WEBVTT 00:00.000 --> 00:12.000 And now I will hand give the floor to Emma, who will give her talk on Jango Atman. 00:12.000 --> 00:13.000 Thank you. 00:13.000 --> 00:14.000 Thank you. 00:20.000 --> 00:24.000 So hi, my name is Emma. 00:24.000 --> 00:27.000 I've been doing Titan for a long time. 00:27.000 --> 00:29.000 Recently, I do more Jango. 00:29.000 --> 00:32.000 I am a member of the PSF, the DSF. 00:32.000 --> 00:36.000 And I'm currently a member of the Jango steering console. 00:36.000 --> 00:38.000 But this is not a steering console talk. 00:38.000 --> 00:40.000 This is my talk. 00:40.000 --> 00:43.000 I also worked for a small company that I confronted. 00:43.000 --> 00:44.000 Left it. 00:44.000 --> 00:45.000 We're in Belgium. 00:45.000 --> 00:52.000 We do mostly heavy data based web application using Jango. 00:52.000 --> 00:58.000 And for bigger projects, I also work with Squats, who gave me these t-shirts. 00:58.000 --> 01:06.000 Squats is the teams of teams where you can basically hire a team and just play what you need. 01:06.000 --> 01:10.000 Since this is a Python demo, I'm first going to introduce Jango. 01:10.000 --> 01:14.000 Who doesn't know what Jango is? 01:14.000 --> 01:19.000 So I have three people into the form of president who doesn't know. 01:19.000 --> 01:26.000 Jango is the web framework for perfectionist's deadline. 01:26.000 --> 01:31.000 It is the most popular, so far, I think it's still the most popular. 01:31.000 --> 01:35.000 Web framework in Python, it's been around for 20 years. 01:35.000 --> 01:39.000 It's very mature, stable, and it's got a good community. 01:39.000 --> 01:43.000 And it also comes as batteries into the things like authentication. 01:43.000 --> 01:46.000 You where I am, everything like that. 01:47.000 --> 01:50.000 It's also used across open source projects. 01:50.000 --> 01:56.000 I think it's like authentic, paperless, but also big names like nasus, 01:56.000 --> 01:59.000 GPL, and so forth. 01:59.000 --> 02:05.000 And one of the most light batteries of Jango is the admin. 02:05.000 --> 02:09.000 And so for the two people in the room, we don't know what Jango was. 02:09.000 --> 02:12.000 This is what the admin looked like. 02:12.000 --> 02:19.000 The admin lets you create a crowd really quickly on any table in your database. 02:19.000 --> 02:22.000 Jango supports many database backends. 02:22.000 --> 02:30.000 And so if nothing else, you can use that to reverse your database if you don't have a correct client. 02:30.000 --> 02:36.000 And it's better than most clients, because it's also this relationship and things like that. 02:37.000 --> 02:39.000 Oops. 02:39.000 --> 02:44.000 And to write such an admin, it's quite easy. 02:44.000 --> 02:50.000 A few lines of code, and you know have something that lets you explore your data. 02:50.000 --> 02:56.000 Dupe thread, 80% of what you will probably want to do with one single table. 02:56.000 --> 03:02.000 You can do that in a few lines of code. 03:03.000 --> 03:07.000 What's not so great is that Jango is 20 years old. 03:07.000 --> 03:10.000 The concept of the admin is also 20 years old. 03:10.000 --> 03:14.000 And so there are things that for not built into the admin. 03:14.000 --> 03:19.000 If you want to do a drag-and-fout, if you want to have nested in lines, 03:19.000 --> 03:25.000 if you want to have extra actions, not just edit and delete with your staff. 03:25.000 --> 03:28.000 There are third party packages for all those. 03:28.000 --> 03:34.000 But the problem is that when you start to get too many third party packages in your Jango application, 03:34.000 --> 03:37.000 you start sometimes to get conflicts. 03:37.000 --> 03:40.000 And there are really more complicated stuff. 03:40.000 --> 03:44.000 So if you want to change the layout of your form, 03:44.000 --> 03:49.000 well, most of the time, the best way to do that is just to do it by hand. 03:49.000 --> 03:55.000 The thing is with the admin, is that a lot of the admin code is in two places. 