Transcript: From Kenya to London with Django - Velda Kiara

Hello, Will.

And we're very pleased to have Velda Chiara join us from London. Welcome.

Even when I started Django. And now I get to hang out with the kookies.

So for those who don't know, just as quick background, you're a software engineer at RevSys, Microsoft MVP.

You've worked with Defna, which runs DjangoCon US. You've worked with DjangoNauts on updates for the Django News newsletter.

You did a fellowship with the Python Software Foundation.

Let me look at my notes. I'm sure I'm forgetting something. And you speak regularly at conferences. We're going to talk about that.

So let me start with how do you do all that? Because I feel tired just listing all that out.

But yeah, it's like all of them are kind of like related. So it doesn't seem like it's a lot of work. Because if I'm doing...

I did DjangoNauts space because I wanted to contribute to Django. And then I got to be a maintainer for DjangoDebugtober through the program because I was an active contributor.

And I got along well with the maintainers. And yeah, it's like open source. And open source is fun, right?

And then the Django Events Foundation of America, I got involved because I attended my first DjangoCon US and I loved it and I wanted to be present and I wanted to be involved with making it better to make other people's experiences better as well.

And that kind of like happened as well. So they had a call for open positions that you could like apply for. And I was like, ooh, that sounds good. And I was like, I hadn't served on a board before. So I didn't know how that's going to be like. But I was like, I mean, it's Django, it's DjangoCon. So it's not going to be difficult. And that's how I got in.

Yeah.

Not difficult and then a lot of work, right?

I had to like justify why exactly I think I'm the right person for the job. And I think, yeah, also like listing out things that I have done and commitments that I had before and how that led me to where I am. Meaning I also did have like some leadership skills as well, which also helped me through the application, if I would say. So yeah, it's like consistently they're all connected. And then I also got to be a Python Software Fellowship Fellow. Yeah, PSF, Python Software Foundation Fellow.

Yeah, Foundation Fellows. A lot of Fs there.

Sometimes I get confused with like, is it Foundation? Is it Fellow? Which is it? But yeah, I got involved.

Oh, Foundation, yeah.

Yeah. So that's how I got involved too.

And that also became like, based on the contributions I had done, also like technically and also community wise, I got to be involved with like the Microsoft Most Valuable Professional program in a way. So all of these things are all connected. The good thing is that it's one thing that's like all these things. So yeah.

I get to meet Carlton and have conversations. And I'm like, what?

So I love it.

And part of the open source thing is like, it's like, how much can you handle? And so people are very open and welcoming, but there is a lot that happens behind the scenes for events, for the code, for everything else.

Yeah.

Yeah.

Social side.

And I think that's how I get to be here. That's how I get to be here. That's how I get saved.

So it's like the sun's up when the sun's up, the sun sets when the sun sets, right? And so everything is like really balanced. And we also have like really good weather. So when I came here, it's like negative something degrees. And they're like, oh, it feels like negative eight, but it's actually negative two. And I'm like, why?

So I had to also meet other people here as well and getting to interact with like other people who are also moving. So I think London is like the seed spot of like most of the people are also like moving in and other people are moving out. But it's always like, like it's so inclusive as well. It's like there are a lot of people who are also moving. There are a lot of people who are also navigating. And yeah, you get, you just find other people who are doing the same and then you start bonding over things that are different.

Yeah.

Yeah.

Right now it's going to be like a year in February 17. So I'm going to be celebrating a whole new year here. I'm excited. I'll probably get a cake or something. But yeah, I want to, I want to celebrate the change in weather and the change in life and everything. And I mean, some of the, some of the things that really shocked me is like, and something that actually I had to figure out was the whole assembling of furniture.

That was the most weird thing that I've ever done. Like normally, so when I order a bookshelf, like back home, I'm always like, I get, I get it delivered because it's already built. So it's just like, okay, here's your bookshelf. Navigate out to get into your room or your house. And then that's it. I'm not assembling anything. I order your bookshelf here. And I almost cried assembling it because it's like different pieces. And it's like a weird color. And it's like, I had to figure out which nails go where, what does it do?

What does it want? And I was like, I can't do this anymore.

I was like, exactly. Are you sure it's the right one? Because I ordered a complete one. So it's like one of those things. I'm like, wow, now, now I can assemble bookshelves and I look forward to assembling things. And I'm like, hmm, interesting.

But here we are.

You didn't even ask me how I knew that. But I was like, yeah, I need to talk to people who also like done move, like moved countries before. And I was like, this is the person to talk to. And weirdly enough, I also had like the option of like trying to figure out other countries other than the UK, maybe figure out Spain. But then when you look at the different opportunities there are and how easy it would be to like get from one job to another, the UK made absolute sense. And you were like, yeah, I mean, you're young. You got this.

I remember you specifically said that you were like, you're young, you have grit, you still have energy, go do it.

People had already started leaving. And I suppose it's the same with any place, right? You sort of ride that curve of like, oh, my God, things are great. And then like, that's okay. Like that, you know, when you get old, like Carlton and me, the thing about transient places is like, you know, people come and go, and it gets a little less exciting. You're like, oh, man, they're gonna come. A long way of saying, I think the next year will be even better for you. And I'll be curious what it's like in, you know, years three, five, you know, because you'll see, you'll see those newcomers. And

be like, oh, I remember that, like, first year energy.

And I'll be like, oh, I remember that, like, first year energy.

So open inbox, come through maybe next week might be fun.

But it's an ongoing thing. Yeah.

Yeah.

So we'll put a link to that. I thought that was great. That was a really great overview.

And let me put it to you. It's your talk. What was the impetus for this talk? And, you know, kind of the highlights, because everyone says, oh, Django performance, but then doesn't really know where to go. And you laid it out really well.

