WEBVTT 00:00:00.020 --> 00:00:05.460 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:05.880 --> 00:00:11.160 This is episode 457, and it's being recorded on November 11th, 2025. 00:00:11.920 --> 00:00:12.700 And I'm Brian Okken. 00:00:12.940 --> 00:00:13.840 And I'm Michael Kennedy. 00:00:14.280 --> 00:00:19.860 Wow, nowhere. Oh, it's 11-11. It's corduroy day. I totally forgot to wear corduroy today. 00:00:20.200 --> 00:00:22.160 I love it. Is that because the vertical stripes? 00:00:22.660 --> 00:00:23.120 Yeah, yeah. 00:00:23.120 --> 00:00:23.420 Vertical grooves. 00:00:23.920 --> 00:00:28.880 Yeah, yeah. I love it. I have no corduroy on either. We missed the opportunity. 00:00:28.980 --> 00:00:32.240 some stripes on your, it's not corduroy, but your sweater. 00:00:32.380 --> 00:00:34.160 Yeah, almost a sweater, almost pulled it off. 00:00:34.800 --> 00:00:40.140 Anyway, we'd like to thank everybody that supports us through going to 00:00:40.610 --> 00:00:43.720 doing Talk Python training or the pytest course, 00:00:44.380 --> 00:00:46.360 Patreon supporters or people that have bought our books. 00:00:47.120 --> 00:00:50.640 We really, really appreciate all of the support and it keeps us going. 00:00:51.000 --> 00:00:57.120 Also love to hear from people on Bluesky, on Mastodon and through email. 00:00:57.400 --> 00:01:01.900 So you can go to the contact us on pythonbytes.fm. 00:01:02.260 --> 00:01:05.580 And there's the links to all the Bluesky Mastermind, 00:01:05.710 --> 00:01:07.020 all those things to get all of us. 00:01:07.640 --> 00:01:10.040 We actually have a couple items today from listeners. 00:01:10.340 --> 00:01:12.940 So feel free to send us things. 00:01:13.700 --> 00:01:15.660 If you'd like to join the show live, 00:01:15.900 --> 00:01:18.400 and even if it's on a weird time like this week, 00:01:19.300 --> 00:01:21.220 you can find out when it's going to happen next 00:01:21.380 --> 00:01:23.800 by going to pythonbytes.fm/live. 00:01:24.820 --> 00:01:28.520 And, you know, jump on and talk with us while we're recording. 00:01:28.680 --> 00:01:28.960 It's fun. 00:01:29.300 --> 00:01:33.760 Also, you don't need to take notes because we're going to send you all the links. 00:01:34.200 --> 00:01:38.320 Well, they're in the show notes, but we'll also send them directly to your inbox if you 00:01:38.480 --> 00:01:44.780 sign up to be a friend of the show and go to pythonbytes.fm and sign up for the newsletter. 00:01:45.220 --> 00:01:46.900 I think that's all of the stuff. 00:01:47.460 --> 00:01:50.160 So let's jump into the first item, Michael. 00:01:50.840 --> 00:01:52.760 Let's talk about it now. 00:01:52.960 --> 00:02:00.940 I want to talk about something called HTTP tap at HTTP tap.dev. Now this one, it's a little bit 00:02:01.260 --> 00:02:08.240 like HTTP pi, HTTP pi, I don't know how you really pronounce that correctly, which is my go to 00:02:08.729 --> 00:02:15.460 anytime I need to test something on the terminal or that sort of thing. HTTP pi is because, you know, 00:02:15.580 --> 00:02:20.240 it's like curl or whatever, but it gives you lots more stats, gives you more information. And this, 00:02:20.380 --> 00:02:21.380 This is a little bit like that. 00:02:21.780 --> 00:02:24.700 If you wanted to go into, say, your Chrome DevTools or something 00:02:25.120 --> 00:02:28.780 and say, "Hey, I'm requesting this page or this URL 00:02:28.920 --> 00:02:29.980 or this API endpoint. 00:02:30.100 --> 00:02:30.880 It's a little bit slow. 00:02:31.380 --> 00:02:32.180 It's a little bit funky. 00:02:32.880 --> 00:02:33.420 What's going on? 00:02:33.460 --> 00:02:34.600 Where's the problem?" 