned them to the newer ones. And I'm like, well, there's a lot of changes here. so they're all running. So go around and I ran every example from the course, probably 25 little apps. And many of them are like, well, here's a web app or here's just a way to do multi-processing. Then we'll look at it. But a lot of them were like, let's actually do this with threads. And in this example with doing I.O, you see it does make a difference.

Starting point is 00:19:40 So this example doing computational stuff, no difference or slower because there's just overhead and it's still running one at a time, right? Well, I decided to type UV-dash-Python 3. 14t run those examples and run them free threaded the ones that used to have no benefit or a net negative are now like seven times faster they would it would go like synchronous one 10.2 seconds async or threaded one rather 10.25 seconds you know and now just put the t on the end three seconds that's pretty cool that is pretty cool man so To me, it's really down to what are the framework, what are the packages and libraries that you're using?

Starting point is 00:20:28 If it's green across the board for free threaded, that's pretty interesting. It opens up some real possibilities. Yeah, also, I'll be looking forward to the, I know, I don't know if we have a deadline or timeline for this, but when we can go back to having one version of Python and it's being the free threaded one, it'll be, you know,

Starting point is 00:20:52 maybe we can switch it to be the default is the free-threaded one. And for a couple versions, there's the other one also. You've got to say 314G. Yeah, or something like that. And I think that would be good because the default way we think about how to program and how to program quickly and the rules of thumb based on the gill are gone. We'll be gone. So we'll teach people different.

Starting point is 00:21:17 Yeah. So, yeah. Anyway. Yeah. And before we move on real quick, a follow-up. to your topic from chart here. Also, if you're on Kubernetes, if you're on Kubernetes,

Starting point is 00:21:27 it makes setting reasonable CPU and memory requests more difficult to have to scale out, like the web garden type stuff. Oh, okay. Yeah, so if you just have one process, like this is the one, you know, I think it's easier. I believe that's what he was getting at.

Starting point is 00:21:41 Yeah, and also, I mean, from the C++ plus world, yes, processes and threads both can work. But thinking in threads is easier than thinking about multi-processes. 100% because you have shared memory. You don't have to proxy stuff over and figure out how to sync it back up and copy it. To queues and all sorts of even.

Starting point is 00:22:02 Anyway, OK. I can't resist. I got to say, I got to put this out there in the world and see what you think. One thing right now, we have the gilded Python and we have the free threaded Python. And I can choose to run the free threaded one as we've been talking about if I want that.

Starting point is 00:22:16 But what about this, Brian? This is what I want to put out in the world. I want, like right now if I have a project, and I got some data and I like I need to process it some parallel on the gilded one. I need to maybe do multi-processing. So that's going to spin up five little baby Python processes all to go jam on that, right? That's how you get around the gill. What about this?

Starting point is 00:22:36 What if you could pass a flag to multi-processing and it would start up re-threaded Python with threads instead of a bunch of different processes that can't coordinate on whatever algorithm of that? You just say run multi-processing, but this time do it in threaded, a free-threated, a free- threaded Python as your little multi-process burst and then come back to me with the answer that would be cool yeah yeah I'm not you're not convinced I mean you could that way because I mean your code has very different though just that one function though and you don't pick you know what I mean you're like this one function that I'm gonna call multi-processing on I'm gonna write it assuming that it can be it will be an advantage to be threaded or it will be an advantage to be concurrent but the rest of my code I don't have

Starting point is 00:23:16 to rewrite like it'd be a cool way to sort of bring kind of like Scython lets you bring and say this one function, if it was just this were way faster, it would make all the difference. You could do that, but for free threading. I would like to see that. It would be interesting. Yeah. Yeah, or just a chunk to say like, yeah, this, this subsystem is a through free threading. With free threading.

