Welcome to svelte radio hello

everyone welcome back to another svelte radio episode i almost

said Svelte episode radio. That'd be wrong. I'm joined by my two co-hosts today. Anthony,

Brittany, hello.

Hello.

Hey, we're finally back,

the three of us. Back all of us together. In one

episode. That's amazing. And today we also have a guest, Ilya. I think that's how you pronounce it?

Yeah, right. Exactly. 100%. Perfect.

I did it. Great. So, Ilya, you are a

Svelte developer or a developer in general, I suppose, right?

You're very active on the Skeet site, Blue Sky.

I love it.

I mean, I really do.

It's so nice because you share a lot of stuff,

not just for Svelte, but also like some Astro stuff

because I think you use Astro.

Oh, yeah, right.

So that's nice because I have no idea what's going on in Astro.

But we invited you to talk today about NGOs and working with NGOs as a developer and stuff.

It seems like an interesting story compared to everyone else that's just working.

I have a dumb question.

What's an NGO?

Oh, right.

Perfect.

I recently gave a talk on this topic and I had a slide.

So NGO is an acronym.

It's for non-governmental organization.

and this is what people most associate with what we actually talk about which is npo and non-profit

organization because every company technically is also an ngo but yeah it's it's a bit more about

like non-profit so civic society foundations so everything which is not governmental and usually

not a company cool

so so like organizations that help out in like disaster relief or like homeless

people or like stuff like that or

wildlife saving

wildlife

just some examples i

guess

this is i think this is the more like uh yeah applied or practical um non-profit organizations

or ngos i mean there are many um mostly they are advocating for something i mean even like

like like greenpeace and its core is like a like an advocating or a company or organization which

is doing advocacy.

So

not all of them have like boots on the ground,

but

as many people with

some expertise

where the government is for whatever reason

not providing the type of service

or knowledge that they should be.

Yeah.

So we all met at Svelte Summit earlier this year.

That was a lot of fun.

And you talked about you wanted to do,

well, actually before I get into like the meetup thing,

Maybe you can just present yourself.

Who are you?

Other than working

with NGOs.

Yeah, right.

Yeah, my name is Ilya.

So I've been living in Berlin now for like 30 years.

And I switched basically careers like multiple times.

So like after high school, I joined the army for a while.

And then I quit the army.

And then I went to university.

And it's a bit similar to this USGI bill.

So basically, they paid for my bachelor's.

And that also means I can only pick one once.

And I chose a social science one.

And then towards the end of this bachelor's,

I decided that I don't want to pursue a career

in political science or something like this.

And then for my master's,

I found a place where I could do computer science as

a minor.

Interesting combination.

Yeah.

I mean, they have a term for it.

It's called computational social science.

It's basically the major is in social science.

And then you apply big data methods.

Back then, this wasn't like a big thing.

and then eventually I started doing like data visualizations as a freelancer basically also

already for NGOs and this like one thing led to another I started building interactives I started

building websites and now yeah I'm also self-hosting my stuff

nice nice

yeah we talked right

we talked

before the show started about self-hosting and and the cloud by the way by the way yeah yeah it's

It's nice.

Yeah, so I just wanted to mention,

me self-hosting did not

save me from the Cloudflare

outage

because I'm using Cloudflare tunnels.

So I'm tunneling all my traffic through Cloudflare

so I don't have to expose my IP and stuff to shady people

that want to DDoS my server.

So now I'm thinking, should I just ditch Cloudflare as well?

I don't know.

Maybe I should. Maybe I should.

So how do you, so you studied social science

and then you moved into computer science

and then data visualizations.

And I assume you found Svelte somewhere around there,

like in making visualizations or when did you discover Svelte?

Yeah, I mean, there are always a bunch of like path dependencies.

And one thing is I first heard of Svelte from Gregor Eich.

So, you know, like

the co

-founder of Data Wrapper, who

also presented

the Svelte Plot Library at the Svelte Summit.

Because in Berlin and many, like many, I think, places where there's lots of media, there is a meetup series, which is called Hacks and Hackers.

And it's developers and journalists meeting.

And back then, I was basically just following what they were doing, like interactive journalism, you know, all those impressive visuals.

And my idea was, how can I apply this in the NGO sector?

So because they have usually they have those policy papers, those reports, it's like PDFs with 30 pages or something like this.

It's just text.

But there is a story and there's data.

And how can you like help them make this become alive?

And he mentioned Svelte.

I didn't use it back then because I was just using jQuery, actually, and

11T.

And I looked into it.

As long as it works,

right?

Yeah, but it was, you know, it was still painful.

then i think i compared like a tiny component to alpine and importing alpine as a library was like

the library itself was bigger than a compiled swell component and i was like okay so that's

it basically so this is what i'm going to use and then i think during covid i also was in an online

workshop um from tan liao he he

did a swell

workshop it must have been like i don't know

five years ago or something like that and then this is how i how i picked it up because it felt

very natural to me because I knew a bit of like HTML and CSS already and you know Svelte's platform

first so it just like totally felt natural to me to

start

using it that makes sense makes sense

did I hear that right though that you were using Eleventy which is like a no JavaScript

framework and then you were putting jQuery into it

yeah but like Eleventy uses JavaScript to

compile to build a website or back then it was

so yeah but like yeah exactly no javascript on the

front end right

yeah yeah yeah and now i'm using astro as well

okay nice right

so so from from

so if

you go from from like visualizations then how did you find like how did you pivot into

ngos or like how did you end up building websites for ngos

yeah doing work for um

right so i like i didn't think about what i would be doing during my masters um and i had like two

job prospects or job opportunities and both would have been uh were related to the government and

both did not happen for reasons you cannot like imagine like in in one case there was like a

government affair on federal level and i was supposed to join as a contractor and then they

just not like didn't what is it like pay pay the money or

make the money available

for this

and the other

one is we had um we have like a think tank which is government sponsored let's

call it government sponsored

and

for it to have a budget there needs to be a government and we had

like government you know coalition negotiations and then one party withdrew and then again i knew

oh shit my job my job entry my post army post bachelors post masters entry in the market it's

not going to happen um and then coincidentally because i was also very online back then on

twitter someone who was like basically working in this space with with a vision for you know new

digital things they invited me and and just introduced me to their to their boss and i was

like oh i need maps for my book oh wow i i created maps for the book and i was like oh we need more

maps and visualizations for a report and i did those as well and then i was like okay and now

we have additional funding we would like to put it online

right i

mean i know how to put an svg

online i mean let me figure out if i can also like you know wrap it in some html with some text

and

this is this

is how it started right and like yeah this was like with eleventy and i figured out

a lot of things, learned a bit, you know, about

accessibility and responsive

design.