00:02:34.660 --> 00:02:35.600 - Yeah. - If it's getting hung up, 00:02:35.700 --> 00:02:37.220 is it connecting to the server 00:02:37.540 --> 00:02:41.260 or is it the backend app server, like Grinny 00:02:41.260 --> 00:02:42.480 and whoever is not responding? 00:02:42.920 --> 00:02:44.380 So that's what this is all about. 00:02:45.460 --> 00:02:48.720 And it's built, I believe it's built with rich or textual. 00:02:48.860 --> 00:02:52.720 I can't remember how far down the Will McGugan train it goes, 00:02:52.800 --> 00:02:54.440 but it definitely is built on some of those. 00:02:54.740 --> 00:02:55.260 Built with Python, 00:02:56.180 --> 00:02:58.640 although that's not necessarily that important, right? 00:02:58.780 --> 00:03:00.660 It's just I guess it's easier to make improvements to, 00:03:00.780 --> 00:03:03.200 but it's more just a tool that anyone doing web stuff 00:03:03.820 --> 00:03:05.820 with Python might be interested in. 00:03:06.020 --> 00:03:09.520 And I pulled up a little example of just asking Talk Python, 00:03:10.040 --> 00:03:11.680 acptptap, then talkpython.fm, 00:03:12.080 --> 00:03:13.240 and it says, here's the analysis. 00:03:13.840 --> 00:03:15.120 And apparently there's only one step, 00:03:15.200 --> 00:03:16.560 maybe if there were redirects or something, 00:03:16.940 --> 00:03:21.980 It would show multi-stage flows, but there's no redirects, so there's not. 00:03:22.330 --> 00:03:25.380 It shows you the IP address, the TLS version, the cipher, 00:03:25.980 --> 00:03:29.700 when the certificate expires, the status, the size, the server. 00:03:30.190 --> 00:03:32.080 It gets all of that stuff right, which is pretty cool. 00:03:33.100 --> 00:03:35.940 Then it has this waterfall-tiered sort of diagram. 00:03:36.290 --> 00:03:37.980 It says the request timeline. 00:03:38.160 --> 00:03:42.680 It took 29 milliseconds for DNS to figure out where the heck that was. 00:03:42.960 --> 00:03:45.080 I've been on vacation. I haven't been going to the website. 00:03:45.440 --> 00:03:47.000 So I had to go get it from the internet. 00:03:47.380 --> 00:03:49.920 And then it took 80 milliseconds to connect to the server. 00:03:50.360 --> 00:03:52.940 By the way, it's about 80 milliseconds ping time. 00:03:53.100 --> 00:03:55.380 So you got to keep that in mind when all these numbers, right? 00:03:55.480 --> 00:03:59.200 Like that's East Coast, West Coast, US, 2,500 miles. 00:03:59.680 --> 00:04:01.820 So connect 80 milliseconds. 00:04:02.300 --> 00:04:04.580 TLS took 130 milliseconds. 00:04:05.060 --> 00:04:08.680 So if you ever see your browser doing like keep alive type of stuff, 00:04:09.120 --> 00:04:10.640 that's a big savings right there. 00:04:10.680 --> 00:04:12.560 That's almost 200 milliseconds savings 00:04:13.060 --> 00:04:14.960 if it doesn't have to keep doing that over and over, right? 00:04:15.040 --> 00:04:15.580 So that's pretty nice. 00:04:15.870 --> 00:04:21.060 And then you can see it's actually sending the request and waiting for the response. 00:04:21.840 --> 00:04:24.880 So it goes out, comes back, takes 170 milliseconds. 00:04:25.420 --> 00:04:27.320 Again, ping time in there as well. 00:04:27.510 --> 00:04:30.720 And then the transfer is a whopping 4.3 milliseconds. 00:04:31.020 --> 00:04:32.100 So this is pretty cool. 00:04:32.240 --> 00:04:34.280 A really quick view of just seeing what's going on, right, Brian? 00:04:34.660 --> 00:04:36.940 Yeah, actually, I love that. 00:04:37.920 --> 00:04:44.840 Very useful, especially if you've got some lag issues or some part of your website that's a little slow. 00:04:44.880 --> 00:04:46.300 you can look at what's going on. 00:04:46.670 --> 00:04:47.720 Is it DNS? 00:04:48.220 --> 00:04:49.560 Is it the transfer? 