What issues do you have? Because, I mean, even Facebook uses like Django in a way and I'm like, is it Facebook or is it Instagram?

Yeah.

So most of the times when I hear like, opinions like this, I'm always.

Yeah.

Always curious. Okay. Why are you saying that? Like, what use case do you have? Because I'm like, I know there's definitely different use cases that might not work. And that's understandable. But just saying it because of just saying it didn't make sense to me. So I went and did a whole talk on it on PyCascades. That was like last year time, like now, actually. So it's really neat that it's also coming out right now, when you're asking me this question. And so I wanted to dig into how you could also like,

increase performance and how that would look like on your app. And not just for people who are

scaling, but people who would just want optimal performance year round. And so when I looked at

that, I was like, okay, what are the things that we could actually talk about other than the common

things like select related, prefetch related, caching and indexing, right? And then also when

it comes to indexing, they're like different trade-offs, right? You can't index every single

column. Like that doesn't even make sense, right? And plus it's also creating like a different data

structure, which means it's also like getting, using up more space than you need. So all these

things, and I'm like, nothing does something perfectly all around, right? That's something

I came to accept in the software space. Like everything does something really well, but not

everything. And that's why we also have like different libraries and different packages.

And I'm like, yeah, it works, but to a point. So you just have to figure out what your point is

and then work on that. And that's how I actually approach the talk as well. And I talk about a lot

of like, um, a sharding as well. I talk about serialization, which also ties out to like

a marshalling that I had on my newsletter last week.

It's super. You write that every week, no?

stories to tell. And like, what's good about my newsletter is it's called the Storyteller by Tales.

So I'm like, I love telling stories. I love telling people about what's going on, what's fun to do.

And this time I had to do it differently because I was doing a course in system design and they

mentioned something about marshalling. And I was like, oh, that seems interesting, but I don't know

what that really means. And I was like, oh, that seems interesting. And I was like, oh, that seems

so I had to like put a pin on it. And then that same week I had a meeting with like Jeff and Frank

and they also mentioned marshalling. I was like, this is a sign. This is a sign. And so on the

newsletter, I had to like define what marshalling is in the concept of in system design and also

in Django. So in terms of system design, it's basically trying to change data into a format

that's easily transmissible. And so I had to like put a pin on it and then that same week I had a

meeting with like Jeff and Frank and they also mentioned marshalling. And so on the newsletter,

I had to like put a pin on it and then that same week I had a meeting with like Jeff and Frank and

and you're like, "Oh, so I've been doing this

this this whole time and i didn't know what that was just some facts you read for it exactly it's

like oh and then also when i started getting into system design i really wanted to understand like

how do people get like the calculations of how many seconds this result how many seconds you're

getting the results how like how everything ties up and how do we actually define how many servers

do we need how many like data data storage do we need and i was like i was curious about that and

that's how i ended up in the system design course and that led me to marshalling and now writing it

on my newsletter so yeah it's like a lot of interesting things that i come across of and

sometimes like in the newsletter i just like like highlight things that i have done during the week

so it might be like a crazy story or maybe how difficult i had assembling a shelf for a quick

laugh and then or maybe like things or maybe if i visited somewhere and i think it's like a good

recommendation for people so yeah

like you there are hundreds of email marketing software services out there and they all pretty

much offer the same thing collect and clean addresses send out broadcasters or drip campaigns

get analytics so you can see what's resonating and what's not button down is designed to hook

into the tools that you already care about everything from static site generators like

jekyll or hugo to payment platforms like stripe and memberful you can hook your site up to button

down with just a form element or you can hook your site up to button down with just a form element or

a simple rest call write emails in markdown and then get on with the actual work you're supposed to

do new customers can save 50 off their first year with button down using the coupon code jango and

if you email support they'll white glove migrate your existing subscribers and archives for free

the secret of longevity is to go every fortnight every two weeks

i want to do that in my life like i want to take vacation and not do a lot of things no it's called

semi-retired or you know taking a year a fun was it fun employment you got to say sabbatical that's

like the nba version it sounds vaguely academic but i mean it does sound fancy it's like yeah

that's going to serve you a massive stead because it doesn't matter what language or framework you're

using like yeah there are some sort of kind of engineering basics which if you haven't got it's

you're kind of guessing whereas if you have got them you can do the calculations up front that's

the same level of engineering all of you are at and you you already have like a lot more years so

i'm trying to compensate by get learning a lot more at a faster rate so i could also like catch up

or one year of experience 10 times and i think doing what you're doing of like going the upper

level of saying oh like there is this pattern i can name it i can see it in django that's how you

get 10 years experience instead of right it is very

easy and i certainly have succumbed to this of just like being so focused on the task at hand

you never get the bigger picture whereas i don't know carlton i would say now i feel like the only

thing i have is some degree of a bigger picture and the confidence to like be able to dive in

because i can't possibly remember everything but i'm like oh i kind of know how this thing

fits together or i know how i would go tackle it if i needed to whereas when i was starting out it

was just everything felt new and overwhelming i think the i think experience

some ways that it might work and it may not you have to implement it and see but yeah

experience gives you that ability to kind of start planning it before you even finish

reading the problem statement like it's you know gently um so i don't know but yeah that

once you start like looking at things and from a system perspective is that nothing is ever tied

to a specific technology that's why i said once you find your point then you start working from

the point that you had it's not like i am attached to this tool it's like no which tool does does it

actually work best with and it like detaches you from like all the tools or everything that needs

to be like any tool that you need to use to build it out and just like looking at the problem and

being like these are the calculations that you need to make and then you start looking at the

problem and being like these are the calculations that you need to make and then you start looking

looking at the problem and being like these are the calculations that you need to make

to accomplish and this is like the output of what we expect the system to to be or to do and that

really helps also when you're working in django because then you know why this is needed and it's

performance and you know people saying django doesn't go fast enough when you've actually got

a service level requirement that says oh we need you know ten thousand requests

a minute or whatever it's like well django will satisfy that quite easily on quite a small box so

were working with and we had like issues with caching a lot so we consistently had to like

bust out of caching and i'm like i thought caching works i thought it would work on every news case

what is going on and i was getting frustrated because jeff was giving me that ticket because

i mean you did it

first so you gotta see it through so it's like i have to make sure that it works and

it's one of the reasons why i think it would break is because we were loading up like

