WEBVTT 00:00:00.020 --> 00:00:03.280 For years, Django and Async came with an asterisk. 00:00:04.000 --> 00:00:06.160 The docs themselves warned you off it. 00:00:06.580 --> 00:00:07.580 Scary performance notes. 00:00:08.180 --> 00:00:09.800 A story that felt half-finished. 00:00:10.120 --> 00:00:12.560 Well, that story just got rewritten, literally. 00:00:12.860 --> 00:00:16.540 And the person who rewrote it is here to tell you why that old framing was wrong. 00:00:17.140 --> 00:00:21.860 Carlton Gibson is a former Django fellow, sat on the security team for eight years, 00:00:22.180 --> 00:00:23.300 and he's on the steering council. 00:00:24.080 --> 00:00:26.960 On this episode, we get into the Async topic doc rewrite. 00:00:27.020 --> 00:00:33.000 what actually remains versus what was just fear and the new tasks framework in 6.0, 00:00:33.560 --> 00:00:36.140 DB-level cascades and fetch models landing in 6.1, 00:00:36.340 --> 00:00:40.160 and why free threading is the bet it's about to pay off big for Django. 00:00:41.260 --> 00:00:43.960 If you've been told that Django's async story isn't ready, 00:00:44.580 --> 00:00:46.160 this episode puts that myth to bed. 00:00:46.760 --> 00:00:51.840 This is Talk Python To Me, episode 556, recorded July 2nd, 2026. 00:01:09.720 --> 00:01:14.040 Welcome to Talk Python To Me, the number one Python podcast for developers and data scientists. 00:01:14.590 --> 00:01:19.880 This is your host, Michael Kennedy. I'm a PSF fellow who's been coding for over 25 years. 00:01:20.540 --> 00:01:21.580 Let's connect on social media. 00:01:22.020 --> 00:01:25.040 You'll find me and Talk Python on Mastodon, Bluesky, and X. 00:01:25.400 --> 00:01:27.220 The social links are all in your show notes. 00:01:27.980 --> 00:01:31.500 You can find over 10 years of past episodes at talkpython.fm. 00:01:31.680 --> 00:01:34.880 And if you want to be part of the show, you can join our recording live streams. 00:01:35.260 --> 00:01:35.700 That's right. 00:01:35.960 --> 00:01:39.160 We live stream the raw uncut version of each episode on YouTube. 00:01:39.760 --> 00:01:44.180 Just visit talkpython.fm/youtube to see the schedule of upcoming events. 00:01:44.370 --> 00:01:48.060 Be sure to subscribe there and press the bell so you'll get notified anytime we're recording. 00:01:48.760 --> 00:01:50.420 This episode is brought to you by Sentry. 00:01:50.640 --> 00:01:54.880 Don't let those errors go unnoticed. Use Sentry like we do here at Talk Python. Sign up at 00:01:55.040 --> 00:02:01.020 Talk Python.fm slash Sentry. Carlton, welcome back to Talk Python To Me. Amazing to have you here. 00:02:01.360 --> 00:02:02.840 Thank you, Michael. Thank you for having me on again. 00:02:03.320 --> 00:02:08.679 Always fun to talk with you. Fun to have you back. We're going to talk about one of my favorite 00:02:08.840 --> 00:02:15.260 topics, async, both for the reasons it's amazing and the reasons it's like, why? Why is it this way? 00:02:15.740 --> 00:02:18.320 Why do we do this to ourselves is the thing, right? 00:02:18.960 --> 00:02:19.220 Exactly. 00:02:20.480 --> 00:02:22.620 Wait, I want to say, oh, but I didn't want it like that. 00:02:22.740 --> 00:02:23.280 No, just kidding. 00:02:24.440 --> 00:02:29.000 Specifically with regard to Django, you did some really interesting updates with Django 00:02:29.180 --> 00:02:30.160 that we're going to talk about. 00:02:30.460 --> 00:02:33.360 But first, your teams, they're still in the World Cup. 00:02:33.460 --> 00:02:33.920 How are we doing? 00:02:34.220 --> 00:02:34.720 Oh, yeah. 00:02:34.760 --> 00:02:37.280 Well, so I'm British by birth, so English by birth. 00:02:37.440 --> 00:02:39.860 So I was supposed to support England, but I live in Spain. 00:02:40.080 --> 00:02:43.900 So depending on who's winning, who's doing better, I get to flip my allegiance. 00:02:44.900 --> 00:02:50.160 you know i think that's really that's really amazing because you if one of your team's not 00:02:50.220 --> 00:02:54.820 doing well you can just like i'm gonna just cheer for the other one it's it's great you get like two 00:02:54.980 --> 00:02:59.340 shots at the thing yeah i mean in the last uh european championships we had the england spain 00:02:59.520 --> 00:03:06.400 final so i couldn't lose really yeah that's amazing yeah so the u.s team um advanced yesterday as well 00:03:06.430 --> 00:03:10.979 they played really well so i'm very very excited for that yeah it's lovely to see new new nation 00:03:11.000 --> 00:03:12.720 emerging nations do well in the World Cup. 00:03:12.980 --> 00:03:15.300 You know, a long time ago, that would have been like some kind of joke. 00:03:15.410 --> 00:03:17.280 People are like, ah, surely you're joking. 