00:04:50.300 --> 00:04:51.480 Where's this delay coming from? 00:04:51.880 --> 00:04:53.400 I mean, it's not super fine-grained, 00:04:53.540 --> 00:04:54.480 but this is pretty good. 00:04:54.640 --> 00:04:56.000 Yeah, well, there's just, yeah, for sure. 00:04:56.070 --> 00:04:57.460 And there's a lot of stuff involved, right? 00:04:57.660 --> 00:04:58.620 There's, like you said, 00:04:58.770 --> 00:05:00.160 the DNS might be taking forever, 00:05:00.290 --> 00:05:01.780 which could be part of the problem, 00:05:02.040 --> 00:05:02.860 or who knows, right? 00:05:02.980 --> 00:05:03.820 There's just a lot of things. 00:05:04.000 --> 00:05:08.740 So really neat to see this little UI built in Python. 00:05:09.219 --> 00:05:12.280 Yeah, and I guess we have to say it. 00:05:12.500 --> 00:05:13.080 It's always DNS. 00:05:13.740 --> 00:05:14.240 It's always DNS. 00:05:14.360 --> 00:05:15.080 I mean, that's... 00:05:16.080 --> 00:05:17.900 It's not always DNS, but yeah. 00:05:18.780 --> 00:05:18.880 Anyway. 00:05:19.260 --> 00:05:20.020 Yeah, but often. 00:05:20.960 --> 00:05:21.240 Often. 00:05:21.510 --> 00:05:22.940 No, I think this is great. 00:05:23.300 --> 00:05:23.800 This is cool. 00:05:24.340 --> 00:05:26.520 I've got actually some side projects 00:05:26.610 --> 00:05:30.100 that I've got some weird speed issues, 00:05:30.230 --> 00:05:31.180 and I don't know why. 00:05:31.540 --> 00:05:34.160 And even just looking at some of this stuff, 00:05:34.300 --> 00:05:36.200 help out to figure out where to start debugging. 00:05:36.680 --> 00:05:38.600 Yeah, you can get some of these types of things 00:05:39.020 --> 00:05:40.880 in the dev tools and browsers, 00:05:41.120 --> 00:05:42.020 but this is really nice. 00:05:42.260 --> 00:05:43.740 It's visual, it's colorful, it's graphical. 00:05:44.240 --> 00:05:44.480 I love it. 00:05:44.480 --> 00:05:44.800 I like it. 00:05:44.960 --> 00:05:45.660 And it's command line. 00:05:46.080 --> 00:05:46.960 Yay, we love command line. 00:05:47.140 --> 00:05:47.440 Yay. 00:05:48.340 --> 00:05:48.500 Okay. 00:05:50.660 --> 00:05:55.120 I'm not usually, I have to admit, I click on listicles all the time. 00:05:55.220 --> 00:05:59.840 But we don't cover very many listicles on the show, but I'm going to cover one today. 00:06:01.100 --> 00:06:03.820 I actually, oh, my aspect ratio is off. 00:06:04.700 --> 00:06:05.220 There we go. 00:06:05.440 --> 00:06:05.740 Here we go. 00:06:07.000 --> 00:06:08.120 This is from PyCharm. 00:06:08.380 --> 00:06:11.119 It's actually from Evgenia Verbena. 00:06:12.040 --> 00:06:12.620 That's a cool name. 00:06:13.400 --> 00:06:19.240 10 smart performance hacks for faster Python code. 00:06:19.560 --> 00:06:25.940 And there's some that, well, actually, I think stuff like this is really good to talk to people about 00:06:26.060 --> 00:06:28.460 because we don't really think about performance too much. 00:06:28.780 --> 00:06:31.880 I mean, when you're starting Python, you don't think about it too much. 00:06:32.060 --> 00:06:34.460 It's hopefully, you know, hopefully you don't. 00:06:34.580 --> 00:06:37.520 There's, we don't want people to prematurely optimize. 00:06:37.960 --> 00:06:44.020 But there's just some habits that maybe are decent to get into that avoid problems later on. 00:06:44.210 --> 00:06:46.940 And so that's one of the reasons why I like going through this. 00:06:48.320 --> 00:06:53.020 Oh, nice menu on the right side, also with all the hacks on the right. 00:06:53.360 --> 00:07:00.840 Anyway, the first one, not too surprising if you've done through any leak code stuff or whatever, 00:07:00.990 --> 00:07:04.020 is leverage set for membership testing. 