30 000 results in one page and inside the 30 000 results are like 15 000 each so i'm like

i mean it can only work to a point so we had to consistently like figure out how

if the user wants results immediately because i mean it's a user and they need

to get what they want they need to get what they paid for so had to figure out okay so we're not

going to cash this for a while or we're going to be busting out of cash in case there are changes

and it's like working with different clients is both interesting and challenging but you always

get the most bizarre feature requests and i'm always like wow well i'm interested to see how

I feel like it's the best because I get to work with like a lot of things that I don't think I

would have worked with if I only did one product or one project. And I really do appreciate that

because I also get to work on open source. I get to contribute to like Django Health Trick. I get

to work with other people's libraries as well and contribute to them. And I also get to build

different projects every couple of months. Maybe this month or the next few months I'm going to be

working on e-commerce and I need to figure out, okay, how do payments work in other places, in

other locations? What are we expecting? And in other situations, like right now we're working

with like a payment, I want to say payment platform service, but yeah, it's like something

in line with payment. And so every project is different. Every need is different and every

request is always different. And that's what I look forward to because

there's always something new to learn. And I feel like that just excites me. I'm like,

okay, let's see how we do this one. Okay, let's see how this one goes. So it's like,

I'm like, I don't get bored is what I'd say. So it's constantly new things, new issues,

new projects. And I like that. And I appreciate that. Yeah. And it's a good, if you've got the

good team of like people who've been in Django. And I'm like, sometimes we're like in meetings

and they say something, I'm like, I don't know what that is. How do you do that? And it's like,

getting like some of that experience from them when they're like listing things and they're like,

oh, you could use this package. And I'm like, I didn't even know that package existed. I don't

RevSys is, you know, one of the better ones. Uh, and also the clients, you know, you're not

just doing crud off real estate websites or something all day long, which, you know, I get it,

the best people who are also actively involved in the community. So they also appreciate like

community efforts and they're like, yeah, they fully support it. And I'm like,

I don't know. It's, it's like an ideal place for a general person to work at. And it was

really difficult to also like figure out like what projects or what, um, when I was choosing

this career, well, actually I chose it. And then I realized that, oh, I needed to specialize

and they're not so many options because most of the options that I was getting were like

JavaScript based. And I'm not saying JavaScript is bad, but that's not what I wanted to do.

I wanted to write Python every day. And so if I wanted to write Python,

I started working with Django and I was like, it would be really cool if I got involved with

a company that does that. And that really reduces the choices that you have because

most of these companies don't also hire every single time. And I'm always like,

wow, how do I get into this job? Like how, how am I going to manage getting a job?

And so I actually met the team at DjangoCon. It was my very first DjangoCon in 2023.

It was in Durham. And I actually didn't know that people had like Jeff or Frank had a company

I was just like talking to everybody who would like listen to me and be like, Hey, I'm doing a

lot of open sourcey just in case you have something, let me know. I'd be happy to do it and

before, uh, thing at like the pizza and beer place? Was that, I think, cause I think, I think

I was, I don't know if that was the first time I met you, but I definitely, we met, we met there

like in person. Um, is that the first time you met Jeff and Frank? Yeah. The first time I was also

actually the first time I was meeting Jeff and I had gotten margaritas like before. And then we

walked over and, um, okay. Interesting. Yeah. Wow. Well, I mean, God, like for me, it's like

a blip. And then for you to go from that to where you are now, that's, that's kind of crazy. That's

little I have or the more than I have, because I'm always like, I don't know if I'm ever going to get

in. So I'm like, I'm going to use it to the best that I can meet a lot of people. And also like

meeting you for the first time, I was like, why is an author talking to me? Like, aren't you like

to answer me and not have like, I want to say pride, but yeah. And be like, so done to it. And

still feel. So, you know, it all just filters around. I think, I think the story you've just

community? How do I get a job in Django? Well, there are a lot of companies out there that are

using Django that don't label themselves as Django shops, but you know, to turn up at the

conference and actually, you know, make sure you introduce yourself to everybody, you know,

make sure you're really out there and just say, Hey, I'm doing this. I'm volunteering. Look at me.

This is my profile. This is what I'm doing, how it's going to work. And it's going to work a lot

better than sending out, you know, cold approaches on LinkedIn or I got, you know, some mass mints

and gist spam the other day, you know, it's got to have zero response, right. To get out there and

actually put yourself in the community. Yeah. That's how you get into it. Yeah. So, you know,

applications I was getting like to the last stages and they'd be like, Oh, because of your location,

it's going to be like difficult to like, and I'm like, but despite my location, you can do,

you can see, I can do the job. So what's going on? And so I was like, wow, if the job is not

going to come to me, I have to go to where the jobs are. And that's, that's the baseline I had

for going to Django Con U.S. I was like, I don't know if anything is going to pan out,

but I just want to go have fun, meet people, interact, and maybe eventually someday in my

lifetime, I'll eventually say that I got this. So yeah. And it worked out great. And most of the

times I'm always like, I cannot believe that I was just talking and it wasn't like an official talk.

It was just like chats. We were having like, I think dinner or was it drinks? And like,

I just mentioned like, Hey, I'm looking for something that this is my job profile.