03:55.000 --> 03:58.000 It's a changeless view and the change form of you. 03:58.000 --> 04:04.000 And those two things are buried deep inside the admin. 04:04.000 --> 04:08.000 And it doesn't really look like anything else in Jango. 04:08.000 --> 04:15.000 If you want to take a step back, the Jango admin is a framework inside of Jango. 04:15.000 --> 04:16.000 It's its own thing. 04:16.000 --> 04:22.000 If you know how to do filters in the admin, and you want to do filters in your regular views, 04:22.000 --> 04:24.000 well, it's completely different. 04:24.000 --> 04:28.000 You, your best bet is to install the third party package, 04:28.000 --> 04:30.000 because it's not there. 04:30.000 --> 04:35.000 Jango is filtering, but only in the admin. 04:35.000 --> 04:40.000 And the community knows that there are a few issues with the admin. 04:40.000 --> 04:46.000 Last year, the steering console introduced the new features people, 04:46.000 --> 04:49.000 which is a place where we can ask for feature, 04:49.000 --> 04:57.000 idea, it's a repo and GitHub, it's organized in a convenient way to follow where the things are along. 04:57.000 --> 05:03.000 And about 30% of the issues that were open there are about the admin. 05:03.000 --> 05:08.000 And there is one issue, specifically, issue number 70 that was open, 05:08.000 --> 05:12.000 that kind of got my eye and sent me down its rabbit hole, 05:12.000 --> 05:17.000 and because of issue 70, this is the reason why I'm standing here today. 05:17.000 --> 05:21.000 I will come back to the rabbit hole later. 05:21.000 --> 05:29.000 And not only the community knows, but the community has been trained for at least 15 years. 05:29.000 --> 05:39.000 There's Django admin too, which is factory revs, Django standard views. 05:39.000 --> 05:45.000 The problem with Django admin too is that it's still very close 05:45.000 --> 05:50.000 in a concept to the regular admin, and it's also been abandoned, 05:50.000 --> 05:54.000 and Django 2.2 is the last version in supports. 05:54.000 --> 05:57.000 Next to that, you have things like Drap Ali in Jet, 05:57.000 --> 06:01.000 that are mostly skins over on top of the admin, 06:01.000 --> 06:04.000 but it's extra functionalities. 06:04.000 --> 06:08.000 And then you have people who do things completely differently. 06:08.000 --> 06:14.000 You have Wactel that is CMS, one of the most successful CMSs is built on Django. 06:14.000 --> 06:19.000 But it has its own admin, and some people have taken that admin, 06:19.000 --> 06:22.000 removed everything that was CMSC from the admin, 06:22.000 --> 06:25.000 and used that instead of Django admin. 06:25.000 --> 06:30.000 And then there are people who just go outside of spite and altogether 06:30.000 --> 06:37.000 to use Django with rest framework, and they built a JavaScript front end. 06:37.000 --> 06:45.000 I wrote a library to do that, and this is one of the forms 06:45.000 --> 06:49.000 that I'm able to manage with that. 06:49.000 --> 06:55.000 And I would love for the admin to be able to do that out of the box. 06:55.000 --> 07:00.000 And it's not just a community, also a side of the community. 07:00.000 --> 07:03.000 Other frameworks have been evolving. 07:03.000 --> 07:07.000 20 years ago, while there was frails, there was Django, 07:07.000 --> 07:13.000 and there was not much else in terms of having an admin. 07:13.000 --> 07:21.000 And even Django, in Django, seeing, hey, we're doing this because frails does that. 07:21.000 --> 07:26.000 But in the past five years or so, things have started changing, 07:26.000 --> 07:29.000 and the other frameworks have mostly caught up, 07:29.000 --> 07:32.000 but they're also starting producing new stuff. 07:32.000 --> 07:36.000 They're starting to introduce dynamic companies there. 07:36.000 --> 07:40.000 They're evolving, while Django is still evolving, 07:40.000 --> 07:42.000 but the others are there, you know, 07:42.000 --> 07:46.000 and they've put up in some areas they have gone further. 