And, yeah, and then one thing led to another

and I was booked for other websites, web products.

So I didn't plan for it. It's just like

I

accidentally became a

web developer.

Right.

And

like this was

one

pillar

and the other pillar is because I was then

implementing interactive

visualizations, media companies

or people who are doing data

journalism they noticed me they saw this on twitter i was like hey we we need support for

a project would you like to join and like you know work on a particular um feature or component

or something like this and this is how i basically started working with with svelte as my framework

full-time

right nice nice and so so what so was it uh was it like the small bundle thing that

pulled you in initially it sounded like that when you compared it to alpine earlier

this was one like one heuristic because i had no clue like how cdns or something like that works

i was like oh no i need to get this bundle size you know down as much as

possible that

was one

thing the other thing is that um as a kid i learned html and css a bit and i was running a website but

that was like early 2000s or so

yeah and

then i didn't touch any of that i'd never learned javascript

or something like that and then when i i'm during my bachelor's i had a student job and i had to write

and CSS. So then, you know, it all came back. And then when I started working on interactive stuff

for the web, and then I saw React, I worked with React like on two or three projects. I was like,

what, what,

what the heck is

this? I mean, there's HTML and there's CSS. You just need to add some,

some behavior, you know, and there is

this,

there is state in a variable. Why is it not rendering?

Why do I need to do extra work to, to render a value of a variable?

And which felt it was

exactly

like this right yeah

yeah just like you have to have i call it a javascript brain to like

understand react you have to think about things from the javascript point of view but it's felt

like you're thinking about it like html css you have your markup just separated and you keep it

separated in your brain

yeah yeah totally agree and

then with um especially with the accessibility

linter and with all those you know bindings on on the dom element like with the the attributes

binding the disabled state binding all those bindings it felt just like so natural it's like

literally i have this state and i just want to be the state applied here and so just along the

ergonomics of writing it felt like super natural because it was just like okay i know html i know

css i just need to some state just a tiny little yeah

how do you feel about back then

um it took me two or three months to adjust but i got i have to say um matthias style from spiegel

um and i was working at the media company like two years ago i just started there and um like

the situation was very intense because like one person quit and there would be four elections

so i i

knew i

was like in for lots of trouble and matthias in april or so april 2004

And four, three, four, four.

He posted, okay, Spiegel is using Svelte 5.

And I was like, okay, I learned so much from him.

If he is going all in on Svelte 5 while

it was still

in beta,

I'm

going

to adopt it as well.

And then it really paid out.

So I'm totally happy.

I'm super happy now.

First two months were a shock, but now I'm a

believer.

Yeah, I think I had kind of a similar reaction as well.

especially when

they eliminated getter setters

like the explicit ones

there used to be

an explicit get set syntax

right, yeah

and this was the moment I was like, okay

you mentioned Matthias Stahl

he actually used to live here in Stockholm

and I met him on the very

I think it was the very first Svelte Stockholm meetup

that we hosted over here.

So that's a fun coincidence.

And now he's at the Spiegel doing data viz.

And I think he also came to the first in-person summit, no?

I don't remember.

Maybe.

Yeah, I don't know.

That was like so long ago.

And I was also very stressed out.

Didn't

he give a talk?

Yeah, I have a fear because, you know,

I remember saying to him that I was literally

having this really weird heated debate or watching heated debate and just joining in now and then to

like troll the people like to fan the flames a bit because i thought it was amusing on twitter

um but it was between dr matthew stall is that on twitter i don't know is it real doctor or not

and someone

else yeah

and and so basically i said to him at the time may have been the first summit

uh like by the way i'm just trolling you on twitter nice to meet you kind of thing

something

along those lines you know so

maybe it was the first summit where that when it happened

yeah i i don't remember i'm my my brain was totally fried there were a

couple of talks from

like newspaper organizations at the first summit and i just i was thinking that they were one of

them

no i so so i think that was bloomberg that was julian and

uh okay yeah yeah

britney

britney

yeah um yeah i i i don't remember but yeah he's he's a it's a really nice guy um you should you

should never get

that name wrong again because i introduced her as like some other name for some

reason and

of course

i get my own name wrong

yeah it's not easy it's not easy doing hard work you

know it's a lot of names to keep in your head and

like oh it's read the correct

notes and

stuff

And make sure you're on the right line.

Yeah.

So I want to mention this early in the episode

because if people drop off,

but you're doing a meetup in Berlin, right?

Maybe we should talk a bit about that.

You just

announced it the

other day, I think.

Yeah.

But I'm going to fully disclaim

you pressured me into doing this at the

summit, basically.

Since then, it's been like a heavy rock on my shoulders

to kick it off.

I

think I shamed you a bit as well

on Discord,

like, oh,

you should stop tweeting so much and do the meetup.

Exactly, exactly, yeah.

All

in good fun.

I don't want to really pressure you.

I mean, in August,

this thing happened where,

for whatever reason, Paolo asked me

on Bluescape if I would like to join the ambassadors.

And I was like, I had

real imposter feelings because

all the core maintainers are there,

and I don't contribute

in terms of PRs or something like that to Svelte.

But then I was like, okay, one thing I can do

is actually spread the word

because every time I talk to people about Svelte,

like afterwards they tell me,

man, you really, it's like, it sounds convincing.

You are like hyped.

You love it.

Sounds good.

I'm going to try it.

And they try and then they start using it.

And then they was like, oh, it's awesome.

And like, yeah, so this is how I like also like,

actually once created my own colleague

because I told them to start learning it.

And then like a year later,

I was able to, you know, co-work with them.

So about the Svelte

meetup,

December 9th should

be a

Tuesday.

We will be hosted by Data Wrapper in Berlin

in their new office.

Nice.

And we will have three talks.

And it's not only me doing this.

It's also Peter from XYFlow.

So basically the person who ported

and implemented SvelteFlow,

this graph visualization, graph library.

We'll have three talks.

One will be Peter introducing SvelteFlow,

talking about in detail,

and also doing like a little side extra discussion

comparing Svelte to React

because they are maintaining two libraries,

one for React, which is the main one, XYFlow,

and the second one is SvelteFlow.

So obviously they have real-world benchmarks,

what it means to basically implement the same library

in two different frameworks.

That's awesome.

We are still in discussions about the second talk,

which might be big, but I can't reveal it.

And then depending on how much basically time

this person would need for their talk,

I will either jump in with the talk or not.

And I would talk about this community survey app

I built with SvelteKit, Superbase,

some BAN middleware, Paraglide, all this stuff.

So there will be three talks.

There will be pizza.

Registration is on the Svelte Guild, Guildhost.