00:03:17.780 --> 00:03:18.980 Surely you're joking about this. 00:03:19.120 --> 00:03:24.700 But honestly, soccer or football, non-American football, is popular in America. 00:03:25.000 --> 00:03:25.740 So super cool. 00:03:26.140 --> 00:03:27.020 Glad to hear it. 00:03:27.240 --> 00:03:28.680 And you're back. 00:03:29.260 --> 00:03:32.040 So tell people just what have you been up to? 00:03:32.440 --> 00:03:34.360 You know, not everyone listens every episode. 00:03:34.600 --> 00:03:35.960 So a quick introduction for yourself, maybe. 00:03:36.240 --> 00:03:36.440 Okay. 00:03:36.510 --> 00:03:37.680 So my name is Carlton. 00:03:38.300 --> 00:03:39.780 I've long time worked with Django. 00:03:40.360 --> 00:03:42.760 I'm a former Django fellow, did that for five years. 00:03:42.960 --> 00:03:44.440 I'm on the Django steering council. 00:03:44.570 --> 00:03:46.260 I was on the security team for eight years. 00:03:46.370 --> 00:03:48.480 I maintain a whole load of packages in the ecosystem. 00:03:48.740 --> 00:03:50.540 So, you know, I'm quite vested there. 00:03:51.420 --> 00:04:00.040 And, yes, you know, basically I use Django and have done for, I don't know, however many years to build web applications to pay the bills. 00:04:01.240 --> 00:04:02.100 Had a busy spring. 00:04:02.610 --> 00:04:04.360 I was everywhere I went to. 00:04:04.670 --> 00:04:09.060 There was a Pi TV event organized by JetBrains in Amsterdam in March that I went to. 00:04:09.620 --> 00:04:10.540 That was all on YouTube. 00:04:10.810 --> 00:04:12.960 There was, I don't know, 11 hours of live streaming or something. 00:04:13.080 --> 00:04:13.840 It was amazing. 00:04:14.700 --> 00:04:18.859 Then I was in DjangoCon Europe in Athens in, I think, May. 00:04:19.350 --> 00:04:19.959 No, April. 00:04:20.340 --> 00:04:23.260 And then in May, I was at Paikon Italia, which is in Bologna. 00:04:23.720 --> 00:04:25.980 Paikon Italia really is a lovely conference. 00:04:26.130 --> 00:04:27.680 You should make your way there. 00:04:27.840 --> 00:04:30.480 And now I'm just home and I'm exhausted. 00:04:30.600 --> 00:04:33.860 I had a quick holiday with the family, but now I'm hiding at home until September, 00:04:34.110 --> 00:04:36.820 where we've got Django on the Med, which is a three-day sprint event, 00:04:37.080 --> 00:04:40.040 which is going to be in Pescara, Italy, which we're looking forward to. 00:04:40.340 --> 00:04:43.000 That sounds really, really nice. 00:04:43.160 --> 00:04:47.600 I love traveling, but the jet lag kind of blows me out eventually. 00:04:47.800 --> 00:04:50.200 I'm like, ah, this is 11 hours jet lag. 00:04:50.240 --> 00:04:51.620 I don't really want to do this anymore. 00:04:52.040 --> 00:04:56.960 It sounds to me like you're kind of able to keep it in the broader Euro neighborhood, which is nice. 00:04:57.580 --> 00:04:58.060 Yes, I was. 00:04:58.240 --> 00:05:01.400 I was, which it does help because there isn't that massive time change. 00:05:01.740 --> 00:05:03.220 But it's still hard work, right? 00:05:03.840 --> 00:05:06.800 Even though you have fun at these events, even though you're seeing your friends, 00:05:06.960 --> 00:05:09.400 even though it's, you know, doing what you're vested in. 00:05:09.620 --> 00:05:14.860 It's their long days and you have to be full on, you know, in action all day, each day. 00:05:15.150 --> 00:05:17.840 And so by the time you get home, you are, you know, kind of exhausted. 00:05:17.890 --> 00:05:22.320 And then obviously you've been away and you get home and, you know, you need to be present for your family too. 00:05:22.450 --> 00:05:24.800 So there's a limit to how many of these I can do. 00:05:24.850 --> 00:05:28.540 So I won't do any other events this year because it's just too much. 00:05:28.540 --> 00:05:28.600 Yeah, that's good. 00:05:28.880 --> 00:05:29.460 Just too much. 00:05:29.640 --> 00:05:29.820 Yeah. 00:05:29.870 --> 00:05:33.640 I remember traveling when we had little kids and that was a whole nother level because you'd come back. 00:05:33.760 --> 00:05:36.800 It's like, you've been gone for a week, so I need some help. 00:05:37.540 --> 00:05:38.400 But I'm exhausted. 00:05:38.500 --> 00:05:38.980 You don't understand. 00:05:39.460 --> 00:05:41.700 But at the same time, it's a fair request, you know? 00:05:41.960 --> 00:05:42.840 Yeah, no, absolutely. 00:05:43.120 --> 00:05:43.240 Absolutely. 00:05:43.620 --> 00:05:44.500 I mean, our kids are older now. 00:05:45.200 --> 00:05:50.460 The eldest is 18, or will be 18 next week, just about to go to university, and then 16, 00:05:50.960 --> 00:05:51.540 and 13, 14. 