07:49.000 --> 07:52.000 So what does this mean for the future? 07:54.000 --> 07:58.000 I would love the future to be extendable, composable, 07:58.000 --> 08:02.000 and I would love the future of the Django admin to be just Django. 08:04.000 --> 08:08.000 What if when you write something for Django, 08:08.000 --> 08:11.000 you can also use it in the admin directly and vice versa, 08:11.000 --> 08:14.000 instead of having to deal with two different paradigms, 08:14.000 --> 08:20.000 that the admin framework inside Django and Django. 08:22.000 --> 08:27.000 So I kind of went into a rabbit hole, like I said, 08:27.000 --> 08:31.000 and I looked more at what would that mean? 08:31.000 --> 08:37.000 Well, it would mean that it would have to be built on Django's generic views. 08:37.000 --> 08:40.000 Thread operations, of course, should be supported, 08:40.000 --> 08:43.000 but they should not be anything special, 08:43.000 --> 08:48.000 and it should not be more special than publish or clone. 08:48.000 --> 08:52.000 And also everything should be composable, 08:52.000 --> 08:57.000 so you should be able to take different parts of the framework 08:57.000 --> 08:59.000 and put them together. 08:59.000 --> 09:02.000 And for that, plugins need to be key, 09:02.000 --> 09:06.000 and I like the Django admin that was first created 09:06.000 --> 09:10.000 inside a newsroom in the US for very specific purpose, 09:10.000 --> 09:15.000 try to think about the plugin as first citizens. 09:16.000 --> 09:19.000 And of course, it should be very familiar. 09:20.000 --> 09:23.000 So you put it in images. 09:23.000 --> 09:27.000 If we take the structure, that would be the Django views at the bottom. 09:27.000 --> 09:31.000 On top of that, you add some niceties, some UX, 09:31.000 --> 09:35.000 some websites, side bars, that kind of thing. 09:35.000 --> 09:39.000 And of course, there's not just visual things. 09:39.000 --> 09:42.000 You also have to handle things that throw behind the scenes, 09:42.000 --> 09:44.000 I permission and all that. 09:44.000 --> 09:48.000 And a way to put that together would be plugin, 09:48.000 --> 09:52.000 using something that is already well known 09:52.000 --> 09:57.000 and supported for the Django plugin is one of those. 09:57.000 --> 10:03.000 And the glue that would come and bring all of that together 10:03.000 --> 10:06.000 would be few factories. 10:06.000 --> 10:09.000 So you register a third-party application, 10:09.000 --> 10:12.000 using the plugin interface, 10:12.000 --> 10:15.000 and then when you're looking at the list view, 10:15.000 --> 10:18.000 Django knows that for your list view you're using this, 10:18.000 --> 10:20.000 that and that plugin. 10:20.000 --> 10:24.000 And it brings out everything together. 10:24.000 --> 10:29.000 And of course, I said earlier, this should all be familiar. 10:29.000 --> 10:33.000 So back to Django admin definition. 10:33.000 --> 10:36.000 This is what an admin definition looks like. 10:36.000 --> 10:41.000 And when I went into that private hole, 10:41.000 --> 10:45.000 I wrote a whole proof of concept, right? 10:45.000 --> 10:48.000 And so this is Django admin do. 10:48.000 --> 10:52.000 And this is what this looks like in Django admin do. 10:52.000 --> 10:57.000 Did you spot the difference? 10:57.000 --> 11:02.000 The only difference is the imports. 11:02.000 --> 11:05.000 So this would stay very familiar. 11:05.000 --> 11:08.000 It would not break anything that you know about yet. 11:08.000 --> 11:11.000 But once you have changed those imports, you can go further. 11:11.000 --> 11:17.000 And you can enhance your admin with things that have been included, 11:17.000 --> 11:20.000 that are not possible right now with Django core admin. 