Well, actually I didn't say job profile, but I said, yeah, you can find me on LinkedIn. I

use my official names. And in like a couple of months, you were like, Hey, yeah, we have this

open source thing that we're looking to upgrade. Would you be open to doing something? And I was

like, yeah. And I mean, I was like, I was very flabbergasted. I was like, you remember?

Like it wasn't like, um, actively being like, Hey, I need a job. Where are you? Here's my CV.

Call me back. And all that. It's like, no, I actually wanted to interact you firstly, because

you were a person.

And I've seen like interesting things you've worked on. And that's where people, they also

say like, you've done a lot of open source stuff. So if you want to talk open source, go to him,

talk to him. So I was like, okay.

This is what I'm looking for. You're not asking everyone for a job. You're just saying,

can you help me? Right. So then someone may not be hiring, but everyone knows everyone

and they can say, Oh, have you. Right. Like that's, I mean, I used to think of this when I was,

you know, in Silicon Valley for a brief period of time.

People from my school would call me and I would always think, you know, you could ask me

how, like what the job is like, or you could ask me for a job. And I'd rather you ask me for a job,

but like, not me specifically, but say, Hey, I'm a business student looking to break into startups.

This is what I want to do. Here's my profile. Who do you know? That's much strong, you know,

or I'm a software engineer. Like it's so much stronger. It's so much easier. And again,

it often is the case that it's like lodged in someone's mind. And then when they hear,

Oh, someone's looking for something. It's like, Oh yeah. Have you talked to them?

really helped because then they could just look at like contributions I had made because that's

like, Oh yeah, you worked on this feature on this project. You've worked on a couple of projects as

well. I do think that really helped because when I was also hunting, I was like, I don't want to

just do nothing and just apply for jobs. I want to actively participate in like things that I want

to do. So eventually something, if something happens, it does.

Otherwise I still get the skills. So yeah, it's like, it's tough out here. And now with the eye,

it's like, it's not going to be a nice, a nice environment because I'm like, it's a lot like

approach system design, like I was telling Carlton before we started recording, like,

so I've been using, really trying to use, especially Claude and starting to use Codex with Django stuff.

And like, I've never, it's never written something that I would pass off as my own in entirety. Now

it's probably trained on my code and Carlton's code, you know, but like, it's like, if I, you

know, I don't know, that's the, that's the mental thing I have is that like, I use it. It's great

for referencing. It's great for ideas, but as a replacement, it seems like an accelerant for

someone who's eager to learn, go to the next step. It's a great tutor, but to say it's going to

it just misunderstands what a software engineer does. Yeah. I think what you said is like really

tremendous amount of experience, people who've seen like different problems, how to approach

different problems. And I think like, as you said, like you and Carlton have a lot of experience,

you've built out systems, you've broken systems. So even when something breaks, you already know

where, where the potential problem is.

And I think with a number, like with a good number of experience and like trials and errors, I think

you have a good baseline to just be like, I don't want to do the grant work. Let me just

write a spec and let AI handle it. And then I could review, right. And that works perfectly

because I mean, you've already accumulated a lot more. And so right now it's just like

basically upgrading what you have or using, constantly using what you know, and learning

some other new things. But then the challenge comes in when people want to replace like

their brain functions with AI was like outsourcing decisions to AI. Like you can't say hello to

someone without referencing AI or telling AI to like, Hey, respond to this. And I'm like, that's

where I think the issue comes in. It's like, I think you can still be human and do like normal

brain functions instead of outsourcing your brain. And the other thing that I'm pretty worried

about is like right now with like early talent, like early talent.

How like you can use it as a good research point, which I think it does really well. But then if you

start using it frequently without absorbing the information that you're getting, then it just

thinks for you. And then I mean, it could think to a point, like it's not going to be like 100%

sure. And how true do you know that whatever it's saying is okay, if you don't know even what the

answer is, if you don't know what the answer is, if you don't know what the answer is, if you don't

know what the answer is, if you don't know what the answer is, if you don't know what the answer is.

So I think that there are some pros and cons, but I do think that people who are thriving the most

with AI are people who are experienced and people who can afford to be like, I don't need to use all

my time doing this groundwork when AI could do it. The other people who are struggling are people

like me who are trying to level up and trying to figure out what's the best way to work with AI

without necessarily, necessarily impacting my brain and, and like having like retaining content

and retaining information. And so I don't know, those are just like different conversations to

have. I mean, I don't know. It's like, it has disrupted how everything used to work.

in a good way and in a bad way as well because it's like okay so how are we going to conduct

hiring so are people going to be allowed to use ai in interviews and then how does that look like

for the company like how do we know that this person is actually really good at what they do

or if it's just a flake or a fluke so i don't know it's like it's a lot i saw there was a headline

was just going to evaluate, evaluate you on your prompts.md file. Like they're like, oh, that's the

only thing that matters, how good your prompts are. And then like, that's not the onion. Like

there are, you know, I mean, that says a lot about YC these days, but yeah. Well, let me ask,

so specifically though, so for you, right. So you do client work, the client just wants it to be

correct. How do you, so that's good in that you, you can use it however you need to use it. You're

not being forced to use it. If it helps you use it, if it doesn't, you don't use it, but how,

how, how are you using it? Cause Jeff has mentioned that you are using it and finding ways. I'm always

curious, like how, you know, what are the positives that you're finding for using it?

Google results and be like, oh, this is, this is a hundred things. These are a hundred articles.

I'm always like, Hey, I'm looking for resources on, let's say Nats, because I used Nats previously.

Like who has written an article on Nats that uses Django or Python that I could look at. And then

I'm like source, because I'm always like source. I don't want you giving me your opinion.

right. Cause it won't unless you ask it for it. Yeah.