00:05:51.680 --> 00:05:53.320 So they are a lot easier than they used to be. 00:05:53.460 --> 00:05:57.820 But still, it's a big ask on my partner to go away and they're alone for the whole week. 00:05:58.940 --> 00:05:59.260 Absolutely. 00:05:59.620 --> 00:06:02.440 In that point position, we might call it. 00:06:03.000 --> 00:06:03.460 Yes, exactly. 00:06:04.520 --> 00:06:09.660 so uh tell people you mentioned django on the med yes i talked to paulo about this a couple episodes 00:06:09.850 --> 00:06:13.700 ago but yeah tell me okay you're going to be part of this too yeah yeah yeah so paulo and i 00:06:14.040 --> 00:06:19.500 this was paulo's idea several years ago to have core developer sprints for django um because at the 00:06:19.630 --> 00:06:23.160 the end of jank on europe or the end of jank on yes we always have a couple of days of sprints 00:06:23.480 --> 00:06:28.100 and they're some of the best parts of the conference um but also it's around the time 00:06:28.310 --> 00:06:32.579 people are leaving and your people are already kind of exhausted you know having the three days 00:06:32.600 --> 00:06:39.040 of talks plus then of the sprints so it's like we wanted more and um we decided to to do this event 00:06:39.270 --> 00:06:44.100 and we were like yeah this is a good idea but it didn't happen and then i think dublin a year a 00:06:44.210 --> 00:06:47.640 year and a half ago now uh we were like no come on let's like it so we said we're going to have 00:06:47.640 --> 00:06:52.580 it in palo vijay which is my hometown um we did that last september and all last october sorry 00:06:52.740 --> 00:06:57.560 beginning of october and we're doing it in paulo's hometown pescara this september and it's as i say 00:06:57.560 --> 00:06:59.440 It's three days to get together and work on Django. 00:06:59.760 --> 00:07:02.200 And we had some really big features landed. 00:07:02.350 --> 00:07:05.880 Like, for instance, Django 6.1 is in pre-release now. 00:07:06.260 --> 00:07:10.860 And it's adding an ability to have DB-level cascade options for the on-delete. 00:07:10.980 --> 00:07:16.300 So on-delete is, well, you've got a foreign key, so I've got a user, and that user has a bookmark. 00:07:16.560 --> 00:07:18.100 Well, what happens when I delete the user? 00:07:18.150 --> 00:07:19.180 What happens to the bookmark? 00:07:19.440 --> 00:07:20.900 It's got a parent relationship. 00:07:21.380 --> 00:07:23.600 And so normally you would cascade that. 00:07:23.880 --> 00:07:28.540 Now, Django's always offered this option, but historically, it's done that in memory. 00:07:28.600 --> 00:07:36.980 In Python, it's pulled, it's walked the tree, and it's got the objects, and then it's deleted the leaves, and then the stems, and only finally down to the branches. 00:07:37.200 --> 00:07:40.040 And now, that can be very costly at scale. 00:07:41.220 --> 00:07:46.380 I'm going to delete this one record that references several tables with a million entries. 00:07:46.700 --> 00:07:47.740 That could be a lot of money. 00:07:47.980 --> 00:07:48.300 Exactly. 00:07:48.500 --> 00:07:53.320 You've got an organization that's got team members and projects, and you delete the organization. 00:07:53.600 --> 00:07:59.380 it just brings down your application basically you get out of memory memory issues so the database 00:07:59.550 --> 00:08:04.720 still has to do that work but the database is much much much better at doing that work than python is 00:08:05.070 --> 00:08:10.640 um and so that's so anyway we've got this new option this ticket had been open for i don't know 00:08:10.860 --> 00:08:15.680 so many years because it's it was really hard and we had marish feliciak who's a former django fellow 00:08:15.690 --> 00:08:21.919 and one of the core um orm contributors simon charette is one of the core open source contributors 00:08:21.940 --> 00:08:26.420 Lily Foote, Jacob Tyler-Wells, who's the new Django fellow. We had the four of them just there working 00:08:27.280 --> 00:08:32.380 together and they were able to get this over the line. And it's literally something which would 00:08:32.760 --> 00:08:37.979 never have happened except for the time that was available at Django on the Met. So we've got 00:08:38.140 --> 00:08:41.940 exciting plans for this time. We're hoping to get people again, people from around the community to 00:08:42.080 --> 00:08:46.900 come. And I know, I'm really excited about it. The first one was a proof of concept. It went really 00:08:47.080 --> 00:08:50.899 well. The second one's lining up to be amazing too. And once we've done two, well, it's kind of 00:08:50.860 --> 00:08:55.920 established and you know we're yeah well it's tradition you gotta do it yeah yeah exactly you 00:08:56.200 --> 00:09:01.280 gotta go hang out with the mediterranean so i've done i've done a lot of remote work in my life 00:09:01.530 --> 00:09:06.480 and i find you know when we do in-person meetups get the people who are all distributed together 00:09:06.770 --> 00:09:11.820 it just changes the energy you know there's something about the bandwidth of face-to-face 00:09:12.450 --> 00:09:18.339 communication which just isn't rec replicable record you which you can't replicate replica 00:09:18.360 --> 00:09:25.660 online yeah thank you mike thank you my but yeah it's it's i don't know i mean look it works online 00:09:25.780 --> 00:09:30.740 but as well we had another it was a good session as well so with um django we've got this um we've 00:09:30.740 --> 00:09:35.880 had recently this new features board which was an attempt to take people suggesting features away 00:09:35.990 --> 00:09:40.700 from django's track instance which is kind of old and scary and is a place for for the work to be 00:09:40.710 --> 00:09:45.839 done and put it into a GitHub project and where we could have a project board and we could have 00:09:45.760 --> 00:09:50.260 upvotes and we could have a Kanban board to see what were being advanced to try and give a bit 00:09:50.340 --> 00:09:54.600 more visibility into the process. And sometimes there's disagreements and those disagreements 00:09:54.700 --> 00:09:58.260 online, they can go on forever and ever, but we were able to have a session and there were 00:09:58.360 --> 00:10:02.980 different voices from different opinions. And we were able to come to agreements in five minutes 00:10:03.400 --> 00:10:10.200 on topics which had been rolling on the boards for months. And that kind of, look, I can see 00:10:10.680 --> 00:10:15.720 why you're saying this. It's really hard to get that online. You know, there's money emojis 00:10:15.740 --> 00:10:20.480 we use as much as we try and communicate openly it's just as much slack as you want to throw out 00:10:20.620 --> 00:10:27.100 there and you know i mean since 2020 it's it's kind of been just oh yeah obviously remote is better 00:10:27.380 --> 00:10:31.900 and so on and there are mega advantages don't get me wrong but it's not always better and in lots of 00:10:31.960 --> 00:10:37.400 ways so yeah so i guess for this good gone you you mentioned 2020 and it's covid it's one thing that 00:10:37.630 --> 00:10:43.259 um i was um django fellow over over that period and um you know it was a really hard time for 00:10:43.280 --> 00:10:48.820 because i think that the pandemic really knocked the back out of the community not we had we 00:10:48.900 --> 00:10:52.460 had older people stepping away we had young roots that were starting to come and then the put the 00:10:52.880 --> 00:10:58.120 the pandemic came and it was kind of like this really really really tough period and it's now 00:10:58.520 --> 00:11:02.080 five six years later that we're looking at it we're thinking actually now the community you know the 00:11:02.180 --> 00:11:06.560 activity is going on with things like jang and on space and the new community the new members that 00:11:06.580 --> 00:11:11.819 are coming through and the people who who are active in the community think well next year when the 00:11:11.840 --> 00:11:16.700 steering council is re-elected. There's obvious candidates who weren't there two years ago. 00:11:18.020 --> 00:11:24.420 And it's taken us, I think, half a decade to recover from the real knock that was COVID. 00:11:24.720 --> 00:11:30.920 Wow. I believe it. I believe it. So is this for just for core devs or is this open to people who 00:11:30.920 --> 00:11:35.660 just want to contribute? No, it's open to everybody. Obviously, if you're an established 00:11:36.160 --> 00:11:41.800 contributor, it's going to be easy to get going. But it's also about bringing on new 00:11:41.800 --> 00:11:43.480 and teaming people up. 00:11:43.760 --> 00:11:45.060 And, you know, if you come 00:11:46.600 --> 00:11:48.820 and you've got a particular interest in Django, 00:11:48.920 --> 00:11:50.200 if you're a total Django novice, 00:11:50.280 --> 00:11:52.740 it would be strange to come to choose Django on the Met 00:11:52.740 --> 00:11:53.740 as your first event. 00:11:54.080 --> 00:11:56.560 But if you're into Django and you know it 00:11:56.640 --> 00:11:57.480 and you want to get contributing, 00:11:57.780 --> 00:11:59.880 it's, yeah, that's part of the plan too, 00:12:00.200 --> 00:12:02.940 is that we have both experienced and new contributors 00:12:03.100 --> 00:12:05.100 and we can team people up and get people working together. 00:12:05.440 --> 00:12:09.500 And, you know, part of the contributing to Django problem 00:12:09.900 --> 00:12:10.660 is the learning curve. 