11:20.000 --> 11:22.000 So for example, instead of using just strings, 11:22.000 --> 11:27.000 you can use colon objects and give more information about your colon, 11:27.000 --> 11:30.000 like change the label or things like that, 11:30.000 --> 11:33.000 which is say it's sortable or filterable. 11:33.000 --> 11:37.000 Or you can, for your form, you can also use a layout object, 11:37.000 --> 11:42.000 just using a list of fields or fields like this right now. 11:42.000 --> 11:47.000 But that's nice and theory. 11:47.000 --> 11:50.000 But would that really work? 11:50.000 --> 11:55.000 Yes, it would really work if I can get my mouse to put that into full screen. 11:55.000 --> 11:59.000 This is a demo of Django admin do. 11:59.000 --> 12:03.000 And when you look at it, you might not see anything. 12:03.000 --> 12:06.000 You might think this is just a Django admin. 12:06.000 --> 12:08.000 And this is good. 12:08.000 --> 12:10.000 This is what we want. 12:10.000 --> 12:16.000 We want it to be funnier to look at something that you know. 12:16.000 --> 12:20.000 And here this is just Django admin.co. 12:20.000 --> 12:25.000 We don't any of the plugins. 12:25.000 --> 12:30.000 And yeah, you see that it works. 12:30.000 --> 12:32.000 And it's not in special. 12:32.000 --> 12:34.000 But here you can see that there is a difference. 12:34.000 --> 12:36.000 There is this change password action. 12:36.000 --> 12:38.000 Right now in the, in the current Django admin, 12:38.000 --> 12:40.000 when you have users, you want to change password. 12:40.000 --> 12:44.000 You have to edit the user, then you have a link to the change password form. 12:44.000 --> 12:47.000 Here you could have that directly as an action. 12:47.000 --> 12:50.000 Because once again, other actions is not just edit and delete. 12:50.000 --> 12:55.000 Other actions list have the same status. 12:55.000 --> 12:58.000 You can find them on any thing here at the bottom right. 12:58.000 --> 13:00.000 You have to delete and you have to edit. 13:00.000 --> 13:04.000 And you have changed password that are accessible. 13:04.000 --> 13:21.000 So here I'm going to delete the user. 13:22.000 --> 13:26.000 And so one of the things that I was telling you about something 13:26.000 --> 13:30.000 that for some reason, very important to me is to be able to have different columns 13:30.000 --> 13:34.000 easily next to each other when you're doing a form. 13:34.000 --> 13:36.000 So this is the case here. 13:36.000 --> 13:41.000 It's just a simple Django form, but you get more flexibility. 13:41.000 --> 13:48.000 That's what you do get out of the box with just no plugins. 13:48.000 --> 13:53.000 But now we're switching to dark mode. 13:53.000 --> 13:55.000 Dark mode is Django admin. 13:55.000 --> 13:57.000 With the recommended plugins. 13:57.000 --> 14:01.000 That's Django filter that Django from set. 14:01.000 --> 14:07.000 So here we can see that you've got all the things that you're used to having in Django admin. 14:07.000 --> 14:14.000 On the right, you've got filters, you can search your views, you can search your list. 14:14.000 --> 14:19.000 The plugin, so the search on the top right, is provided by the core. 14:19.000 --> 14:23.000 But the filters is provided by the filter plugin. 14:23.000 --> 14:29.000 And those two, they work together without any issue. 14:29.000 --> 14:34.000 The form here is rendered by Django from set. 14:34.000 --> 14:35.000 Django from set. 14:35.000 --> 14:38.000 If you don't know, it's a library for doing forms in Django. 14:38.000 --> 14:39.000 Django is very powerful. 14:39.000 --> 14:44.000 You can do things like conditional display in that kind of thing. 14:44.000 --> 14:50.000 And other than it's dark, you probably wouldn't have seen the difference if I had 14:50.000 --> 14:53.000 Untiled you, it was rendered by another library. 14:53.000 --> 14:55.000 And this is the idea. 14:55.000 --> 14:56.000 It should be transparent. 14:56.000 --> 14:58.000 It should stay familiar. 14:58.000 --> 15:03.000 It should not bring anything different. 