00:07:04.600 --> 00:07:08.580 So if you've got a big list, you can throw the list into a set, 00:07:08.780 --> 00:07:13.960 and then you can find out easily if something's in the list by putting it in a set. 00:07:14.839 --> 00:07:19.160 And oddly, that speeds things up quite a bit, like drastically. 00:07:20.240 --> 00:07:23.140 List lookup versus set lookup is super fast. 00:07:24.500 --> 00:07:25.800 Same predictionaries, too. 00:07:26.040 --> 00:07:29.720 If you're looking up by some kind of key, either one, it's ridiculous. 00:07:30.260 --> 00:07:30.440 Yeah. 00:07:32.000 --> 00:07:34.080 Like I said, some not too surprising. 00:07:34.420 --> 00:07:38.500 Slots are great if you're going to use classes or data classes even. 00:07:39.740 --> 00:07:43.940 If it's not going to grow over time, you can define slots and stuff. 00:07:44.900 --> 00:07:50.320 And then the lookup for each element within a data class is a lot faster with slots. 00:07:52.099 --> 00:07:55.900 I want to jump around because there were some that I didn't really realize 00:07:56.680 --> 00:07:58.680 or I just didn't think about before. 00:07:59.140 --> 00:08:01.680 And one of them is use math functions instead of operators. 00:08:01.760 --> 00:08:03.760 and I guess I wouldn't 00:08:04.080 --> 00:08:05.020 some of these are weird 00:08:06.080 --> 00:08:08.380 so math.square 00:08:08.820 --> 00:08:10.240 sqrt for square root 00:08:10.620 --> 00:08:12.160 is faster and more accurate 00:08:12.440 --> 00:08:14.260 than raising a number to the power of 00:08:14.700 --> 00:08:16.060 0.5 using 00:08:16.500 --> 00:08:17.660 exponentation operator 00:08:17.900 --> 00:08:19.860 I wouldn't do that anyway because that's weird 00:08:20.100 --> 00:08:21.280 it just looks weird 00:08:22.199 --> 00:08:23.900 I don't know, bad code style 00:08:24.420 --> 00:08:25.640 but there's a whole bunch of others 00:08:25.960 --> 00:08:27.600 similarly the sine function exponent 00:08:28.200 --> 00:08:30.239 log, those are really fast operators 00:08:30.260 --> 00:08:36.520 in math so go ahead and utilize math and uh so that's a good pointer pre-allocate memory 00:08:36.840 --> 00:08:41.780 with known sizes now this is some like an old trick that i think that i didn't really think 00:08:41.820 --> 00:08:45.780 that it was a huge difference and it looks like one of the things that's it's cool in this article 00:08:45.860 --> 00:08:51.000 it does the pre-allocated time versus or the uh before and after times so there's a bit of 00:08:51.320 --> 00:08:59.719 you know uh 0.01 second speed up um but so the trick is just to uh i guess have an array with 00:08:59.740 --> 00:09:01.800 If you're going to fill an array anyway, 00:09:02.000 --> 00:09:05.260 may as well just fill it with some zeros or something initially. 00:09:05.900 --> 00:09:09.460 And then it seems like a waste of time to fill it with zeros 00:09:09.760 --> 00:09:10.960 because you're going to fill everything up. 00:09:11.220 --> 00:09:14.160 But it isn't because it's the allocation that takes a little bit longer 00:09:14.360 --> 00:09:15.860 than the filling it with zeros. 00:09:16.880 --> 00:09:20.380 Yeah, and Brian, it's worth thinking through how that goes a little bit 00:09:20.560 --> 00:09:22.680 because if you just go append, append, append, 00:09:22.940 --> 00:09:25.300 and this is something I don't do and it's inspiring me to consider, 00:09:25.460 --> 00:09:26.480 well, maybe I should. 00:09:26.840 --> 00:09:30.140 But the array doesn't know if you're going to put a million items in there. 