00:12:10.760 --> 00:12:12.620 So it feels like it's a big, scary thing. 00:12:12.750 --> 00:12:17.640 Whereas if you sit down with somebody in a room and you work on what they're working on, you're like, actually, I could work on this side ticket. 00:12:17.800 --> 00:12:19.200 And all of a sudden people are off. 00:12:19.300 --> 00:12:23.000 Well, it suffers from what many of the successful projects out there do. 00:12:23.420 --> 00:12:24.880 It's super polished. 00:12:25.280 --> 00:12:27.260 It's used by millions and millions of projects. 00:12:27.980 --> 00:12:33.900 It's hard to just look at it and go, oh, I'll just change this without regard to all that knock on effect, right? 00:12:33.960 --> 00:12:45.740 Yes. And that's where part of the new features repo was about giving a space where people could put their ideas forward without having them immediately closed as won't fix. 00:12:45.890 --> 00:12:51.840 Because if you went and, you know, oh, can we integrate React? No, we're not going to do that. 00:12:52.020 --> 00:12:57.080 You can use React and here's how we use React, but we're not going to integrate it into JankRusate as an example. 00:12:57.620 --> 00:12:59.860 And that would be open, it'd be no, no, close, won't fix. 00:13:00.010 --> 00:13:01.939 And people felt that that was a little bit too... 00:13:03.060 --> 00:13:04.140 A little too stack over flowy. 00:13:04.660 --> 00:13:06.020 Well, yeah, it wasn't ideal. 00:13:06.180 --> 00:13:10.880 So the new features repo is to allow space to develop for a conversation. 00:13:11.060 --> 00:13:14.120 And then those ideas which have a bit of traction, we can say, no, do you know what? 00:13:14.340 --> 00:13:17.520 A lot of people have said they support this one or we can advance it. 00:13:18.400 --> 00:13:18.940 Hey, one second. 00:13:19.399 --> 00:13:23.180 Normally, this would be an ad break from Sentry, but not this time. 00:13:23.560 --> 00:13:27.120 Let's just thank them for supporting the show and get right back to the conversation. 00:13:27.620 --> 00:13:31.000 Also, visit talkpython.fm/sentry after the show. 00:13:31.340 --> 00:13:31.800 Thanks, Sentry. 00:13:32.280 --> 00:13:35.440 You have another project coming along, The Mantle. 00:13:35.710 --> 00:13:36.400 Tell us what this is. 00:13:37.000 --> 00:13:37.600 Django Mantle. 00:13:38.600 --> 00:13:42.840 So it's my take on a modern serialization story for Django. 00:13:43.020 --> 00:13:46.900 So the big topic we've had is about REST Framework, and REST Framework is aging. 00:13:46.950 --> 00:13:49.280 And REST Framework is wonderful, and it has been wonderful. 00:13:49.480 --> 00:13:51.860 But the serializers in particular, they're kind of last generation. 00:13:52.190 --> 00:14:01.040 And there are much more modern approaches available now, involved in Pydantic, in msgspec, in Attrs and Catters, which is what I use, 00:14:01.180 --> 00:14:03.820 are much, much faster than they used to be. 00:14:04.650 --> 00:14:06.580 They generate code which is almost as quick 00:14:07.550 --> 00:14:09.780 as you would see if you hand-wrote it. 00:14:09.940 --> 00:14:12.580 If you wrote the code to literally pull this attribute, 00:14:12.820 --> 00:14:13.680 this attribute, this attribute, 00:14:13.920 --> 00:14:16.300 the code that's generated for you by, say, caters 00:14:16.550 --> 00:14:19.440 is as quick as that code, which is amazing. 00:14:19.610 --> 00:14:21.680 Whereas REST Framework serializes, 00:14:21.730 --> 00:14:24.420 they go through and get the field object from the model. 00:14:24.580 --> 00:14:26.100 They get the value from it. 00:14:26.500 --> 00:14:28.400 It's like, hang on, all of those loopings 00:14:28.680 --> 00:14:30.200 are very slow by comparison. 00:14:31.300 --> 00:14:33.880 So then the question is, well, why don't we just use Pydantic? 00:14:34.120 --> 00:14:37.440 Well, Pydantic's fantastic, but it's not ORM aware. 00:14:37.870 --> 00:14:42.000 It doesn't have any insight into how Django works or anything. 00:14:42.050 --> 00:14:44.360 So Django Mantle is my take on that. 00:14:44.540 --> 00:14:49.540 It's the type-save wrapper around Django's liquid core. 00:14:49.550 --> 00:14:54.320 So you write these atters classes, and then you just instantiate a query object, 00:14:54.390 --> 00:15:00.260 and you go fetch, and it fully instantiates your object to the shape you define, 00:15:00.520 --> 00:15:02.140 and it fetches all the data all in one go, 00:15:02.540 --> 00:15:03.900 and it can do that for nested objects, 00:15:04.040 --> 00:15:06.300 and you can do writes with validation, all these things. 