Yeah, we'll have a

link.

in the in the network and on blue sky and we are like we have like i don't know like 25 seats maybe

and 20 or 21 already signed up so yeah get the

four four last seats might actually be full by

the time this goes out next next thursday so hopefully not but we'll see we'll see and yeah

but we have no idea right

exactly exactly but we

have no idea like

how many people are in berlin

And so this is like where we are like a bit cautious, you know, about like, do we get a full room or not?

And now like within, we just announced it yesterday.

There were already 21 people who basically signed up within 24 hours.

So let's see.

I think Svelte is popular.

Yeah.

I think especially in Germany and Berlin area, like there's quite a few companies and things that use it in that area.

So I think you'll get quite a few.

Yeah.

How is the Berlin tech scene, like other than Svelte?

Is it

a lot of meetups

or?

From what I hear from other people living in other places,

it might be a bit similar.

So before COVID, there were like,

you could go to a meetup every day somewhere.

There would be one on JavaScript, on TypeScript,

on map making, on whatever, on every topic.

And then with COVID, I think there was some like,

what is it, like it hit the dent or so a bit.

Now it's recovering.

And yes, there's like quite a big tech scene.

So, you know, we know Git Butler has an office in Berlin.

Data Wrapper is in Berlin.

All the media companies are there.

All the consultancies are there.

McKinsey, Accenture, whatever.

Plus we had like three or four different bootcamps.

And all the people from those bootcamps,

they basically joined all those, you know,

mobile first companies.

The one

for Lieferando, Amazon.

So there are like, I don't know, thousands of developers actually.

Yeah.

So it will take some time to convince every one of them,

each one of them to join and start using Svelte.

But yeah, I think

maybe...

One person at a time.

Right, exactly.

And then we'll take

over the work.

Use

a megaphone and go to the canteen and just shout.

I feel like, wouldn't it be a very Berlin-ish

to do like a Svelte Tech rave meetup or something like

that?

Yeah,

right, right.

I like that.

With some

techno,

obviously.

I'll

put my flights.

Oh, right.

You're an old rave person, right?

An old rave person.

Well, I mean,

I meant you used to go to raves and stuff.

I used to go to

raves.

I mean, I still would

if I didn't have two children after all time.

Yeah.

What's a

rave like?

I've never been to a rave.

What's a

rave like?

uh how do you describe it it's i don't know it's probably not for everyone the music's

loud music it's very

dark and often feels kind of cold in there till you get dancing a bit

and then

it's kind of like i don't know it's a

weird experience if you could explain it maybe

it wouldn't be as fun right so it's more that you can't really put you i can't put a pin on why it's

good it's just kind of good i mean it's basically like a nightclub vibe but different it is but

Like, yeah, I feel like a lot of nightclubs are cheesy

and kind of like, you know, the music's not as great.

And in a rave, you're kind of in that zone

and the music continues and stuff like that.

Yes, pretty good.

Yeah.

Yeah.

Yeah, I mean,

you can

simulate it.

Like, you know, just code until 3 a.m.,

put on some music in the background and dim the lights.

Or

I don't know if you have like...

Yeah, I don't know about the raves you go to,

but the raves I go to don't feel like that.

I feel like you need

those Philips Hue lights

that blink in different colours maybe

and then you can have the music on

the strobe lights

right

smoke machine

is that what it's called?

smoke machine, lasers

you need the people

you need to be surrounded by people

just invite all

your friends

have them stand

around you

while you do coding

but I think

but I think

most people around you

they need to be

in varying states

of consciousness

in order to make it

feel

like a rave

as well

fair enough

it sounds

like

we should have a rave

at the next

World Summit

I think so

I think so

and I think

I already

know

who the DJ will be

right

who's the DJ

going to be

I can't DJ

by the way

not in my skill set

we'll invite

Brady

of course

the

guy

that made uh the svelte radio intro song that's true that's yeah he's not been to one of the

svelte summits yet

no so i

think yeah i think i hope he can come for the for the next one that'd

be fun

he can do some live like

live production like

you gotta have a break from touring

yeah yeah yeah um okay sorry so a bit of a tangent there

sorry listeners

but so so okay so what kind of stuff are you building now for ngos like give us some examples

and what what how is it different well i don't know how much you've worked with like regular

business clients and then ngos like what's the difference yeah

so oh there are so many differences

i feel it's like um it's a bit like like really was it like a stark contrast basically between

working for those two environments but then also i met some employers where they are like right in

they feel like they have a for-profit mission.

So they try to earn money,

but then they work like an NGO or administration

or something like this.

So I think

I've seen basically the three worlds.

I think one thing is that the first thing about NGOs

is mostly they don't have their own funding.

They usually depend on grants, on donations.

This is why they are not for profit.

They don't create any revenue, right?

And there are two things.

One thing is that they don't have a product to sell.

They have insights to sell or they are trying to advocate for something.

So then they are not trying to sell something.

They are basically trying to pursue people and learning about something or understanding it or trying to see it from a different point of view.

So work for them mostly is like story first.

So you need to figure out how to help them to tell a particular story.

It could be like a report or for a website.

basically you need to help them figure out like to present themselves so that like especially if

they have different audiences because usually like like one NGO I'm working with right now they have

two audiences one is their community because it's like it's the community basically of um

black BIPOC people in Germany and first of all it's like a huge community it's like one million

people that they estimate like Germany has

80 million people

and and and they they assume or

the estimate is that like one one million has like some some some BIPOC roots like African German or

or African-American.

That's one thing.

And one thing is obviously they try to create stuff

for their own community, for themselves,

but then they also need to attract funders.

So basically, you know, was it like white-collar people,

people

who are not

black,

and people who are maybe like super senior

in some federal administration or something like this.

And you need to figure out how to make a website work

for both of them, for both of those communities,

But, and this is the third thing, since they don't have their own budget, they usually have project budget.

Project budget means they have applied for funding maybe a year ago, maybe two years ago.

So whatever you are doing, like then when you are implementing all the preprocessing, all the preparations happened two years ago.

And you know what

happens like in the

tech world within two years, right?

And so they arrive and they're like, hey, we have like a budget of, I don't know, like 6,000 euros.

Can you get us a website and a dashboard?

And we need a newsletter and this and this and this and this and this.

So

obviously, this is why you have to use the most efficient technologies you can find.

And this is where Svelte comes in as one of the most efficient pieces of the puzzle.

This is, I

think, one of the biggest things that you need to understand first is that

it's not that we have a project budget, which we can adjust.