Right. And you know, AI wasn't built to tell you the truth. It was built on existing data. It just

gives you output, just gives you feedback. That's like, I got into a hole of like researching

MLMs with how they train everything. So I do know that it was not trained to tell you the truth. It

was trained to give you information. The information is entirely up to you to figure out if it's true

or not. So I'm very cautious about that. So if I'm looking, um, I usually use it like that and

like resource stuff, because sometimes I'm like, I would like to use it in like code and be like,

Hey, write this, write that. But I'm always like iffy about it for some reason, because I'm like,

okay, is it going to think for me? Is it like, how do I work with it? And I think I've mentioned to

Jeff this as well, because I'm always like, most people like to use it. But at the same time, I'm

like, I'm still like doing a lot interesting things. And I don't even want to outsource that

trial and error, that struggle of learning to AI. So I'm like, I don't know if I want to outsource

that, like using it to debug like JavaScript, because I do not like writing JavaScript, but I

do use it to like debug JavaScript research. And I think the other use case was like, if I'm trying

out a model, I'd want to see how the tests would look like. So it's like, shows me like a mockup of

tests. And then I could be like, okay, maybe I could think in that line, but touching my code,

nah, it's I'm always like, no, I'm not trying to do anything funny. And I'm always like using it

on chat. I'm not audacious enough to have it in my, in my terminal. I'm not audacious enough to

change anything.

I'm just like, keeping it at arm's length and trying to also make sure that I'm in the correct

space of consistently learning and trying out new things. Because I mean, how else are we going to

learn at this point?

to progress so that in five, 10 years time, you're a top notch engineer that is going to

make dividends whether you're using AI or not. Like, however it rolls out, it's not going to be

the case. Well, not unless these machines get suddenly much more capable than they are. But it's

not going to be the case that the human isn't needed that the skilled engineer isn't. And if

you just lean into them, totally, you won't be a skilled engineer. Whereas if you force yourself

to learn, you will.

doing performance in Django, there is like trade offs that you need to think about. So technically,

it's like the same thing. But I'm like, I

don't know if you also know enough that, like how everybody's saying like, oh, we're going to do a lot

with like medicine and all that. And I'm like, the same thing that's giving me weird results is what

we're depending on for like medications right now. I think we have some long way to go before that's

possible. But okay.

can't they you know, they know the you know, they know the facts in a way they can pull out their

training data, but they're not able to take a history and interpret it in a way that leads to a

diagnosis reliably in some limited cases, maybe but not, you know, okay, yeah, not like your general

practitioner is going to be replaced by a machine anytime soon.

delivery coming in with like a different system. And so to get back to the customer service, you

had to go through a bot, but a bot doesn't have all the use cases. And so I was stuck on the belt for

a while. And then I had to like, I had to like use weird, like, issues that people had hired that had

like shared on Reddit, so that I could like, figure out how to bypass the bots because it was an actual

problem that as long as your issues are not meeting what the bot needs, it's just going to be stuck in

the same process. And so you don't get like, also to contact like customer care and I'm like, but I

know we want to replace like the the customer support.

And with AI, but there is no like learning loop. And there's no like human who's like on standby,

if like there's a persistent user who's having problems. So I was like, you're actually turning

away clients and customers, because from my experience, I'm never using that company again.

And I don't care if they build a new world, or they do whatever I am never touching them

ever again. And so it comes down to like, also in terms of like businesses that want to embrace

like AI and other different tools, I think it is okay, because sometimes it makes sense business

wise, right? But it might not make sense to like, for me as a user, but business wise probably makes

sense. But then when I have problems, how are we going to work together? Because I'm technically

giving you money, and you're giving me a service. So if I'm not giving you money for your service,

so how is the business working again? So I don't know, it's kind of like,

also like different edge cases, maybe to somebody else, it does really make sense. But to me,

at that point, it was like, no. And considering that there was a long Reddit thread, I do believe

of us as engineers, it's, it's, most of my friends are not programmers. And so it's interesting,

because we know we have a knowledge of trying the tools for like the one use case where people are

right now. And so you read popular things, and it's like, oh, it's taking over coding, therefore,

it's gonna just one to one replace all these other things. And I find that it's hard. It's an

interesting place to be, it's hard to have like an actual conversation with people about what it

might, you know, there's a lot of medical people near here, what it might do, because, A, they don't

understand the underlying technology, but B, they're just reading, oh, it's replaced all the programmers.

And it's hard to have any nuance in there. Like, it's hard to do things like, you know, when you can have closed

systems with reinforcement learning. So yeah, better at chess, better at Go, like, absolutely

machines. And I'm sure there's gonna be a growing case of where AI can take over there. But most of

the world and most business problems are not closed systems. And so yeah, it's interesting,

like we're living in the one area that's being disrupted. That is the harbinger for everything

else. And yeah, it's, yeah, it's, I don't know, just an observation. I don't know, Carlton, if you have lots of AI fanatic

near you and the coast of Spain.

I may do. I don't speak to people. So it's hard to know.