00:15:07.460 --> 00:15:10.880 And it's ORM aware in that it only selects the fields that it needs, 00:15:11.260 --> 00:15:14.280 and it does the pre-fetch relateds for you 00:15:14.420 --> 00:15:16.900 so that you don't get this N plus one query problem, right? 00:15:17.140 --> 00:15:20.520 And it's very good for performant fetches 00:15:20.660 --> 00:15:22.400 of just the exact data you want. 00:15:23.580 --> 00:15:26.460 You end up with your objects being fully Python typed. 00:15:27.199 --> 00:15:28.700 They're at its classes, so they're typed. 00:15:28.880 --> 00:15:30.240 Your type checker knows the way around them. 00:15:30.800 --> 00:15:35.620 mypy or PyRefly is now adding support for Atis classes too, I see this week. 00:15:36.380 --> 00:15:36.760 Oh, cool. 00:15:36.960 --> 00:15:43.120 So anyway, it's an attempt to show what a serialization story using these modern approaches 00:15:43.320 --> 00:15:43.840 can look like. 00:15:44.080 --> 00:15:48.360 And then from there, it's like, well, okay, what does an API framework in 2026 look? 00:15:48.460 --> 00:15:48.980 I love it. 00:15:49.030 --> 00:15:49.680 I love this idea. 00:15:49.850 --> 00:15:55.979 I think there's a lot of value in taking some of these very focused, lightweight, dedicated 00:15:56.440 --> 00:16:03.600 like data structures right like c attrs or pydantic or just straight data classes and somehow 00:16:03.880 --> 00:16:07.320 integrating them in here without just this heavyweight stuff like you talked about the 00:16:07.320 --> 00:16:12.180 eager loading that's and some of the other things like the typing can be weird as well like sql 00:16:12.380 --> 00:16:18.420 alchemy you kind of have to lie to it yes somehow do you either lie about it being a database thing 00:16:18.420 --> 00:16:24.320 or do you lie about it being an integer when it actually gets the value right like yeah yeah yeah 00:16:24.320 --> 00:16:28.300 Yeah, because the actual type is a Django deferred field, right? 00:16:28.400 --> 00:16:30.460 For a Django modifier, it's a deferred field. 00:16:30.540 --> 00:16:31.940 What on earth is a deferred field? 00:16:32.160 --> 00:16:34.900 Where it's an object that has the descriptive protocol 00:16:35.120 --> 00:16:37.660 so that when you access it, it gives you the converted value. 00:16:38.640 --> 00:16:39.420 I don't want to type that. 00:16:40.380 --> 00:16:40.680 Exactly. 00:16:41.100 --> 00:16:43.080 Even if you do, then you try to add it to an integer. 00:16:43.240 --> 00:16:44.460 It's like, you can't add that to an integer. 00:16:44.640 --> 00:16:45.400 That's not the same thing. 00:16:45.440 --> 00:16:46.460 You're like, but it's going to be. 00:16:46.780 --> 00:16:47.860 Yeah, it's going to be. 00:16:47.940 --> 00:16:51.680 So this was the topic of my talk that I gave in Amsterdam 00:16:52.020 --> 00:16:53.600 and Django in Europe and PyCon in Thailand. 00:16:53.820 --> 00:16:57.420 It's slightly different versions because I had more time at DjangoCon Europe. 00:16:57.470 --> 00:17:02.500 So I went more into the way we grow from, say, an active record pattern, 00:17:02.550 --> 00:17:03.640 which is like a Django model, right? 00:17:03.640 --> 00:17:04.319 It's an active record. 00:17:04.500 --> 00:17:09.300 It's your database row, but it's also an object that's got logical behaviors 00:17:09.680 --> 00:17:12.439 to a more separated approach where you're like, hang on, 00:17:12.470 --> 00:17:16.300 I don't want my display logic and my aggregation logic 00:17:16.500 --> 00:17:20.060 and my user validation logic all living on this one database class. 00:17:20.699 --> 00:17:21.959 So how do we separate those off? 00:17:22.040 --> 00:17:23.819 That's kind of where mantle came from originally. 00:17:24.240 --> 00:17:32.800 But the idea was that, and my take is that Django's ORM is built on these very powerful dynamic patterns that Python allows. 00:17:33.260 --> 00:17:36.660 And retrofitting static typing onto that is going to be a real problem. 00:17:36.710 --> 00:17:38.700 So instead of that, that's also a dynamic C. 00:17:38.840 --> 00:17:44.260 On top of that, we have these static islands where we can have all the nice type safety features of modern Python, 00:17:44.550 --> 00:17:49.940 but we're not trying to eliminate the very powerful dynamic features that we have underneath. 00:17:50.520 --> 00:17:55.180 And so the mantles, the shape classes that you define, they're your static islands. 00:17:55.290 --> 00:17:58.580 And there you can add all your domain logic that you want to, and you can have it fully 00:17:58.860 --> 00:17:59.660 separated from the database. 00:18:00.030 --> 00:18:01.500 And they are just pure Python classes. 00:18:01.720 --> 00:18:03.860 There's no data mutability in there. 