And the deadline basically has been set two years ago.

something like this usually that's one thing and the other thing is they are not used to doing

product so they don't have like a po or something like that if you're lucky there they have maybe

they have like a fixed was it um in their organization maybe they have a person full-time

responsible for communications if you are lucky

that means

that this person knows what the jpeg is

and that you can't upload five megabyte images but usually they don't and usually don't they

don't have their own IT department they like don't have a domain or something

like this so you need to set up everything for them and the other thing

is it's project based so that means that if you launch a website but funding is

done you get paid once

there is not no

such thing that they are going to pay

you for the next four years to maintain it so

first of all

you need to negotiate

it in advance but second you need to make sure that your website still runs or

can be deployed in four years

with tiny updates maybe right right so interesting so so you're

very budget constrained and you know you you have like a hard limit compared to like

if you're working for a client at a at a company doing like a marketing website or something

right

exactly but then on the other hand what i see in comparison to like for-profit companies

says usually the deadline is not really a hard deadline unless there is like a particular date

that they want to launch a report website because there will be some summit.

If this is not the

case,

there is no real deadline. And that means that even if the budget is smaller, what you then have

is time. And this is where basically how I, I think, grew into this full stack developer role

and why I had this opportunity to try out everything is basically, okay, look, your budget

is very limited but if you give me three months

of time for this and we do

like a weekly

iteration that means that

I can study some super base I can study

some SQL I can figure out how to

like what stack we could use

and then it becomes

a bit more like a you know like I don't know like a

paid

traineeship or

something like

this and then the day rate

can be like 500 or even

less

right if you're lucky

you

can get maybe like a 700

euro day rate or something like this

but you can figure out what to do

how to basically max out on time.

Yeah, interesting.

That's a very different style of working with clients.

I've only

ever worked with like paid,

like, well, not paid, but like

businesses,

right?

So it's very, very different.

So what kind of stuff have you noticed that are like,

so I assume since like the budget is constrained,

you kind of have to prioritize, right?

What do you often work with like minimizing the scope,

changing the scope of what the NGO wants?

Because you mentioned like, we want a

newsletter,

we

want blah, blah, blah, blah, blah, blah.

But then like, there's only so much money

to implement stuff, right?

So what, in terms of like, yeah, where would you,

not cut corners, but like, what would you prioritize

in your experience?

That's

a tough question because it depends.

Right.

Now, lucky me is that I also did this

like one design thinking certificate

and then it helped me with some,

I mean, I know like design thinking,

it's very similar to human computer interaction

or something like this.

It's like, you know, this marketing buzzword

for a particular way of thinking about

what you're going to do.

What I'm trying to do as a principle is,

you know, there's this project triangle

product triangle was it feasibility desirability and the third pillar feasibility desirability

so does it make sense can we afford it and what do we actually want and

what i'm trying

to do with

them is to ignore feasibility because we will find ways to to make it feasible so to implement it

and focus on desirability and then they have many many ideas huge wish list second thing is that we

do is we prioritize which is like essential for the launch for product launch which is essential

to achieve a particular goal and everything else is basically on top if we have time or if there's

something that i want to basically learn in my free time then then we're gonna bake it in um but

otherwise we are trying to focus on the core and this is also like one thing that you need to to

teach them or to figure out with them is that for themselves they need to figure out what the unique

selling point

of something

is or what the core is.

And they probably don't know.

Right, exactly.

And because I think in nine out of 10 cases,

it's academics.

It's people with a political science degree

or public policy degree or something like this.

They have so much knowledge in their brain,

but no one ever taught them at uni

or during their PhD or whatever,

how to tell a story,

how to focus on the essential parts.

And basically, how to tell them

how to turn a page into a single paragraph.

Right.

this is actually where the most most of the work is figuring out the story and then you realize

okay maybe we don't need this feature it doesn't make sense and if you like also all those techniques

like how might we or why why why you know the why why why method like ask someone why they want

something three

times

like why right and then it will be most like nine out of ten times it will

eliminate itself and

the feature

will disappear

yeah it's a difference between it between like a

highly technical person or a discipline of being highly technical versus being uh you know product

orientated and i think that there's there's a huge void there that the people that you want on your

team are not necessarily the most clever or technically adept people are the people the

biggest product focus when you're a small company or when you're a small team so

maybe there's a

there's a comparison to draw there like between very very small companies

and ngos

like because

there's only so much knowledge

that you can have inside of a small company, right?

I mean, sure, there are like geniuses

in small companies that know everything, of course,

but I assume that's the case in NGOs as well.

One thing I notice is I think in companies,

for whatever reason, even in small ones,

there's still some type of like a division of labor

or hierarchy.

And

they have like at least an implicit understanding

of who decides and who maybe can veto something

and

who is basically

just there

so that their advisory can be heard,

but it's not a blocker.

In NGOs, maybe because of this community spirit

or something like this, it's usually very flat.

So there are many people who can veto something

and there are also many people who

can voice ideas,

right?

And yeah, this is the hardest part basically

because sometimes it's like contrarian opinions

or contrarian wishes.

And this is not, you can do the regression to the mean,

then you will have like a mediocre product

or you need to kind of do some Game of Thrones thing

behind the scenes, like a

particular,

yeah, yeah, it's lots, it's still,

it's lots of politics at those companies.

And it's personal because this is also one thing

at NGOs, people very often are very like personally invested

in a particular topic.

Yeah, they care a lot.

Right, they care a lot because it's either they are impacted

or it's someone whom they know who is impacted

by some social policy or so everything is quite personal like

lots of this

stuff this is why also

they like very often burn out

because there is

no separation between work and and free time

right so yeah there are like many very special

things

but i think

it prepares me for the for

profit market

if you're if you're very invested in something as well but you but the company

structure means that you can't change it that's also quite tiring and it can burn you out very

quickly right so you know you if you're given and i guess in my world i would call that that being

invested in something ownership um if you're given ownership of something or you believe you have

ownership of something you should be able to change it at your will without having to go through lots

of red tape which is probably what i call the other bit i know flat structures generally i i prefer

flat structures and they work but i've never really tried a true flat structure at scale

and i think that if you try to this is what this is what i i interviewed with meta i mean that was

you know i was young and i needed the money and i didn't i was just i just did but one of the

things they told me was that basically the whole company is flat and and if there's any problems

you just you go straight to zuckerberg and i said sure that sounds

that sounds like what happens

there that sounds that sounds realistic i believe you

and it's that kind of thing

is like um you know

you could you could it's you were laughing right because there's no way that works right that

doesn't work right it's just nonsense so i think i think that yeah ownership

is

why i like to work

small companies because you can get ownership you can move things you know you may even still be a

cog in a wheel feel like a cog in a wheel but as long as that cog can turn whereas if you're not

it will turn that cognitive

yeah i there is a counter like example of of like a company

large company that is i mean not as large as meta but valve right they're they're supposedly a very