I gotta try. I gotta try that.

things. I did want to ask you about Neapolitan. So Velda, this is Carlton's package. I understand

you've actually been using it in the real world. So can we talk about that a little bit?

mention that I found it hard to use your library. And well, part of it is because I also didn't read

the documentation. But beside that,

But at the same time, it's like, oh, God, I don't know how this works. So I think also the problem, well, not the problem, but the problem is that I don't know how this works.

not the problem or like the the project I had I had to do a lot of customizing and so I had to

actually get into like the base templates more to define and I was finding it hard to locate where

those files were and also figuring out like which one does what and as I said I didn't read the

documentation but I had to because I had a custom problem to solve and so I had to also look for the

template text specifically for the views because there were a lot there were different ways that

we wanted to custom make the the views and the forms as well and I found it so difficult I was

like why why could we just create them normally and it's like we want to use just one for like

Neapolitan they're just sort of like um what do I call bootstrapping templates so they get it up

and I immediately replace them with my own

you

the ones for that view so like I won't use the yeah edge and table the list view or the the

object view I would certainly don't try and customize the template tag I'm just like no no

I just want to put in my own template here the the bit that I think is of genuine event and I

love it I'm literally in love with it is crud view itself because it gives you that namespace

it gives you all the methods but yeah no the first thing I do is I bootstrap the view like

you know I put it in you know my four line thing and I've rooted it and I load the page yep

it loads right then like okay here's my own template yeah I mean that does work and it does

out testing the views as well each and every one for create update and delete but then I think once

we we got into like the actual like there is a certain way that we wanted the forms to look like

and I think I realized they did the project was using the default like templates that you had so

I was like okay that's going to be a bit of spice inside there but yeah so I had to like figure out

like how do I actually edit out this form how I want to add more fields but we're using the same

ones that are there so how do I add that like going to the tags and figuring it out and so

I was like why but then at the same time like yes it does make sense but why

form renderer which brings in the templates that I want to use for the forms and I'll I basically

just use the stock ones I've got some small net small customization so I can override the field

template much more easily yeah yeah and then I just define some css that styles them how I want

so I use the div templates and in my my actual templates that I'm using I just use the mustache

thinking this whole time I did I think I don't know if I 100 promised you but I told you I'd

write a tutorial of neapolitan and I still haven't done it so yeah okay so the bottom line is it's

really really long list there's a kind of how I django mega section on my site that just has

never happened because I've never had the capacity to to to put it together but it's how I do forms

how I do you know templates how I do you know um I mean the thing I like though is that you said

you had to get into the source because the whole idea of neapolitan is that the source code is in

that one views file and you just open the views bar and you just read it and you sit down and

you read it and when you've only it's only like 300 lines it's only like 300 lines it's only like

300 lines on you read the whole thing and you understand it entirely and you can just go right

I'm just going to override that I'm going to override that yeah but you're assuming we know

didn't have when I was looking at that project for the first time and so even when I was going to

Jeff I was like I don't know how you use this but then after doing a lot like reading out the

documentation I was like makes sense I understand why Carlton did that and I was like I'm definitely

beginner on-ramp here look if you're just getting started with views if you're getting started with

Django all of that stuff is just totally missing because I haven't literally haven't had the

capacity to write it I thought you were going to tell me off about the fact that I haven't updated

like where we are right now even without updating it I don't know what to do I don't know what to do

it needs a lot more information because I was like if I could have just schemed this page or they go

out of like gotten it easier but then I also skipped the documentation and assumed that

stuff I'm working on now which I've been working on for a little while which is why I haven't

updated Neapolitan because it's pending those changes um so I've got that to get out and then

Neapolitan will get its kind of um phase two of all those things

that I said were on the roadmap that I never did um including you know my take on you know Neapolitan

API and things like this that I've people have asked me about and I'm like oh I don't last the

thing is I'm loath to just put out a you know put it out until it's right I want the API to

be just so how I want it so that you know yeah because it's stable and I like it stable if I'm

going to leverage my projects on it it has to be stable and then the sort of long list of semi

deployment is to you know improve the the story for other people who don't know Django so well yeah

we I had to actually do and override like different tags and everything else it did it did work well

so I applaud you for that but everything else like getting on board and everything else it was like

nah okay okay well I'll have to I'll have to better lock will in a cage and make him write that uh

get some pie charm time for it then it's easier actually I think this year it's getting a little

bit easier I was gonna I was gonna say as you're talking Carlton I think the key to Neapolitan just

like Django itself is like Django was built to solve a real real world problem right that's what

Jacob and um Adrian would say and Simon right and and you yourself you're using Neapolitan yourself

right it's not some academic thing so I think that just bleeds through with like it's actually meant

to solve a problem and that's

the North Star um and that's where a lot of these other projects fall flat maybe is they're just done

to be interesting as opposed to like nope this solves the solves my problem and I'm using it in

time I'm just like yes this is what I want and there are bits I've got in the work project I

haven't yet up sourced back to Neapolitan because of these other changes that I want to get out it's

that meme of giving a medal to yourself right like Obama it's not just giving

it's like it's like my future my past self having already walked the path for me and laid down the

stuff right and then my current self just has to walk along it and it's like oh wow yeah thank you

I would type the question I would pull up a tutorial I wrote on it asking the question

exactly the same way and it's sort of like this like you know doppelganger thing I'm like huh like

I phrased it exactly the same way four years later I mean enough time passed and I forgot that I

reason why he puts everything into an open source package is because he doesn't have to want to solve

the same problem twice he wants it there available for him in the future when he's got the same

problem again and I really like that projects and books should we switch to that yeah um maybe start

and and I'm it's sort of like puts everything into the suspicious glaze for me but yeah

to go use it now

um and so I helped maintain the toolbar with Tim Mathias and Elenita so we're like a team of four

and right now what interesting thing I'm working on for the projects is the redesign phase because

I realized that I think I'm the most from all of us I'm like more passionate about making sure

everything looks good and feels good and so that's where we are we're trying to we're starting out

with like revamping our logo so you might be seeing a new logo in a few months in your

apps and we're also trying to figure out what the next steps are because you want to do like

an end-to-end design from getting our own colors getting the themes that we like and seeing if we

could also like improve our documentation as well and maybe build out a website who knows where this

is going to go but yeah I'm excited about it because now I had to I had to do like everything