Starting point is 00:23:36 Yeah. No, all right. Go ahead. Okay. Completely different tangent is we're going to go back in time by a couple of one or two episodes. We've been, I, the last few episodes, or two or three talking about. lazy imports because I thought they were coming in 315 they're proposed in 315 but they're not here so or they're not accepted yet I haven't checked I don't think that anything's changed

Starting point is 00:24:02 so so last episode I did talk about lazy imports that you can use today responding to that Bob Belderboss wrote up was had a discussion on LinkedIn and part of that discussion we had Willa McCugan hop in and said that he's excited about the pep and that he uses lazy he has a lazy loading mechanism for textuals widgets which totally makes sense like widgets you might just need a button but you don't need like all of the all of the other widgets be cool if you could just load what you needed and he's got that so I checked it out I was checking this out and I think this is sort of a small topic but I think it's cool. So the idea that I want to highlight is I presented last week in A Method for doing lazy loading now or lazy importing now on things that you depend on. But what if you're like, you don't have to wait. You can be lazy today.

Starting point is 00:25:06 But what about like, yeah, exactly. What about if you're the package? If you have a big package that you have a bunch of stuff that people might want to import, making it so that you're net and you're not the problem, that your package is going to import really fast because behind the scenes you're doing lazy importing for people. And that's essentially what Textual does. And the implementation is like really easy.

Starting point is 00:25:32 He's using a basically when you access anything, he's overriding the get adder function. So if you access a widget, it's going to try to just grab it out of a dictionary. And of course, right away, it's going to fail. And then if it fails, he goes ahead and does it like goes ahead and imports it. And then stores it in. So the next time somebody tries to grab it, it's going to be there. And so there's a little bit, a little bit of misdirection here. But in the end, you get lazy loading on, on every widget access. So pretty cool. And so I thought, you know, this is pretty neat. I don't want to

Starting point is 00:26:17 just to be hidden it's not hidden it's an open source project so i went ahead and wrote all this up in a new or a new post called polite lazy imports for python package maintainers very nice link to that also but i just thought that the implementation is totally clever and um and maybe it's maybe for other package maintainers this is like well yeah that's the way you do it but um it was new to me and i thought it was neat so there we go also want to shout out i didn't realize there was both get adder get Adder is used for if it's not found, for missing items. But also, there's another one called Get Attribute. And I did look up.

Starting point is 00:26:56 I really appreciate that Trey Hunter had posted this every Dunder Method article. So pretty cool. I totally have this bookmarked so that I'm like, if I want to understand a Dunder method in Python, I go here. Yeah, very cool. I know it for classes, but for modules. as well, interesting. Yeah, I don't think, and packages. So this is, you go ahead and throw it.

Starting point is 00:27:22 Well, I guess it is for modules, but if you put that in a Dendor Init file, then it's for your entire package. Yeah, exactly. It's great cool. Anyway. I love it. Nice.

Starting point is 00:27:34 All right. Is that all your extras? Oh, that's just. That's my main thing. I do. I do have one extra. The one extra, so one of the things I've been doing is, is I've been,

Starting point is 00:27:46 And I paused Python people a long time ago, and I totally stopped doing testing code. I'm focused on this and focused on work, but I'm also writing more. And one of the things I'm doing is writing a book on test-driven development. And I don't have an announcement yet, but in the next couple weeks, I think I might have an announcement. But if you'd like to know, I guess what I'm going to do is I'm going to write the book, but to motivate me and to make people not have to wait, I'm doing. Doing it is like a, you know, like Manning does like the early access books and Pragmatic has beta books.

Starting point is 00:28:24 I'm kind of doing that. I'm doing a lean, trying to do a lean startup applied to books. And I'm going to release it after I've got, I'm like, you know, do a rough first draft, clean it up a little bit and then release it for every chapter. And then as I go along, incorporate feedback from people. And then once the whole thing is good and polished, I might bring on an editor or something or maybe just release it. But if you'd like to know more about that,

Starting point is 00:28:51 I'll announce it here, but also if you'd like to know the minute it's available, join the newsletter over at Python Test, and I'll let you know. Yeah, we really do appreciate when you join our newsletter. You might think, oh, we can just announce these things on social media or on the podcast, but it's not the same people.