00:09:30.390 --> 00:09:34.880 Every time you do like a array.append or list.append, it doesn't go, guess what? 00:09:35.040 --> 00:09:35.900 We're going to add a million, right? 00:09:35.990 --> 00:09:37.180 It's like, well, what about 16? 00:09:37.960 --> 00:09:39.020 What about 32 of them? 00:09:39.300 --> 00:09:40.040 Oh, now we need more. 00:09:40.110 --> 00:09:41.360 Okay, what about 64 of them? 00:09:41.480 --> 00:09:45.760 You know, it just, it has to keep reallocating and copying as you build this up. 00:09:45.900 --> 00:09:46.400 So, yeah. 00:09:46.520 --> 00:09:49.380 And I don't remember how, I mean, it's a little smarter than that. 00:09:49.530 --> 00:09:52.680 So I think it like does, like doubles it in size or takes a chunk. 00:09:53.160 --> 00:09:54.360 I can't remember what it does. 00:09:54.380 --> 00:09:56.100 Yeah, it's not allocating every one, right? 00:09:56.140 --> 00:09:56.640 It's got an awkward. 00:09:56.740 --> 00:09:56.780 Right. 00:09:57.080 --> 00:09:57.560 Or heuristic. 00:09:57.900 --> 00:10:00.460 But if you already know how big it's going to be, 00:10:00.770 --> 00:10:02.960 you can just tell it to do it once, which is cool. 00:10:03.110 --> 00:10:05.280 But I don't really think about that too much. 00:10:06.640 --> 00:10:09.360 A couple of the ones down below I really wanted to cover. 00:10:09.600 --> 00:10:13.240 This was avoid exception handling and hot loops. 00:10:13.780 --> 00:10:15.360 And I am guilty of this too, 00:10:15.490 --> 00:10:17.580 because I try to have my exception handling 00:10:17.830 --> 00:10:20.960 as close to the place where it can go wrong as possible. 00:10:21.460 --> 00:10:23.300 Just as it's a visual thing 00:10:23.860 --> 00:10:26.180 and a code readability thing to have this tight. 00:10:26.980 --> 00:10:28.520 have the try right around it. 00:10:28.680 --> 00:10:31.120 Like the example has a zero division error, 00:10:31.520 --> 00:10:33.380 has the try and accept right there. 00:10:34.079 --> 00:10:37.600 However, that might slow down a little bit. 00:10:38.480 --> 00:10:40.560 And it did, in this case, 00:10:40.820 --> 00:10:42.000 was the try and accept on this one. 00:10:42.240 --> 00:10:42.880 It's just not there. 00:10:43.100 --> 00:10:44.760 I think the top one versus the bottom one. 00:10:44.980 --> 00:10:46.600 Yeah, but there's no try and accept at the top. 00:10:46.860 --> 00:10:47.920 That's the baseline. 00:10:48.360 --> 00:10:48.760 Oh, okay. 00:10:48.860 --> 00:10:50.620 The conditional check versus. 00:10:51.280 --> 00:10:52.700 It's like checking for the condition 00:10:52.840 --> 00:10:55.200 that would cause the exception to avoid the exception. 00:10:55.760 --> 00:10:56.720 Oh, OK. 00:10:57.070 --> 00:10:59.580 I guess I would just put the try and accept at the top 00:11:00.330 --> 00:11:03.020 or around the larger one as a better comparison. 00:11:03.360 --> 00:11:03.560 Anyway. 00:11:04.300 --> 00:11:05.100 Yeah, I agree. 00:11:05.300 --> 00:11:06.000 But this is a big one. 00:11:06.320 --> 00:11:07.680 But that's a big one. 00:11:08.500 --> 00:11:12.720 Then we get into some really, except there's exception handling, 00:11:13.620 --> 00:11:16.360 using iter tools for combinatorial operations. 00:11:16.700 --> 00:11:18.100 And this is huge. 00:11:18.650 --> 00:11:23.480 If you're working with chunks of lists or chunks of stuff, 00:11:23.860 --> 00:11:26.580 using iter tools and knowing what's in there, 00:11:26.910 --> 00:11:28.660 it speeds up a huge amount. 00:11:29.030 --> 00:11:32.120 So using iter tools versus loops for, 00:11:32.510 --> 00:11:34.000 I forget what the example is here, 00:11:34.360 --> 00:11:36.900 but the speed up is using product 00:11:37.500 --> 00:11:40.920 and the speed up is massive from, you know, 00:11:40.930 --> 00:11:43.480 like what, 440 times bigger or something? 