I had to figure out where we're going to get designers who do we want to get involved and

like open source design is like one of the hardest things to do because it's open source

so I'm not going to pay you for your time and I feel bad about that because I do think

like designers are God's favorites because I don't know how you

ideas from thin air and just like create something beautiful and have people enjoy it so i'm like

i would love to pay you but right now i can't

job and so all that is still in progress so we're having a lot of back and forth right now to figure

out what the next steps are what colors we're are we looking at what does that say to a user

because we also need to be mindful of what the user's mental model is when they look at the

django debug toolbar and so it's like i'm picking up design skills on the side i'm like oh open

source for the win product skills yeah yeah and product skills because i'm like oh now i need to

come up with a timeline and i'm like oh how long does it take ideally well this is

going to figure out what the next steps are what colors are what colors are we looking at

what does that say to a user because we also need to be mindful of what the user's mental

is now like a thing of the past and i was getting so annoyed because i was like come on like give

me a break like it's not true at all it's dealing with people like there's no prompt that can deal

because you need to have that conversation of like if you're not liking a color why are you

not liking the color or if you're not liking the design why are you not liking it right

because it's like yeah everything needs to be good but we also need to define what good means

to me that's different to the maintainers and to the designer as well so it's like a weird i don't

know it's like a nice spot to be at where we're like all thinking about like almost the same thing

i mean like oh but when i look at green i see this when i look at orange i see that and it's like

wow we all see color very differently and we all like different color

as well and yeah it's like also getting to the user's perspective and being like oh this is

gonna be too smart for the user or this is going to be like enough for the user because i think

there's a statement that jeff likes to use every time i do like very complex things he's like no

that's too smart we do not do too smart things you need to like see that's very wise of him to

opinion on anything so from my experience what i try to do is i give them a task i say

accomplish this thing with this tool and talk out loud and that's better than just like showing them

two designs and be like which one do you like better they're just like they have no idea but

if i'm like okay go from here to here um that seems to work better because people can't articulate

my in my head too smart for us yeah too smart no there's really should do a deep take of that

that'd be amazing yeah but yeah also like putting yourself in the in the shoes of a user and now

this is not just any user this is we're going to get people who are not using django try to use

django debugtober we're also going to get veterans who've been using django for the long

time and i know those ones are going to have a lot more opinions and i'm like that's fine but i need

to make sure that everybody is somehow happy i might not get everybody very happy but somewhat

happy is is good so that's what there's sent there's century for the advanced use cases right

but if it's just like random being like oh interesting

color or i don't like that color i'm like okay any suggestions you have you know yeah yeah cool

very exciting yeah and also calling out uh other people who want to get involved because

once we have that like the designs in place we would want people who would want to implement them

into our project we'll be open to getting more contributors there are also new issues that you

could work on and you could reach out to me or anybody else on github we're pretty nice we're not

mean we're nice

we are accommodating so you could reach out and start making an issue all right good pitch

new project django lsp so a language service language server protocol in django it's i don't

know support um autocomplete for your um settings for i think he's getting into models he's looking

he's looking at all sorts

it's another several of these now there's like that one josh josh thomas who's got um another one

he's got uh django language server which he's got um yeah completions there's another one which is

um django template lsp so there are several in this space as well as um what is it the

hang on i can't think it it's called um zuban which is a language server written by the guy

who did jedi so the gem jedi is the um long time static analysis and autocomplete library

for python he's written a new version in rust that's got django model support as well

and it just goes to show that you know you can do tooling for dynamic you know you know libraries

yes they leverage stubs where stubs are available but you know they also are doing custom logic to

get those all the completion in these other places and it's not like we have to rewrite the whole

language in rust in order to get decent autocomplete in python it's really exciting time so there's four

there but the adam's one came up this week and i was like wow him and josh need to get together in

the room because adam's got his jangles project and josh has got his um django bird project and

then they've both got language servers it's like do we need some sort of celebrity death match at

django contest which one i'm going to use or something i don't know like x factor you're

going to have like one direction whatever i mean ideally they could collaborate so they're not

reworking but you know if we have to steal cage match to decide the winner then well but this is

there were three leading ones at the time

and borrowed from one another and then south from andrew godwin was decided but you know there are

three really strong ones done so there's going to be the same with apis and it's you know you know

yeah yeah and when when carlton said jedi

mcgee did one of one of the three um andrew godwin did one and i'm sorry i forget who the third one

but there were three big ones at the time and i i'm pretty sure there's some talks that i think

russell and andrew have given about as well there's a panel there was something i saw where

they went into the history of it um it's taken me a while to sort of piece that all together too

so um yeah sorry to the third one we'll look you up um for me

packages django link check which checks broken links on your django site goes through your models

um pretty cool i should use it more uh and i actually was thinking of it because i was

looking through something recently and i saw a broken link and i was like oh

so this is a long-running um project and it doesn't have a ton of stars but like we all have

broken links in our stuff so check it out seems pretty mature yeah and then i'll switch to books

i'll go first

i'm gonna pick a lightweight one this is called a history of the bible

by john barton who i think is a what is he's like a cambridge or oxford professor

one of those oxford um but actually interesting so treating the bible not

you know as a religious text per se but as a book and talking about the different authors

and how it's interpreted in um judaic and christian traditions anyways sort of nerdy but i like that

kind of stuff and this is actually my mom's copy i mentioned my mom's really into this stuff i mentioned

my mom i was like this book sounds good and she's like i've got a copy so um wow little little light

to me it like enriches whatever your belief system is to know about these things um so yeah

um yeah yeah maybe i would i know i've actually i have been reading a bunch this year so

i just finished a couple books and i was like should i do a write-up and i was like

kind of lazy but i was like there's some you know this is like the post-lm world like

just yeah the exercise of writing an essay forces you to come to terms with your own thoughts and