flat uh company and that but they also kind of like we work on whatever we want to work on which

is very nice that's

kind of nice that's kind of nice and and you know if i believe that kind of

company works and i'm gonna get shot down in flames this but that's that's apparently how i

understand that palantir works as well right is that you kind of work on where you think your

efforts could be best put and the way that the reason that works is because the people they hire

are the kind of people who can do that it wouldn't work in every company where you know people are

hired uh let's say with a lesser bar like a lower bar to get it to entry because those people wouldn't

be able to correctly prioritize and be able to correctly figure out what is and isn't important

it's a certain caliber of person and i think that maybe valve has those people right and that's and

that's great i love i love that right you know complete autonomy and if you've got complete

autonomy like that and the people are right then you can have a flat structure but yeah i don't

think anyone believes it can happen at scale the scale of meta even though i'm sure meta has a high

as well but still it's just so there's so many people

right yeah someone has to to like make a

decision somewhere right yeah and

it's not always stuck

right and

the thing is you can also slice

a company up to provide more autonomy i worked in sky which is a huge company our team had autonomy

but our team was like 16 people there were other teams that had absolutely zero autonomy and it's

just because the way things are sliced and dice and i think if you if you can slice and dice it

even within a large organization where you have full control and we had to break down some barriers

like they had a provisioning computer provisioning thing i joined the company as a contractor getting

paid a load of money and it was going to take four weeks for them to get a computer through this

provisioning service and it was going to be just a regular pc that would cost like 800 quid at the

shop and it was five grand because it's being provisioned via british telecom or some crap like

that and they just went they went to pc world and they expensed the computer for me and they bought

the computer and put it on my desk within the same day right and it's that kind of thing that if you

can do that it wasn't really allowed but if you can do that they can work in a big company you can

have the autonomy and the ownership and you can control things yourself yeah like there were teams

just wouldn't do that and be like well you just sit there for four weeks getting paid do nothing

whilst we provision a

really expensive

computer for you

sounds very boring

as well very very boring

very slow nothing changes you get disrupted easily that kind of thing so that's why sky were doing

this thing where they where they would have this autonomy in certain teams to try and change things

and it you know it works it works it definitely works

yeah so ilia we met we mentioned kind of

struggles of working with an NGO?

Well, some of the hard stuff with working with NGOs.

What are some of the good things that you've

experienced working with these?

I think it was super helpful to hear Anthony explaining his

past employers. Because that's the upside.

If you work for NGOs, you are usually the only person.

and you are the only technical person.

And also like the way I worked is like,

I also did this product vision design creative thing.

So basically also, you know,

like leading the story or trying to come up with a narrative.

First of all, you can be fully creative.

If you are a creative person,

there's basically no limit what you can contribute to them.

And obviously as a tech person,

you can preserve your sanity because you can choose your own stack.

You know, there's, you are not inheriting some,

some React code base from 2019.

You are writing from scratch

mostly,

right?

You will start from scratch,

but there is no tailwind.

You just write the stuff

the way you imagine it should be written.

And I think this is like also,

especially now with Svelte,

with the rapid development of Svelte,

that means that you can try out

the cutting edge stuff.

You can try

out Svelte

5 server-side rendering

with Bun or like

the thing that

I tried out

is Svelte in Astro server islands

for pre-rendering Astro content and it works.

And I can just present it to them.

And then look, we have something,

no one has done it before, but it works and it's awesome.

And we can make your landing page

display for 15 events dynamically from a DB

and you don't need to redeploy on every event

every time you

edit an event.

So this freedom gives, like, I don't know,

it gives me a lot of joy.

Yeah.

Yeah, I can imagine, like, being able to just

also experimenting, right?

With new features and things that you might not be able to do.

And if you're stuck on Svelte 4, for example,

which some companies are, right?

I think

Brittany, you

mentioned like a while back

that you guys were stuck on Svelte 4

and

just upgraded to

Svelte 5 or something?

My previous company,

we were stuck on

Svelte 3 on the main app.

And then, yeah, they couldn't even upgrade it to Svelte 4.

And then we started working on a separate project in Svelte 4 right before Svelte 5 came out.

And that one, from what I've heard from people that still work there, is still not

migrated.

They still haven't worked on this React project they were going to switch to.

But, yeah, the new company I work for, they have like an old plugin that uses Svelte 4.

And I'm migrating it to Svelte 5, but in a completely new repo.

we didn't

go

through the migration just working on it

completely fresh so there's like that that

that's usually what happens in in a company you're you're stuck on an

old version of something

yeah

like anthony and zapper well have you have you do you have

news and we're still doing it we

are we

now have

a dedicated resource basically and it's peru right who's who's

migrating this page

by page

yeah he's getting there there's a

lot to go

but he's getting there

that's good

Sapper is like ancient at this time

at this point right

yeah well yeah

I bet it still runs

though

it still runs

it still boots really slowly

still you start the app up and it takes like

20 seconds to

start maybe more

was it Finn McCann that like had a bunch of stuff

on Sapper for a long time

I don't know

possibly

who

does ElderJS

Oh, that's Nick.

Nick.

Yeah.

But

wasn't ElderJS?

No, that's ElderJS

is like his own thing.

So he

built another framework.

I swear somebody was on Sapper for like a really long time and had,

I don't remember,

a bunch

of sites like just still running it.

I mean, if it works, it works.

Oh, yeah.

And

it does work in production.

It works, right?

I think Decathlon had their whole site written in Sapper.

I think they were using Svelte.

I'm not 100% sure they were just using Svelte and something else.

But they rebuilt their site, which was really creaking.

And it was amazing in Svelte.

And then they recently rebranded and built it again in something else.

And it's shit.

It's just terrible.

Like, I love the cafe.

I think they're brilliant.

But I don't know why they decided to move away from Svelte.

But their new site, it just sucks.

It looks like Shopify, like it's that bad.

Maybe it is Shopify.

Yeah.

I mean, again, I

don't think Shopify is necessarily bad.

but I think it has a certain format if you don't configure it.

And I think it looks like that.

It's just very, very basic.

Yeah, it's probably, which in itself is, right.

We're getting down a

tangent here.

But my

point being that it looks like a basic

and configured Shopify instance, and it operates like one.

And you used to be able to really dig into the search a lot

and find things you wanted.

Now you can't.

Yeah, yeah.

You just can't.

This discussion kind of brings up another thought.

So working with these NGOs, I assume you don't have to really worry that much about like, oh, in the future, we need to upgrade to Svelte 6 because they're probably just going to stick to the website that you build.

Or like, how does that?

Like, of course, you want to build something that you can maintain, right?