00:11:44.140 --> 00:11:46.840 Or 4 divided by, oh, 5 divided by 20. 00:11:47.140 --> 00:11:47.440 That's 4. 00:11:47.710 --> 00:11:48.820 So it's 4 times speed up. 00:11:49.860 --> 00:11:50.500 Math is hard. 00:11:51.460 --> 00:11:54.240 BISect is also a great thing to look at. 00:11:54.760 --> 00:12:01.040 And yeah, anyway, just some cool, wow, that's a massive speedup, 100 times speedup for using 00:12:01.240 --> 00:12:02.440 BISect in the right places. 00:12:03.000 --> 00:12:08.080 So finding out places where you can use BISect or iter tools and just knowing that those 00:12:08.140 --> 00:12:10.200 are there to look them up, those are good speedups. 00:12:10.520 --> 00:12:13.960 Anyway, occasionally it's good to think about how fast is my code going? 00:12:14.220 --> 00:12:21.420 You know, I think we could summarize up that article by basically saying, know the standard 00:12:21.800 --> 00:12:23.520 know your data structures, mostly. 00:12:24.080 --> 00:12:24.320 Yeah. 00:12:24.700 --> 00:12:26.980 Use DICT, use Zitter Tools, et cetera. 00:12:27.779 --> 00:12:30.300 And the standard library is still pretty big. 00:12:30.400 --> 00:12:32.540 I would probably limit that to the standard library 00:12:32.760 --> 00:12:35.180 that manipulates data structures. 00:12:35.920 --> 00:12:37.560 Yeah, it kind of intersects with data structures. 00:12:37.640 --> 00:12:38.960 Yeah, I guess that's kind of what I was getting at. 00:12:39.280 --> 00:12:42.220 And by the way, this was written by Dito Grigorov 00:12:42.860 --> 00:12:44.340 and posted by Uvigena. 00:12:44.880 --> 00:12:45.560 Oh, okay. 00:12:45.640 --> 00:12:48.400 Just want to make sure Dito gets a little credit for writing it. 00:12:48.580 --> 00:12:49.760 Thanks for that clarification. 00:12:50.120 --> 00:12:50.780 Yeah, no worries. 00:12:51.400 --> 00:12:53.040 - It's tricky, I think it's 'cause the way 00:12:53.200 --> 00:12:55.060 the JetBrains CMS works. 00:12:55.510 --> 00:12:56.780 Yeah, so cool, cool. 00:12:57.000 --> 00:12:59.660 Well, maybe we should, Brian, 00:12:59.770 --> 00:13:00.800 maybe we should talk to him about it. 00:13:00.980 --> 00:13:02.780 Put him up on a little video call, you know what I mean? 00:13:03.060 --> 00:13:03.460 What do you think? 00:13:03.570 --> 00:13:04.520 Do a little WebRTC? 00:13:05.240 --> 00:13:06.580 - I've kind of forgot about RTC. 00:13:07.140 --> 00:13:09.300 - I know, but you and I are sitting here, 00:13:09.480 --> 00:13:11.660 people don't necessarily know how we're doing this. 00:13:12.480 --> 00:13:15.120 It's magic, we basically, you and I, we speak through magic, 00:13:15.660 --> 00:13:18.840 but some of that magic is WebRTC 00:13:18.860 --> 00:13:22.800 because all of this is recording, streaming, et cetera, 00:13:23.450 --> 00:13:26.260 video and voice, multiple channels, multiple screens, 00:13:26.260 --> 00:13:30.320 and so on over WebRTC in the browser, which is ridiculous. 00:13:30.960 --> 00:13:32.780 It's just ridiculous that this works. 00:13:33.220 --> 00:13:36.480 So I want to talk about a library from Freddie Bolton, 00:13:36.980 --> 00:13:39.760 who is behind the Gradio project, Gradio, Gradio, 00:13:39.880 --> 00:13:43.080 Gradio project, and I had him on Talk Python 00:13:43.280 --> 00:13:46.140 a couple of years ago, and this is called FastRTC. 00:13:46.760 --> 00:13:50.180 So kind of like, hey, I'm a fan of integrating with FastAPI, 00:13:50.520 --> 00:13:51.760 and I want to do WebRTC. 00:13:52.480 --> 00:13:55.540 It's kind of ridiculous how simple some of these things are. 00:13:55.600 --> 00:13:56.340 So it's pretty popular. 