Starting point is 00:29:07 A lot of people miss it because they miss an episode or like social media is just a screaming feed of stuff. It really makes a difference. So join the newsletter. Um, also I want to like talk about the new, there's, there's newsletters and then there's newsletters. So, um, this isn't really, I don't do a weekly like eight tips on testing. That would be cool. I just don't have time to do that. But, um, so it's mostly announcements if there's something to, to keep track of. But people like you and me and others, um, we're using

Starting point is 00:29:35 things like, like I'm using kit, but there's others, uh, I've used other email things before. I don't keep track of who's subscribed. So I'm not going to sell us to anything. anybody or anything. It's just an announcement thing, and you can unsubscribe anytime. And I don't even know, like once you're unsubscribed, you're gone. I don't have the access to that. So there's people that abuse newsletters and people that I don't think there's a lot of people in the Python tech space that do though.

Starting point is 00:30:00 So it's not spammy. Anyway, moving on, do you have any extras for us? I actually have a couple. Let's jump over and hear about it. Very exciting news. I am really happy with this new course that I created called Agentic, A. AI programming for Python devs and data scientists. Cool.

Starting point is 00:30:18 Yeah, and the idea is basically, so we have these agentic AI tools like code and cursor and Juni and all them. But how do you actually be successful with them, right? I've talked a few times about just insane productivity and really good outcomes that I've had. And people are like, how are you doing that? So this course is like a real world set of guidelines and examples and best practices for working with agentic AI.

Starting point is 00:30:47 So spend an hour building a really detailed application. But we also, it also talks like almost for an hour about like guardrails and road maps and how do you get it to do exactly what you want? You're like, when I say build an app, don't just build the app, but build it with UV, ripe high test tests, format it with Ruff, with this Tommel and you don't have to do any,

Starting point is 00:31:08 you have to, but you know, like how do you get it to give you what you want, not something in the VIG, general space of what you ask for you know what i mean so sort of a a practical agentic a programing course so i really i'm getting like crazy amounts of good fees back on this course so people check it out the link is in the um in the show notes i think it's talk python dotfm dash slash agentic dash a i so that one's fun what else i'm also going to be on talking with hugo bowen anderson he's running the vash vanishing gradients podcast he was on talk python a while ago and now i'm going to be on his show data science meets agentic AI but sort of a follow up from my talk python

Starting point is 00:31:49 and production book plus this course i just talked about those those kind of ideas we're going to spend some time riffing on that so that's tomorrow night u.s time so uh check that out all right um also open ai introduced a i ii browser wrapping google chrome called atlas this is interesting i played with it i'm not convinced you i mean it's kind of fine it's it's i don't know i don't even know what to think about these things. That's not why I included it. Maybe it's sort of interesting as a side note, but I linked to the Ars Technica article. Holy smokes, you guys, check out the comments. They are not loving it. And it's not open AI's fault exactly. It's just this AI, this idea of AI taking over everything is not a, is not, at least if there's not everyone hating on it, there's a very vocal

Starting point is 00:32:35 group of people who don't love it. So, and I just noticed that Kyle Orlin wrote this. And he actually was a whole book on minesweeper which you know props to him for that but oh that's awesome i got to check that out i don't feel like geez it's been a long time to thought about mindsweber anyway this is actually just a really interesting cultural like touch point i think so people can check that out all right james abel um one of the i think he was behind pie bay this year but yeah in san francisco area uh said hey we somehow we were talking about dunder name if dunder name equals dunder main today i have a package it's real real simple but uh instead of writing this if you want to like get people started you can just import my thing and say if is main as a simpler saner way

Starting point is 00:33:19 you should that maybe should be a built-in don't you think python yeah or what i would rather see is if is main i here's what i would like to see but i don't know that python has a a mechanism for it i would like to say at the top something to the effect of register this function as may and then when it's done parsing the file, then it runs it. Because even this, which is super helpful, it's still, if you accidentally put code below it, it's a problem, you know what I mean? I would like Python to enforce like, we load the whole file and then it runs, like a cool mechanism for that.

Starting point is 00:33:52 But still, this is pretty nice. And at first I thought I was just like wrapping if name is main, right? But in fact, it's like doing some a little bit different here. It's using the inspect stack to go back and then look at the value. because obviously it itself is not main, so it can't do that test. So I don't know, kind of cute. Yeah.