15:03.000 --> 15:12.000 So here I'm going to be trying the filters. 15:12.000 --> 15:15.000 This is very powerful. 15:15.000 --> 15:23.000 And yesterday again somebody was saying, yeah, but the problem when you start to use plugins and plugin 15:23.000 --> 15:26.000 is that you don't always know what's happening. 15:26.000 --> 15:31.000 You don't know what plugin introduces what into your view. 15:31.000 --> 15:35.000 You have a hard time allocating the code and so on. 15:35.000 --> 15:39.000 And so with great productivity. 15:39.000 --> 15:47.000 And because of that, Django had me do it comes with another plugin that documents. 15:47.000 --> 15:53.000 Every view that has been created that has been assembled from your base view and all the plugins 15:53.000 --> 16:00.000 that come over that, it creates a full documentation website with everything. 16:00.000 --> 16:07.000 And it looks a little bit like CCBV, except that it's focused on the plugin system. 16:07.000 --> 16:10.000 It tells you, yes, this is the base view. 16:10.000 --> 16:12.000 This is a base view class. 16:12.000 --> 16:17.000 If this plugin that plugin and this is where each function code, which matters come from, 16:17.000 --> 16:21.000 each property comes from and how they ended up there. 16:21.000 --> 16:28.000 This is a plugin that generates a website, but there's also a common live version that can give you 16:28.000 --> 16:35.000 a few smaller details about that. 16:35.000 --> 16:39.000 In last year, somebody in this room asked, 16:39.000 --> 16:46.000 why did we still need to rest third-party application to do rest in Django? 16:46.000 --> 16:49.000 And so I wrote an article saying that you did not. 16:49.000 --> 16:53.000 And then I wrote a library that did it, so it defeated the purpose. 16:53.000 --> 16:58.000 But I thought, should I try to break this? 16:58.000 --> 17:00.000 And so I wrote a plugin. 17:00.000 --> 17:06.000 I made a plugin version of that library that can stick into Django admin. 17:06.000 --> 17:13.000 And so you can, why am I speaking so fast compared to my slides? 17:13.000 --> 17:20.000 I made a plugin that's just asked rest to anything from Django admin. 17:20.000 --> 17:22.000 So now you have rest. 17:22.000 --> 17:25.000 Let's just like in the other action. 17:25.000 --> 17:29.000 Getting, doing a get, doing a get-feed quest with JSON. 17:29.000 --> 17:35.000 You're getting exactly what you would get with any other rest framework. 17:35.000 --> 17:41.000 And just to test it, I also, because I think it's still cool to do to do applications, 17:41.000 --> 17:44.000 when you want to do JavaScript and rest. 17:44.000 --> 17:52.000 So I did a small-to-do list application that uses that trust framework from Django admin. 17:52.000 --> 17:57.000 And I reload it several times just to show you that it saved some disk in the database, 17:57.000 --> 18:05.000 not just JavaScript metric. 18:05.000 --> 18:09.000 Yeah, that's it for that. 18:09.000 --> 18:13.000 So this is the list of the plugins that you saw at work. 18:13.000 --> 18:21.000 And this shows, this is to show that plugins have been designed into this proof of concept, 18:21.000 --> 18:24.000 as being first-class citizen, even the core. 18:24.000 --> 18:27.000 It's Django admin, the plugins core. 18:27.000 --> 18:31.000 Everything is plugins. 18:31.000 --> 18:34.000 The other thing is that you can try this without any risk, 18:34.000 --> 18:37.000 because it's another application altogether. 18:37.000 --> 18:40.000 So it can live right next to your admin. 18:40.000 --> 18:43.000 And if you want to try it, and if you want to migrate, 18:43.000 --> 18:49.000 you don't have to switch from the stack admin to Django admin. 18:49.000 --> 18:56.000 You can just do one model at a time and start your migration. 18:56.000 --> 19:04.000 Of course, the admin itself is nice, but there is a whole lot of other things around the admin. 19:04.000 --> 19:07.000 There is a full ecosystem of applications. 