Yeah.

Since it's project based,

it must be different.

both questions are very good especially britney's because the oldest website i built was like in

2020 or something like this like a whole website which i built and i don't need to redeploy it

the only thing that breaks is if safari releases a new version of the browser then something always

breaks for whatever reason um like one website i built i think in 2021 the academic freedom index

um their pair contract i need to do daily updates because every year there's a new index

and I went actually the whole story with it so far.

It's like it's Astro,

but then from Svelte 3 to Svelte 4,

from Svelte 4 to Svelte 5.

And I think like that there will be no big changes

between Svelte 5 and 6

because the focus is now on SvelteKit,

which brings me to like the most up-to-date topic.

Right now I'm helping an NGO

basically relaunch their website.

And this NGO has existed for almost 10 years.

They have like 500 sub pages and content collections

in directors.

It's like events, projects, people, pages, all the stuff,

like 12 different content collections.

And because, first of all, it's a mostly volunteer-run organization,

so they, again, also don't have permanent stuff.

Right now they have a tech person,

and this tech person is the only person who knows their way around in directors.

And we're going to leave it as it is for now

because we're focusing on the website relaunch.

but right now we are migrating away from SvelteKit 1 to Astro

just for the sake of the stability for the next year or so.

Like I think in, I don't know, like when SvelteKit 3 is out,

I think if it would have been out, I would have made the decision differently.

But right now, especially because I want to basically make sure

that they can operate next year on their own

and do

tiny updates,

I'm migrating them to Astro.

I think what I like about both SvelteKit and Astro is the way how pages are structured, how slacks work, file-based routing, it's actually very easy then to translate back and forth

as long as your

content collection is somewhat stable or in an organized state.

So there is no big loss for them right now.

And those pages are mostly either fully static or they just need a tiny dynamic part.

Yeah.

So, yeah, I mean, right.

Astro works and SvelteKit works as well.

Those websites where you need server-side rendering

or something dynamic,

obviously those are a bigger concern.

And yeah, I need to make up my mind yet

because this year I built this whole ecosystem

for a community survey app.

It's like everything.

It's like the Superbase with database.

Everything is self-hosted also for privacy reasons.

And I'm operating the server

and I will continue doing so next year.

again there's this opportunity as soon as all the breaking changes arrive in SvelteKit

I can test it out for myself and I can use this production app basically obviously on a

development branch right to test all of it to figure out what works what doesn't and if there

is something broken you know I can open an issue or talk to the maintainers or something like this

and this is I think also like something where I can I can just create my own time you know take

my own time to do this because i can kind of you know control the timeline when things have to

happen and i'm not sure how this works in the company unless you are very very senior and

you know your your pm or whoever is

like

trusting you a lot

i mean in in a company you you always

have to it's all about profit in the end right like it's

something costs

money you have to earn

money so there's some priority of like and even like an upgrade

has to have like some cost value

to them to be able to get it done yeah and with the work that you're doing with the ngos like

yeah it's like you have this free playground but you also have consumers of the app you have to

think of and like you said you're putting them on astro for stability so that they have access to

the things that they need when you're not there anymore

yeah um so any other thoughts and questions

and stuff about NGOs.

I've run all out of my questions.

I can add one because you asked about the upsides

of working with NGOs.

And there's like this obvious thing is like you are helping

usually good people do good things.

And

if you help them, sometimes they have a very particular idea.

Like once we built a website for a basically meetup

with parliamentarians because they wanted to impact legislation

with regards to Telegram.

So we had like a hard deadline and then we launched the website exactly for this event, like in particular for this event.

And so they succeeded.

This legislation then followed or

it happened.

Oh, nice.

Right.

Or sometimes you enable them to do, you know, like even just a simple, this is your own self-hosted newsletter or management thing.

And they're like so happy because, you know, they promise to their community that this is like privacy first and nothing, you know, is collected by other companies.

And then they can follow up and basically deliver on this promise.

Yeah.

And they don't have to pay thousands of dollars for an email marketing solution.

Especially

like this big project, which is called the AfroZensus, which was like a large

end survey of this German African community.

This product, we relaunched it.

It's still online.

It's been online for three or four years.

And this is like also a dashboard for the community itself, because all those people, many of them

have suffered discrimination

or they have been,

you know, I don't know,

like been discriminated

by the police

on the housing market

and whatever.

And this data set

that the motivation

for this dashboard

was actually to revalidate

the people,

the people's experiences,

you know?

So we are not selling something.

We are basically

just helping people

to understand

that what they are like,

what is it?

Living through

is not unique to them

and it's actually a thing.

So there is

discrimination

in society

and they are validated.

and just this

I think actually it uplifted the whole

community you know and

then they

basically think about bigger plans

and bigger ideas and they have this thing

that they can use for advocacy

they can go to politicians show them the

website and the dashboard and I was like

no we are talking about thousands of

cases of discrimination just

by police forces or something like this

right yeah so it's very rewarding

actually

yeah that does sound very fun

well I mean not the

issues themselves, but rewarding.

Yeah, yeah, yeah.

All right.

So I think we, unless anyone has another question,

then we move on to the controversial parts of the

episode.

Oh, no.

So unpopular opinions.

Do you guys have one?

Who wants to go first?

Anthony?

About self-hosting.

It's going to be a health leader.

I have a bond,

yes.

Let's hear it.

It's not that controversial.

It's not that controversial.

It's just something that was relevant to me today,

so I was thinking about it and made people disagree.

I'm sure people can argue over it because they argue over everything, right?

My controversial opinion is that people should store states, not effects.

And by this, I mean something like if you're dependent,

if you're wondering about the state of something being approved,

the effects of something being approved is that it gets published or promoted or featured, whatever.

There was a lot of discussion with my team today about what it should be called,

whether it should be called published or published at the date of when it's published,

or the designer wants to call it featured because she's going to use featured stuff on the front end.

Promotion, because it might relate to promoting things at some point.

And all this kind of backshedding about what the name should be,

it occurred to me that the actual thing here is a state of something being

approved or not approved is what we should be storing because everything else

is kind of like things that you can do with that thing because it is approved.

It's just the effects of it being approved rather than the actual state of the

object.

So we store approved in database or approval and then we store who approved it

and then we store a date and everything else is kind of like hanging off that.