00:13:56.460 --> 00:13:58.900 It's got 4,000 GitHub stars, 400 forks. 00:13:59.280 --> 00:14:00.660 It's been around for a while. 00:14:01.020 --> 00:14:03.900 Freddie pointed this out to me a decent while ago, 00:14:04.020 --> 00:14:05.460 and I finally ain't getting around to talking about it. 00:14:05.800 --> 00:14:06.820 But it's quite neat. 00:14:07.060 --> 00:14:08.800 There's a little video you can watch about how to do it, 00:14:08.920 --> 00:14:10.580 but it's got some examples for you. 00:14:10.720 --> 00:14:13.240 Like if you want to do an Echo Audio, 00:14:13.340 --> 00:14:17.440 which is kind of a sadistic example, you know what I mean? 00:14:17.640 --> 00:14:19.940 Because that's usually the worst thing about a video meeting, 00:14:20.180 --> 00:14:20.700 is the echo. 00:14:21.380 --> 00:14:23.120 But if you want to listen to yourself echo back 00:14:23.220 --> 00:14:26.200 just to prove that it's going round and round, 00:14:26.700 --> 00:14:28.580 then how many lines is that? 00:14:28.640 --> 00:14:30.620 I mean, there's a huge comma in there, 00:14:30.780 --> 00:14:32.440 but if you take out the comma 00:14:33.020 --> 00:14:34.520 and you format it a little bit wider, 00:14:34.660 --> 00:14:37.340 it's probably five lines of code to set up a server, 00:14:38.120 --> 00:14:39.660 connect to it, and echo back. 00:14:39.780 --> 00:14:41.300 Like, that's insane. 00:14:41.779 --> 00:14:42.180 Insane. 00:14:42.240 --> 00:14:46.660 And there's also ways to set up LLM voice chat if you care about that. 00:14:46.960 --> 00:14:52.260 You know, that's obviously the thing du jour is I want to talk to my LLM. 00:14:52.440 --> 00:14:55.480 But, you know, you could do that pretty easily, pretty quickly here. 00:14:55.800 --> 00:14:59.360 But also just stream your webcam back or object detection. 00:14:59.360 --> 00:15:00.460 It's just like a handful of lines of code. 00:15:00.740 --> 00:15:01.480 It's insane. 00:15:01.720 --> 00:15:04.020 But this stuff seems so complicated, doesn't it, Brian? 00:15:04.740 --> 00:15:05.040 It's cool. 00:15:05.200 --> 00:15:05.520 It's cool. 00:15:05.640 --> 00:15:07.580 You're wrapping stuff up in a library. 00:15:08.460 --> 00:15:08.560 Yeah. 00:15:08.680 --> 00:15:13.440 So it comes with automatic voice detection, turn-taking built-in, which is interesting. 00:15:14.360 --> 00:15:20.420 It's got a UI to launch it in built-in Gradio UI, if you want to do that. 00:15:20.450 --> 00:15:22.000 I believe that's probably JavaScript. 00:15:22.920 --> 00:15:24.660 Automatic Web RTC support. 00:15:25.360 --> 00:15:30.820 So mount it onto a FastAPI app and get a Web RTC endpoint for your own front end. 00:15:31.050 --> 00:15:33.640 So you can kind of go Gradio or you can go this way. 00:15:33.940 --> 00:15:34.940 It's got WebSocket support. 00:15:35.450 --> 00:15:35.860 Listen to this. 00:15:36.120 --> 00:15:36.960 Automatic telephone support. 00:15:37.320 --> 00:15:41.840 use FastPhone on the string and get your own temporary phone number. 00:15:42.600 --> 00:15:43.720 Actually, that's pretty cool. 00:15:44.480 --> 00:15:44.580 Yeah. 00:15:45.300 --> 00:15:47.860 So, yeah, there's a lot of stuff going on here. 00:15:48.020 --> 00:15:50.640 It says you can easily be mounted on your own FastAPI app 00:15:50.640 --> 00:15:52.100 and customize it entirely, 00:15:52.680 --> 00:15:55.180 extend it to fit within your production app, 00:15:55.500 --> 00:15:57.080 I guess, assuming that's a FastAPI app. 00:15:57.340 --> 00:16:00.460 Although you could certainly squeeze it in using, like, 00:16:00.720 --> 00:16:02.680 Nginx routes to make it look like one app, 00:16:02.780 --> 00:16:05.320 but it's not exactly the same back end and so on. 00:16:05.420 --> 00:16:06.240 So there's examples for that. 00:16:06.500 --> 00:16:06.660 Wow. 00:16:06.960 --> 00:16:08.520 Yeah, so I thought it was pretty cool. 00:16:09.460 --> 00:16:10.100 That's pretty neat. 00:16:10.350 --> 00:16:15.040 I interviewed somebody in like three years ago, 2012, about WebRTC. 