00:18:03.960 --> 00:18:09.120 There's no like, say, oh, I modified a property and saved, and now I've updated something. 00:18:09.310 --> 00:18:11.580 No, there's none of that active record stuff going on. 00:18:11.630 --> 00:18:17.199 So it's a kind of way of getting some of the benefits of this sort of separation layers 00:18:17.220 --> 00:18:22.520 type approach whilst still keeping everything you love from the aura uh spoken like someone who 00:18:23.260 --> 00:18:28.760 appreciates some nice architecture and layers are very cool well the issue is as your application 00:18:28.940 --> 00:18:35.100 grows the complexity grows with it and you need more structure to manage that complexity like you 00:18:35.240 --> 00:18:40.440 simply do like it the issue isn't that you started with a jango model it's that your jango model grew 00:18:40.640 --> 00:18:44.999 and then grew and it grew and all of a sudden it's got it's got methods on it which are several 00:18:45.020 --> 00:18:50.060 distinct responsibilities. And then you think, oh, I'm getting a lot of code here. So you start 00:18:50.480 --> 00:18:55.300 doing dry, you start abstracting methods. But that abstracted method wasn't for a single 00:18:55.580 --> 00:19:01.140 responsibility. So it's not clean so that you've got, ah. And everybody who's grown a Django app 00:19:01.150 --> 00:19:05.460 has seen this. And it's perfectly natural and it's perfectly normal. It's just how applications grow. 00:19:05.840 --> 00:19:10.819 And Mantle is kind of my take on, well, how do we deal with that? How do we grow from the active 00:19:10.840 --> 00:19:15.080 record pattern that we began to, to something that's more robust. And it's not that you should 00:19:15.300 --> 00:19:19.960 start with the more robust patterns because you're over-engineering them. There's a nice road we have 00:19:20.020 --> 00:19:25.600 to walk between these two problems. And that's the joy of crafting the application over time, 00:19:25.700 --> 00:19:35.580 right? That's why we're programmed. I love it. I love it. All right. Let's talk about Django 6.1. 00:19:35.940 --> 00:19:36.020 Yes. 00:19:36.500 --> 00:19:38.820 I think it's coming out. 00:19:38.960 --> 00:19:39.740 It's not quite finished, 00:19:39.900 --> 00:19:41.940 but there's a lot of features we're going to talk about 00:19:42.010 --> 00:19:43.800 that are pretty exciting and related here, right? 00:19:43.900 --> 00:19:45.640 A lot of new things coming out of the database, for example. 00:19:45.920 --> 00:19:46.160 Yes. 00:19:46.270 --> 00:19:47.640 So I think it's in beta now. 00:19:47.730 --> 00:19:49.860 So it's going to be a release candidate in a couple of weeks, 00:19:49.930 --> 00:19:52.000 and then we're about a month away from the final release. 00:19:52.080 --> 00:19:54.440 Oh, by the way, I see you've got the page up there. 00:19:54.470 --> 00:19:56.480 The Django Developer Survey is on right now. 00:19:56.650 --> 00:19:57.360 Do take it, folks. 00:19:57.390 --> 00:19:58.640 We want as many replies as we can. 00:19:58.750 --> 00:20:01.840 It really does help drive forward the development of Django. 00:20:01.890 --> 00:20:05.000 So do give that five minutes there of your time to answer the questions. 00:20:05.640 --> 00:20:06.840 But what's exciting? 00:20:07.400 --> 00:20:09.940 Well, we already talked about database-level cascades, right? 00:20:10.140 --> 00:20:13.660 That's my big feature because that's something that you just couldn't work around. 00:20:13.850 --> 00:20:16.300 And there's no in Python solution to that. 00:20:16.350 --> 00:20:17.080 We wanted that. 00:20:17.660 --> 00:20:23.000 You'd have to, well, in the past, I have created a custom database backend and changed the way it creates. 00:20:23.240 --> 00:20:25.200 No, that's not a sustainable way. 00:20:25.970 --> 00:20:28.200 So database-level cascades are my favorite one. 00:20:28.200 --> 00:20:32.860 The one that I think more people are excited about is called fetch modes, which are on query sets. 00:20:33.280 --> 00:20:37.100 So we talked about the prefetch related with the way Mantle handles it. 00:20:37.600 --> 00:20:43.900 But the idea is you've got an author and so you fetch a user and they've got a list of bookmarks. 00:20:43.940 --> 00:20:50.960 And so you get the author, you get the user, and then you iterate for, bookmark in users.bookmark, do display the bookmark. 00:20:51.120 --> 00:20:57.480 And so what happens is it goes through and you get this N plus one query where it goes and fetches the lazy related object. 00:20:57.660 --> 00:21:05.260 And that's great at small scales and when you're getting going, but it's a performance nightmare, basically. 