So

store states,

not effects.

interesting i

thought you were gonna talk about uh effect.ts like no

no no type script library

as part of our huge migration to salt 5 we are um i'm now learning about effect and everything else

and it's all interesting stuff

all right does anyone else have a have a

and unpopular opinions

Brittany no

I am

trying to think

I'm

in a pretty good mood

today

nice

yeah I mean I

I have my standing unpopular

opinion about more people

should self host

by the way

by the way

do away with all the CDNs

yeah

so you don't go down when Cloudflare goes down

have you seen those memes online

so you've seen that

XKCD comic thing right

where it's like

boxes and stuff are stacked on top of each other

and then there's like this small

small box at the bottom

right

there's been a bunch

of those memes now

lately

I mean it was like GitHub was down

and like you couldn't code yesterday

with any like

GitHub was down as

well

yeah there's a lot of stuff down should i self-host git is this by

the way i'm

hearing from this yeah

oh that would be i mean it's not that bad there's

like lots of solutions for it but

you already self-host git right that's literally what git is

sure but but you know what i mean right you would have to

share it if you wanted to work with anyone

but yeah yeah but you

can share it you just get into remote into your machine and pull from your

your git yeah that

seems safe

you are but you but you're self-hosting by the way but you are

um you are the the server right and and

git habit is

just another server

and you're

just syncing them when you when you pull right let me let me clarify let me clarify

should i self-host a git repository gui and hosting solution for my git repositories

that I

can reach online

from

other...

That sounds like another side project

that'll take you five years.

What?

No, there's like

lots of them,

like

GitLab,

Gidea.

I used to use one.

I used to self-host,

by the way,

on DreamHost when I was on DreamHost

and I had all the Git repositories on DreamHost

and I was using one there

and I had some kind of UI

and I couldn't remember what it was.

It wasn't very good.

It wasn't anything like GitHub.

I was very excited to not self-host,

by the way, onto GitHub.

Yeah.

Which is free, by the way.

GitHub is free.

Until we run into limits.

Well, yes,

but I've

run the whole of Biancon GitHub free

for literally until like three months ago.

I mean, I don't think the limits are that harsh.

Those

free credits, man.

Do you have like popular opinion?

I don't know.

Like, did anyone have access to Copilot

and it just got taken away?

I do have access to Copilot.

I've not checked because Copilot,

I don't find it that useful.

so i i don't either as like an ai tool but for like when you're typing something it would auto

generate it and typically know what i was typing so it was much

faster you

just hit tab i have lost

that now and i feel like my

productivity has gone down like so i've had

free access for

three years or however long

it's why i

don't know like i was a member of an organization

that got free membership.

I would get a monthly email from them saying,

your membership has been renewed.

And then last month, just all of a sudden stripped it away.

I

don't know what organization it was.

It never like had any of that in the email.

Yeah.

So I don't know, like, what do I need to do to get this back?

I don't want

to pay $100 in a month.

You need to pay for it.

I don't know that it's worth that,

but like I missed that easy productivity.

Like I have the, I use the at component syntax in a lot of my components at the bottom that

has like the component documentation.

And

it's basically

marked down in a comment in Svelte.

I need to, I need to, a lot of that would just auto generate for me.

And so like things like that, it's so helpful for.

Yeah.

That's kind of, it's like one of those hidden features that I always forget about.

Like the documenting the

components.

Oh yeah.

It's so clutch for a design system.

I'm going to put it in the show notes.

It works

so well with our IntelliSense.

You hover over the component when you use it

and you can see all the properties that it has.

I have this nice table format I use

and it'll show you basically like a storybook thing

that's not interactive, obviously,

but you can

see everything that

you need.

Yeah, yeah.

It's nice.

Taching feature.

That'll

be my pick.

I'm sure I've picked that before, but I'll put that.

It's worth picking again.

add components, yeah.

So Ilya, do you have an unpopular opinion?

Yeah, I'm not sure if I should say it out loud or not.

So I like types and I like IntelliSense and all this stuff,

but I think I'm with Rich and I'm not sure

if it's most of the core maintainers,

but basically I prefer JSDoc over TypeScript

for many, many, many reasons.

And the main reason is because JSDoc helps me

to keep the code readable.

so I don't need to think about types

when I'm reading the code

I'm reading

the code

when I'm reading the code

and I think that the way how TypeScript works

is not the best solution

for providing types or defining types

so I'm a big JS docs fan

I actually agree with that

but I hate writing JS docs

so I like

writing TypeScript better

see this is why

we are paying for Copilot

because it then infers all the stuff

or rather Cursor

I'm using Chorazone.

Yeah.

All right.

Picks?

Brittany?

I have one of those too.

Brittany?

Brittany can go

first.

Mine's already done.

Yeah.

Add component documentation.

I

don't have to talk about it again.

All right.

Anthony?

Yeah.

So my pick this week,

Shock, Horror, and Surprise is a technical pick.

My pick is Friendly Capture.

And the reason it's Friendly Capture

is because it is GDPR compliant,

European That's

an oxymoron.

Based.

How do

you mean?

A friendly

CAPTCHA is an oxymoron.

Well,

no, it's not. Because one

of the things that I like about it, which I

didn't, it's not why I chose it, but I

like about it, is it's self-solving. So if

you go to the site and you see the CAPTCHA, it will,

whilst you're filling in the signup form,

whatever it will it will solve by doing cryptographic puzzles in your browser and it

will

doesn't that defeat the purpose of the captcha no

because i am not i am not a bot so

if you were a

bot if if you were a bot it would not self-solve right this is this is how it works it's the same

way as recapture will just not appear if uh if it detects you're not high risk this does something

a little bit different maybe you know arguably i don't know how it works exactly but a little bit

in that it will self-solve in your browser

because it will know that you're not a bot

and you're a low risk,

but it's just eradicating any remaining risk, right?

If you were using some automation or something like that,

I imagine it wouldn't solve.

Yeah.

I need something like

that.

I

know I've talked about

the Pokemon stuff before,

but like the recapture is getting insane on Pokemon Center

and you have to like,

the latest one has been like a bird house

and you have to pick which animal will go in this thing in the picture and it's a birdhouse so you

have to pick the birds but you have to do it over like three times

right so you have to pick four

birds

three times in a row to get through this recapture

it's getting crazy like the

you

need a

phd at some point to do captures yeah i've i've just got a new pick

as well so i just because my

wife just messaged me she says that because we're trying to book holiday we're trying to book a

first holiday and like it's literally been like five years since i've been a real holiday um and

she's found one she's the place we want to go they have a they have a plan you can go there like an

all-inclusive thing and it comes with a 24-hour butler

oh yeah that sounds horrible it sounds

amazing that sounds amazing you just call them up and

yeah and they they deliver i think tattinger

champagne to your room every day i mean it just sounds very expensive i

don't remember mcl mc it's

like a cruise line um

mcl they

um we were supposed to go to rome italy and leave off and

do a mediterranean cruise right

when covid hit

and it got canceled but it was gonna have a private

concierge 24

hours in

the room and i'm

like do you do you actually end up using those like whenever

i'm at a hotel

i've never had the opportunity to but i probably would

right i've

caught i've

called room service before

i mean

it's like similar but more personal

i

will definitely

use the bottle of champagne

every day i

mean because because you can say to the butler

we're going to go to the beach take our towels on the beach for us yeah

like so nice

yeah we'll

see you there the amount

of your own towels i want on

my vacation

no but why i don't want to bring

my house

I have children to carry

take my children

take my children to the beach

so your second pick is 24-7

Butler

well my second pick is

this holiday,

it sounds great

where is that?