00:16:15.410 --> 00:16:17.240 They wrote a book on it. 00:16:17.680 --> 00:16:17.840 Nice. 00:16:19.020 --> 00:16:20.980 I thought, like, I don't think I'd ever use that. 00:16:21.220 --> 00:16:24.060 But of course I use tools that use WebRTC all the time. 00:16:24.540 --> 00:16:26.300 So like you said, we're using it for. 00:16:26.520 --> 00:16:27.460 We're using it right now. 00:16:27.880 --> 00:16:28.520 We're using it right now. 00:16:29.100 --> 00:16:30.880 And yeah, it's one of those things where it's like, 00:16:30.940 --> 00:16:32.220 I don't think I'll ever build one of those. 00:16:33.070 --> 00:16:36.280 And then you're like, oh, but what if I could do it for this project 00:16:36.300 --> 00:16:39.160 or something comes up or, you know, just knowing about it as a... 00:16:39.160 --> 00:16:39.660 I think it's... 00:16:40.110 --> 00:16:47.840 Yeah, and right now, like you said, with things like hooking up LLMs to, you know, some vibe coding. 00:16:48.660 --> 00:16:49.340 Yeah, exactly. 00:16:49.960 --> 00:16:51.320 I mean, no comment. 00:16:51.920 --> 00:16:52.740 No comment. 00:16:54.000 --> 00:16:56.540 Okay, so this is a listener suggestion, actually. 00:16:57.860 --> 00:17:01.100 Comes to us as a request from Nicholas Kersner. 00:17:01.500 --> 00:17:05.920 He said, hey, I'm using PipDepTree a lot lately. 00:17:06.180 --> 00:17:07.860 And it's pretty cool if you covered that. 00:17:07.930 --> 00:17:09.180 And I'm like, yeah, we covered it. 00:17:09.500 --> 00:17:10.560 But I looked it up. 00:17:11.260 --> 00:17:13.540 It was, I should have had the episode up. 00:17:13.819 --> 00:17:16.520 But we covered it in episode 17. 00:17:17.660 --> 00:17:19.040 That was what? 00:17:20.140 --> 00:17:20.439 2017. 00:17:21.260 --> 00:17:23.800 So yeah, it was a minute ago that we've covered it. 00:17:23.939 --> 00:17:25.760 So it's totally fine to cover it again 00:17:26.160 --> 00:17:27.100 because there's been changes. 00:17:27.420 --> 00:17:30.720 So I did want to talk about both pip Deptree, 00:17:31.080 --> 00:17:34.300 but also I have been, I used pip Deptree for a long time, 00:17:34.480 --> 00:17:36.320 But now usually I use uv pip tree. 00:17:36.740 --> 00:17:39.520 And so I just decided to write an article on it just to, 00:17:40.120 --> 00:17:41.540 because I wanted to show some code examples. 00:17:42.700 --> 00:17:46.040 One of the cool things I did not know that pip dep tree did. 00:17:46.120 --> 00:17:51.080 One of the things that I would always do is I'd have a project install pip dep tree. 00:17:51.100 --> 00:17:52.440 And what it does is when you run it, 00:17:52.700 --> 00:17:56.240 it looks at all the stuff you have installed and then shows, shows them. 00:17:56.580 --> 00:17:59.220 It's kind of like pip list, but it shows all the, 00:17:59.720 --> 00:18:04.440 it's like a tree structure and it shows them we're like in a tree structure with 00:18:04.460 --> 00:18:10.320 versions that are installed and it kind of it's extra information like the uh like if i'm using 00:18:10.320 --> 00:18:16.440 it from pyproject.toml i have an example of the cards project why it chose the particular version 00:18:16.580 --> 00:18:23.380 that's installed um so it's pretty handy but the i would always install it in the project and it 00:18:23.460 --> 00:18:29.059