19:07.000 --> 19:12.000 And some of the third-party application will be easy to migrate. 19:12.000 --> 19:17.000 Some will be less easy, and some will be hard. 19:17.000 --> 19:21.000 The easy ones are the ones that either not necessary anymore, 19:21.000 --> 19:26.000 or don't do anything visually or that kind of thing. 19:26.000 --> 19:30.000 And the more visual things that a third-party application is going to be doing, 19:30.000 --> 19:34.000 the hardest is going to be to migrate it right now, 19:34.000 --> 19:38.000 because it's a whole different system. 19:38.000 --> 19:43.000 And under the hood, even if the API, the final user's API is the same. 19:43.000 --> 19:47.000 Under the hood, we got a threat of the changes view. 19:47.000 --> 19:50.000 Of the changes view, of the change from view, 19:50.000 --> 19:54.000 those concepts that are only part of the current admin, 19:54.000 --> 19:56.000 we got a threat of those. 19:56.000 --> 20:01.000 So, teams, for example, I would not advise anybody to do a team right now, 20:01.000 --> 20:06.000 except maybe some CSS customization, because the templates are changing a lot. 20:06.000 --> 20:13.000 Four things like Django Reversion that adds a sidebar in the admin. 20:13.000 --> 20:17.000 Well, we have a plugin that the sidebar thinks, 20:17.000 --> 20:21.000 but this will have to be translated to be compatible, 20:21.000 --> 20:23.000 which is sort of the plugin. 20:23.000 --> 20:26.000 It's going to take a bit of work. 20:26.000 --> 20:32.000 So, if there is enough success in people like the idea, 20:32.000 --> 20:37.000 this is roughly the roadmap, that's a project like that we need. 20:37.000 --> 20:41.000 It decor itself with still needs several months. 20:41.000 --> 20:42.000 Mostly clean up. 20:42.000 --> 20:44.000 I did the architecture. 20:44.000 --> 20:47.000 I delegated a lot of the codes to a robot, 20:47.000 --> 20:50.000 so no, I really have to do a lot of clean up. 20:51.000 --> 20:57.000 Taking care of the third or key positioning is important. 20:57.000 --> 21:00.000 But most importantly, if this goes forward, 21:00.000 --> 21:04.000 we will have to accompany the whole community, 21:04.000 --> 21:08.000 help people do the transfer, and not just say, 21:08.000 --> 21:10.000 hey, this is a new admin. 21:10.000 --> 21:12.000 No, you're on your own. 21:12.000 --> 21:16.000 So, this is going to be what's taking the most time. 21:17.000 --> 21:20.000 Unfortunately, time is money. 21:20.000 --> 21:25.000 And this is probably why, up until now, 21:25.000 --> 21:29.000 there is not been any change of doing something like this. 21:29.000 --> 21:33.000 Jacob Caton must say that it would take about a million-dollar 21:33.000 --> 21:35.000 to rewrite the admin. 21:35.000 --> 21:38.000 And that was 10 years ago before inflation. 21:38.000 --> 21:43.000 So, the DSF does not have money for that. 21:43.000 --> 21:46.000 The DSF barely raised 300K this year. 21:46.000 --> 21:50.000 And the DSF uses its money to maintain Django, 21:50.000 --> 21:52.000 not to add new features. 21:52.000 --> 21:56.000 The only feature that I'm aware of in Django 21:56.000 --> 21:59.000 that kind of got some funding with the migrations. 21:59.000 --> 22:03.000 And that was about 10 years ago as well. 22:03.000 --> 22:05.000 And this was brought from it. 22:05.000 --> 22:07.000 It was not funded by the DSF. 22:07.000 --> 22:11.000 But now, we have things like the severing textants 22:11.000 --> 22:14.000 and where people try to, of the Europe, 22:14.000 --> 22:18.000 if willing to invest money in open source. 22:18.000 --> 22:22.000 So, I'm thinking of maybe reaching to that. 22:22.000 --> 22:24.000 There's also other systems. 22:24.000 --> 22:25.000 Django is framework. 