00:21:05.600 --> 00:21:09.740 And this N plus one problem that comes up is something that people hit all and all again. 00:21:10.160 --> 00:21:13.800 So fetch modes let you say, look, fetch peers, which is the exciting one. 00:21:13.800 --> 00:21:19.760 It lets you say, if I access the first element in a list of related objects, go and get them all at once. 00:21:20.220 --> 00:21:22.480 And so it's like a deferred pre-fetch related. 00:21:22.700 --> 00:21:32.820 And that's going to be a lot easier to manage in a lot of cases than having to manually add the pre-fetch related in each case, because those often become difficult to do by hand. 00:21:33.560 --> 00:21:36.320 They're all right to add the first time, but they become difficult to keep updated. 00:21:36.500 --> 00:21:39.720 And then you're suddenly pre-fetch relating something which you're no longer using. 00:21:40.060 --> 00:21:41.500 And so fetch mode. 00:21:41.510 --> 00:21:44.800 Yeah, your data access layer, maybe you're right, says get me all the books. 00:21:45.290 --> 00:21:49.840 But then you might have to add like some kind of flag that says and the authors. 00:21:50.300 --> 00:21:53.560 Because sometimes you want to call that function and you want the authors, but sometimes you 00:21:53.620 --> 00:21:55.780 want to just list the books, you know, and you don't need that. 00:21:55.920 --> 00:22:00.040 And it's like, yeah, so this way kind of just lets you says, if you're ever going to do 00:22:00.480 --> 00:22:04.040 this lazy loading, then basically upgrade it to an eager load. 00:22:04.260 --> 00:22:04.700 Yes. 00:22:04.980 --> 00:22:09.080 And there's, I don't think the videos are quite out yet, but Jacob Tyler was, you found 00:22:09.160 --> 00:22:12.400 a great talk at DjangoCon Europe on the, on how fetch modes work. 00:22:12.840 --> 00:22:13.760 They'll, they'll be out very soon. 00:22:13.760 --> 00:22:17.340 I think they'll be on YouTube, but do look out for them in the interest, but how, how, 00:22:17.480 --> 00:22:22.000 how fetchmates work and what the limitations were and you know where it's not you know it's not a 00:22:22.250 --> 00:22:27.660 magic solution it's not just oh let's enable fetch peers in all cases you know without thinking about 00:22:27.660 --> 00:22:32.860 it no but it it's a good change and um when i've spoken to people about what they're excited about 00:22:32.860 --> 00:22:39.020 in 6.1 most people have pointed to this as their headline feature nice nice so that's good and then 00:22:39.090 --> 00:22:43.420 i guess um a little bit yeah go ahead well i was going to say about the other thing that i'm 00:22:43.520 --> 00:22:47.440 excited about that you know there's an old list of features the thing with django releases every 00:22:47.460 --> 00:22:50.540 there's this big long release pack go and read it go and pick your favorite features but there's 00:22:50.620 --> 00:22:57.460 always lots in there for everything but you we can now customize the multi-part parser class for 00:22:57.900 --> 00:23:03.640 on the request object now whoever wants to do that well it turns out but that you can when you're 00:23:03.640 --> 00:23:08.780 getting multi-part requests if you're getting big ones with lots of multi-parts in swapping out a 00:23:08.860 --> 00:23:14.679 different multi-parts class could save you maybe 30 on your passing time which is 30 cpu on pass it 00:23:14.580 --> 00:23:15.880 on request for a passing time, 00:23:16.320 --> 00:23:18.080 that may be a non-negagable game. 00:23:18.610 --> 00:23:19.560 So I'm excited about that. 00:23:19.590 --> 00:23:23.240 But I'm excited because it starts to open the pathway 00:23:23.500 --> 00:23:25.340 to more customizations on the request. 00:23:25.390 --> 00:23:27.960 And so, you know, Django's got this implementation, 00:23:28.400 --> 00:23:30.720 but can we swap out that implementation and compare it 00:23:31.070 --> 00:23:32.900 and see if we can get some performance gains 00:23:33.030 --> 00:23:36.220 by being able to turn some of these nods under the hood. 00:23:36.440 --> 00:23:37.100 That's really nice. 00:23:37.500 --> 00:23:39.820 And having it be customizable is really great. 00:23:39.940 --> 00:23:40.460 I'm working on this. 00:23:40.740 --> 00:23:43.440 I'm working with this project that I just want to use. 00:23:43.560 --> 00:23:48.000 I don't want to be an author of it, but it has no real extensibility. 00:23:48.120 --> 00:23:52.540 So anything I want to do, I have to write little custom bits and change the source code and run it. 00:23:52.600 --> 00:23:55.980 And I'm like, as soon as there's a new version, how am I going to pay for this? 00:23:56.040 --> 00:23:56.560