yeah you know there's

you know there's a lot of things that i've read and i've read a lot of things that i've read and

i'm sure it solidifies a book in my head to do even a quick take on it like for example i just

finished um normal people by sally rooney which i know is like a big thing and i had thoughts about

it so um yeah maybe i'll do that my tldr is i i enjoyed it i think i would love it if i was 25

at 45 i couldn't help but being like she's so 25 but um i mean you know come on all right anyways

uh

github so uh we'll put the link in the show notes but it's called python mastery um and there's a

pdf and there's a whole folder full of exercises and you sit down when you're you you look through

the pdf and you sit down with your editor and you write the thing by hand and it gets into

all of the tasty bits of python which so it's skates where the puck is going to be

carlton would be the told that we shouldn't go anywhere near but metaclasses and

you know the data model and you know get really in there it's it's practical python is the name

of the other one and david beasley is taught this course um as a paid course is stop teaching it

um and made it freely available and i i really recommend it he's also got an introductory one

for people who just want to learn python um which is worth checking out as well but that's not the

one that i'm super interested in i'm super interested in python mastery which is about

these more dynamic aspects with the analogy yeah

yeah

where i think the excitement is um now i know you know people are more interested in static

types and whatnot but i think actually what's python superpower well you know

yeah i agree well something will skate to where the puck was 15 years ago i think is in this case

you know like crossword puzzle for python yeah yeah that's right that's right but um this one's

called python mastery and the reason i wanted to recommend it so we talked to time and time

again about lmms and then to you know they're taking over the role of coding and then dumbing

us down and it's like well okay if we stopped going out and doing manual labor so we need to

go to the gym to keep fit well here's a way of doing some mental exercise to keep your programming

chops and learning something which is perhaps a little left field that you wouldn't do in your

day-to-day work but is really going to make you a better python programmer belda you've got a book

and it's like historical fiction and this book is gut-wrenching but it does he writes beautifully

and you get to have like a history of how afghanistan was and he's also american so

he has both perspectives and you see how they were like how people's lives were well

yeah how people's lives were before the war after the war people who moved how the different culture

customs some of them were maintained others weren't and it follows one particular person

who goes through a lot of things a lot of gut-wrenching things so it's if you want to know

it's like as i said historical fiction so most of the places actually do exist

the places that he mentions actually do exist but yeah it gives you a different

perspective of like things specifically in afghanistan being also from like a place where

it's war-torn and it's it's gut-wrenching to be honest so i would say you read it at your own risk

but the literature like the literature itself is very rich in terms of how the user tells you the

story the perspective he gives you the story and and how he builds up like anticipation of

one of the things that's going to be really impactful in the book and how he referenced it

it at the end as well and it's so good

it's so good it is so good i like it because it's like it's a fic it's fiction but it it also has

some history and it's like half the things i thought would happen didn't happen but they

happened in a different way and which is like more mind-blowing i'm like wow wow but yeah i'm trying

to read more of khalid husseini's work because he has written other books like a thousand splendid

songs and something about the mountain moved or something but yeah i'm gonna take a time out before i

pick up his next book because that was just too much yeah they're a little intense

they are very much so so i'm like i'm gonna be like i'm gonna pick up other books in like two

three months and then i'll i'll get back to the a thousand splendid sons but i highly recommend

liberally so code community

docs packages podcasts whatever what would you improve timing like timings of how things

out the website page or improve like how our interface looks and stuff like that and i feel

like we need i don't know if it's the time that we need to shorten the process of or how all these

things work but yeah if we could change like

the timings of what we're doing to like a little bit hasten the process so we're not talking about

a website in like two three years still and nothing's changed is something i'd like to improve

so don't come at me i am still in the company no no no no no i think we're in agreement

so it's like i do like i love the community i i wish more people also like get involved i wish we

also have a lot more money well

i'd say time and money i'd give jango a lot more money and i'll give it time

to hasten the process so i'm like yeah other than that i think everything else we have it like

know jeff is the jeff triplett is the new jango president there's a new board they're doing a lot

of things behind the scenes um pie charm jet brains is doing some sponsorship stuff i know

there's ongoing efforts to actually have an executive director so

um i yeah i just want to mention that i know i know firsthand people are working really hard on

this and i do think things are going to come to fruition but it takes longer than if it was one

person and you know pc funding so yeah the trouble is the resource for any given task is very large

we'll just be we'll be going on we would rule the world we'll be going to mars and jupiter and be like

be there um anyways thank you so much for making the time um and for coming on again it feels like

god what episode is this 190 something we should have had you on earlier but so glad you could make

that we were trying we were trying for a while so as part of it was scheduling you were busy and

there was stuff last year

you know it's like i remember my parents saying oh when i grew up and i was like what are you

talking about you're so old and now i'm living it i'm like oh god like you don't know anything but

it's to be discovered right you can read about it but then you have to live it and

yeah yeah i'd also love to work with either of you on like open sourcing things and yeah i think

neapolitan needs some docs carlton has a specific way he wants to do things so i'd be happy to help

the right way so if you did it he would probably he was like is that fair to say carlton

you know a nice tutorial i just haven't ever had i'll do this i'll do this tutorial i'm

so like you know if but you know velda comes along with i actually use your product

the things that bugged me and these are what i'd like to suggest that is valuable what what i find

frustrating is when people turn up and they've got no familiarity with the project they're not

using it and then they suggest something which you know these increasingly these days has come

out of a machine it's like no that's not helpful um the builder from our own experience turning up

official blog post on client stuff using open source tools referencing neapolitan

or make a br to his whatever documentation and get it done and then when he saw you at the at the

conference before i did he's like that's the person you should drag and i was like i just

i'm meeting him for the first time i don't want to be like hey i hate you don't be shy don't be shy

newsletter