22:25.000 --> 22:29.000 There's proof that you can help other companies 22:29.000 --> 22:32.000 integrate the product into their own product 22:32.000 --> 22:34.000 and that they're happy to pay for this. 22:34.000 --> 22:38.000 Consulting time to do that. 22:39.000 --> 22:43.000 So, what could come next? 22:43.000 --> 22:47.000 We have, I said, we have some clean up to do this. 22:47.000 --> 22:51.000 More permission plugins to do more granular permission 22:51.000 --> 22:53.000 that was currently implemented. 22:53.000 --> 22:58.000 What we have is already more, has more choices already 22:58.000 --> 23:00.000 than the current admin. 23:00.000 --> 23:02.000 But we want to object a little permission 23:02.000 --> 23:05.000 to build permission, that kind of thing. 23:06.000 --> 23:09.000 We want to improve with a model, 23:09.000 --> 23:12.000 related model, functionality, 23:12.000 --> 23:15.000 internationalization, and much more. 23:15.000 --> 23:17.000 And if you're interested, if you want to try this, 23:17.000 --> 23:20.000 maybe your feature, maybe your feature you need. 23:20.000 --> 23:24.000 If you want to, go contact me after the talk. 23:24.000 --> 23:29.000 Because in order to move any further, 23:29.000 --> 23:31.000 this project needs you. 23:31.000 --> 23:33.000 It needs people to try the project. 23:33.000 --> 23:36.000 It needs people to give some feedback. 23:36.000 --> 23:40.000 It needs people to find bugs 23:40.000 --> 23:44.000 and find possible improvements. 23:44.000 --> 23:48.000 The best way is to give feedback is to either go directly 23:48.000 --> 23:52.000 on the project's repo or go on that issue 23:52.000 --> 23:57.000 70 on the new ideas repo. 23:57.000 --> 24:01.000 So, it looks like I finished in time 24:01.000 --> 24:04.000 and we have time for questions. 24:12.000 --> 24:13.000 Hello. 24:13.000 --> 24:15.000 Thank you for the talk. 24:15.000 --> 24:16.000 Thank you. 24:16.000 --> 24:19.000 I have a question that might be controversial, 24:19.000 --> 24:22.000 but you talked about core. 24:22.000 --> 24:23.000 Yes. 24:23.000 --> 24:26.000 Is there any chance that some of these core 24:26.000 --> 24:28.000 could be reused for a fast API, 24:28.000 --> 24:35.000 so that we could help and contribute for both communities? 24:35.000 --> 24:37.000 I said controversial. 24:37.000 --> 24:38.000 No. 24:38.000 --> 24:42.000 It's just, I don't know. 24:42.000 --> 24:46.000 I think maybe. 24:46.000 --> 24:50.000 So, this core, we believe that there 24:50.000 --> 24:55.000 would need to be a similar core for fast API 24:55.000 --> 24:58.000 and then have plugins that work for both. 24:58.000 --> 25:00.000 That might be a solution, 25:00.000 --> 25:04.000 but then how do you deal with the ORM and things like that? 25:04.000 --> 25:05.000 Yeah. 25:05.000 --> 25:09.000 So, fast API doesn't come with its ORM, 25:09.000 --> 25:12.000 but you could use, like, a small community or something. 25:12.000 --> 25:13.000 Yeah. 25:13.000 --> 25:18.000 Or you could actually use a Django ORM with fast API. 25:18.000 --> 25:19.000 Yes. 25:19.000 --> 25:20.000 It would work. 25:20.000 --> 25:22.000 And then it might work. 25:22.000 --> 25:23.000 It might. 25:23.000 --> 25:24.000 Thank you. 25:24.000 --> 25:27.000 You're welcome. 25:27.000 --> 25:30.000 I want more questions. 25:35.000 --> 25:36.000 Hey, Emma. 25:36.000 --> 25:41.000 I've put the API part of the project is really cool. 25:41.000 --> 25:45.000 To what extent do you think it would be useful 25:45.000 --> 25:49.000 in, sort of, large scale production code with, 25:49.000 --> 25:52.000 mainly API features, for example, 25:52.000 --> 25:54.000 is it your vision that someday it could replace 25:54.000 --> 25:56.000 the need for Ninja ORM framework? 25:56.000 --> 25:58.000 Or is it just something basic 25:58.000 --> 26:00.000 that's kind of a helper utility? 26:00.000 --> 26:01.000 How do you see that feature? 26:01.000 --> 26:04.000 So, it would definitely not replace 26:04.000 --> 26:06.000 Django ORM framework, 26:06.000 --> 26:08.000 but it could come on top of 26:08.000 --> 26:11.000 Django ORM framework. 