Starting point is 00:34:12 If you don't want to have a dependency on this library, you could take that line of code and put it somewhere as a utility class in yours. But, okay, I'm just going to throw this out there. If you want to do something really fast, make sure you time this, because every time I've added inspect, I love the inspect library,

Starting point is 00:34:30 but whenever I added, it slows things down. Okay. Um, just, just, you know, benchmark it. Benchmark it. All right. In line, in line. All right. Um, if you're an IntelliJ person like Pie Charm, wouldn't it be nice, Brian?

Starting point is 00:34:47 Wouldn't it be incredible? If while you're working, a little pet could walk around in the bottom. Anthony Shaw, I believe is the one who added this to Vs code. Yeah. Pie Charm people can love animals too. So you could install the pet into your Pie Charm now. Oh, cool. Nice.

Starting point is 00:35:04 I don't know that I'm going to be doing it. This is the last minute edition. I just got a brand new M5 iPad. How insane is that? I've had, the last one I had I got five years ago. I'm like, ah, it's probably time. And I got a 13-inch iPad Pro, which is an insanely large iPad. But I've started using it as the second monitor for my laptop,

Starting point is 00:35:23 if I'm traveling or if I'm at a coffee shop or something. And that's a super cool experience. Just put them side by side. If they're on the same network or the cable, they just do, you know, real, real low latency. dual monitor stuff and yeah it's pretty neat so just a little shout out to that that's incredible it's like it's like having a new laptop a second a laptop for the price of a second laptop yes it is but here's the thing here yes i agree but it's also a really nice reading device which i do a lot of reading

Starting point is 00:35:50 and i was going to get one of the cheaper ones one of the things that drives me i have a macbook air is my main laptop computer and the same chip from my main computer i don't have a high-end computer and it's plenty good for Docker stuff, for programming stuff, all these things. It's totally fine. But here's the thing. The iPad Air, while great, its peak brightness is something like five or 600 nits. And if you're working, like I like to work in my back outside area in the summer, like that's kind of not relevant right now. But in general, you know, sit outside, enjoy the weather, get out of the office. And if it's at all bright through a window or somewhere, it's like really a pain. This thing is like the best screen you can buy on a Mac, period, whatever.

Starting point is 00:36:31 a thousand nits. So you could push the computer to the side and just put the laptop in front of you and type on it. It's really nice. Anyway, that's the main reason. I want a brighter screen without having a MacBook Pro. I feel so bad for you having to deal with like working outside in such bright light. It's really horrible. I know. You should really, it's hard. It's hard doing me. All right. Carrying on with the jokes. No, go ahead. Well, before we get to the joke, I want to, I guess, to highlight, going back to my announcement, of possibly a book. Japanol 7 says a TDD book by Brian.

Starting point is 00:37:08 Can't wait. Also, Talk Python training courses are great. Kudos. Yeah. I'm looking forward to getting that book at and I'm looking forward to that AI course. Yeah. Yeah, same.

Starting point is 00:37:20 All right. I'm not looking forward to surgery. I'll tell you what. And it's getting to be weird, Brian. I mean, like doctors using AI and stuff. Actually, we probably are going to get better diagnoses. as for certain things with that but here's a surgeon situation there's a person who just they're on post-op okay they're laying there like ah man a little woozy from the anesthesia coming out of it

Starting point is 00:37:42 the the doctor which is a robot with the chat tpt like i don't think it's is it the same i don't know some an ai logo for a robot face and the the patient says but why is the scar on the left if the appendix is on the right the a i surgeon says you're absolutely right let me try that one more time. Please don't try it one more time. It's so bad. It's pretty bad. It's pretty funny.

Starting point is 00:38:10 That actually drives me nuts when I'm like, this doesn't sound right. Is this, you know, the user is showing you out. I've made a mistake. Oh, you're right. Yeah. Oh, well. Yeah. So, yeah, just get a second opinion.

Starting point is 00:38:25 So if opening is going to operate on you, have Anthropic be the best. backup, I guess, is the moral of the store. I don't know. I don't think that's the moral of this world. You don't think so? Okay. Maybe somebody trained in medicine. Yeah. Trained on medicine? I'm sure they are. Trained with heavy medication. Yeah, exactly. All right, cool. Well, fun as always. Well, definitely fun as always. And we'll see everybody next week.

Python Bytes - #455 Gilded Python and Beyond

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.