26:11.000 --> 26:13.000 Right now, 26:13.000 --> 26:16.000 Django, why do I say Django in the day? 26:16.000 --> 26:20.000 Ninja is a lot of manual work to create stuff. 26:20.000 --> 26:25.000 So, there is, among the long list of things to look 26:25.000 --> 26:28.000 at, a Ninja plugin, 26:28.000 --> 26:30.000 that would help, 26:30.000 --> 26:32.000 and maybe manually, 26:32.000 --> 26:34.000 automatically create the serializers 26:34.000 --> 26:36.000 that kind of think. 26:36.000 --> 26:38.000 And that would definitely be something 26:38.000 --> 26:40.000 that's this kind of plugin system 26:40.000 --> 26:42.000 or the load to do. 26:42.000 --> 26:44.000 Same thing with Django or framework. 26:45.000 --> 26:47.000 Django or framework is much easier to do 26:47.000 --> 26:49.000 because there's already an application 26:49.000 --> 26:52.000 that does that, is the RSKM adapter. 26:52.000 --> 26:55.000 And it automates a lot of things. 26:55.000 --> 27:00.000 So, that's something that I could do much faster. 27:00.000 --> 27:03.000 But yeah, the rest application is, 27:03.000 --> 27:06.000 mostly, I'm not going to say the size. 27:06.000 --> 27:09.000 The small resting is something 27:09.000 --> 27:12.000 that you would use if you just have one or two 27:13.000 --> 27:14.000 rest views. 27:14.000 --> 27:16.000 And you don't need an old thing. 27:16.000 --> 27:19.000 If your whole application is based on an API, 27:19.000 --> 27:21.000 well, you should probably use a third party package 27:21.000 --> 27:24.000 that is specializes in APIs. 27:24.000 --> 27:28.000 But this could really come on top of that 27:28.000 --> 27:32.000 and help create factories for the views 27:32.000 --> 27:36.000 that you need to use a Ninja behind the scenes. 27:36.000 --> 27:37.000 Yeah, thank you. 27:37.000 --> 27:38.000 I think it's really cool. 27:38.000 --> 27:39.000 Really exciting idea. 27:39.000 --> 27:40.000 Thanks. Thank you. 27:43.000 --> 27:45.000 Do we have questions from the chat room? 27:45.000 --> 27:46.000 Do you know? 27:46.000 --> 27:48.000 No questions from the chat room. 27:48.000 --> 27:49.000 Okay. 27:49.000 --> 27:51.000 If that were all questions. 27:51.000 --> 27:53.000 One of the hands last question. 27:53.000 --> 27:54.000 And then we go. 28:01.000 --> 28:03.000 Really cool concept. 28:03.000 --> 28:07.000 If there was the funding to get this third and third 28:07.000 --> 28:10.000 of would this be something that could replace 28:10.000 --> 28:14.000 Django admin, would this be a final goal for a project? 28:14.000 --> 28:17.000 So, yeah, so if there is funding, 28:17.000 --> 28:19.000 so I'm one of the things I'm going to 28:19.000 --> 28:22.000 walk into is getting funding 28:22.000 --> 28:25.000 either from the surroundings or 28:25.000 --> 28:27.000 I was not kidding. 28:27.000 --> 28:30.000 If any of you has a company that is interested in the project 28:30.000 --> 28:33.000 and wants help to integrate it, call me. 28:33.000 --> 28:35.000 So I'm hoping to get funding 28:35.000 --> 28:38.000 and if there is funding and if there is enough. 28:38.000 --> 28:43.000 There is other people in the steering council who think 28:43.000 --> 28:45.000 the idea is interesting. 28:45.000 --> 28:49.000 So there could be something that lands in Django 28:49.000 --> 28:52.000 to replace the Django admin. 28:52.000 --> 28:56.000 If that happens and nothing is going to happen. 28:56.000 --> 29:00.000 But if that happens, we will need to set a new 29:00.000 --> 29:03.000 deprecation policy in place to 29:03.000 --> 29:08.000 deprecate the current admin in maybe two or three version 29:08.000 --> 29:13.000 to major version instead of the current one major version. 29:13.000 --> 29:17.000 And probably what would land in core would probably be the core of 29:17.000 --> 29:22.000 Django admin do and leave the plugins outside. 29:22.000 --> 29:27.000 Right now, this is not going to happen. 29:27.000 --> 29:30.000 I hope it's going to happen later, but I don't know.