it's in the Maldives

it's a

little out of

the Maldives

it'd be beautiful, well

when we book it

it will be, it's still not booked it

we're still there's still a millionaring but we book it for hopefully next month

i'm sure that's

very similar to hawaii that's what we're going to january yeah

well hawaii was on our list but we

i mean we love the maldives so i'm excited to go again especially be the first time the kids have

been to like well first time they've been a holiday which is shocking really but um definitely first

time maldives so i think they'll like it we get we're not gonna have a water bill this time because

the thought of having kids in a water villa treading out into the sea is not

amazing so

kids are a little young for that yeah

yeah this all just

sounds very

expensive we're gonna go

kayaking

and snorkeling and all the things exactly a lot of fun paddling

we like to flip it on its head kev we don't go it's very expensive might be fun we say it's fun

but it'll be expensive right because

you know okay

i i think there's no better way to spend

money than on travel i

100 agree yeah

so i i think i would agree as well

the cost worth the cost but

where i disagree is like i would much rather do a long vacation

yeah that

that is cheaper per day than a short one that is more expensive

but it's not that short

still be two weeks but

you could just a good amount of time you could do two months

with two months we can't really take two children away from their school and everything for two

months it's

not really not

practical i

i have fantasized about just letting them do like virtual

school or homeschool or something and just going anywhere i would do a six-month cruise around the

world i would i would just travel

like those uh like those pensioners that are like uh

like the seniors that are like

yeah it would be a bunch of senior citizens and me but yeah

sounds great although yeah i i don't know how you wouldn't get bored on a cruise ship though

like for six months you got yeah you gotta

go yeah i guess i

mean

oh i see so you would take

like one of those travel around the world cruise ships not

yeah where it stops at a bunch of

different okay yeah you

wouldn't go back and forth between miami and the caribbean

no

absolutely not

what so

how long does this stay in a

port for um

like a day or two like a normal

cruise you only stay for like six to ten hours depending on the

port sure that

is very short

but some of these longer cruises like this that we're talking about they do stay overnight sometimes

and

they'll still it's still short you're still spending six months on a ship with with a couple

the days here but

it's like

yes it feels like it's quite weighted it's not enough

to like experience

everything about the place that you're going but you get a little bit of the culture and then you

kind of

figure out where

do i want to go back to and spend more

time that's true it's like a tasting

flight yes

it's a flight of places

jesus

this was one of the longest tangents we've ever been on i

This is definitely a tangent of a tangent.

We can talk about a charity as well.

Welcome

to Svelte Radio with y'all.

Don't worry, we will chat.

We'll chat, Kev, about your long, cheap holidays really soon.

Have I been shamed for

preferring cheaper holidays that

go on for longer?

I have been on many cheap holidays, and I mean cheap holidays as in self-hosting, by the way.

No, self-accommodating in

hostels and

stuff like that,

by the way.

You know why I can go on holiday all the time?

it's because i'm self-hosting by the way self-hosting by the way

exactly not paying for

sell sixteen

thousand dollars a year yeah

or sorry

no no uh did you oh by the way did you uh

uh solve the the google big data big query thing that you were talking about the

other day well

first of all we put a bunch of slots in place we've got a slot reservation so we've reduced our

bills by by controlling them a bit but i'm gonna move the whole lot i think to postgres because i

don't think we use BigQuery for

what it's designed for.

Yeah, so I will do whenever I get 10 seconds.

I will actually go back to working on that,

which is my primary thing right now.

Yeah, no, totally get it.

All right, Ilya, do you have a pick?

How many picks do I get?

You can have as many as you want,

but we will be here for hours if you have 10,000 of them.

I mean, I probably have 10,000.

So for all Svelte developers, all Svelte developers

need to use Svelte Render Scan from Stanislav

and Svelte Inspect Value from iRig.

Those are the

two most

useful libraries that exist in Svelte ecosystem.

Svelte Inspect.

Svelte Inspect Value, right.

This is the one which gives you this pane where you just plug all your state into it

and it renders it reactively.

And if it's an object, it will render the whole object.

You never have to add another code position fix details thing.

Are they packages like you install in your library?

Yeah, right.

The packages, you install them,

and then you just mount them as a regular component.

And Render Scan will render everything.

Basically, if you wrap your whatever section or component

in Svelte Render Scan,

it will render everything that happens within it.

It highlights every re-render.

Oh, nice. Okay.

It's brilliant.

That's brilliant, especially for,

you know, effect and stuff like this.

Yeah.

So those were my, like, two Svelte-related ones.

The third one is Sarah Swaydan, you know, the accessibility expert.

Her course, Practical Accessibility, is on discount right now.

And this only happens, like, once a year or so.

And it's a hefty discount of, like, 30%.

So I think everyone should go and buy it.

Yeah.

That's my three picks for this week.

All right.

Cool, cool.

All right.

so I have two picks

so one is

you can just do things

and I did just that this week

because I've long

been talking about like having a

a SvelteKit adapter

that renders to

a

service worker

and I've always

been intimidated by the adapter

thing like how do you even build

one and stuff but

using the power of LLMs,

I managed to actually create one,

which was very nice.

So I did some experimentation

and I built an adapter

that basically takes your whole SvelteKit application

and puts it inside of this service worker.

So you have server endpoints,

you have form actions,

and all of that good stuff.

Which LLM did you use?

So I used Claude Code.

Yeah, yeah.

I could have guessed.

Yeah, that's my go-to.

It's just very nice.

It is very nice.

So I'm going to pick that.

And then you reminded me of another one of Stanislav's projects.

And it's like a showcase of how to build a progressive web app

that kind of feels native.

And so I met him yesterday here in Stockholm.

We had like a Svelte lunch that we usually do once a month.

They would just go

somewhere and eat

lunch, a bunch of Svelte people.

And so he told us about this thing called Sparkle Post,

which is like a Twitter clone that he made.

And it's just like this very nice experience that you can install on your phone as like a PWA.

And it feels native, which is nice.

Then I have one last pick.

Sorry.

this was my original pick so i have started reading a series called the hierarchy which

is like a fantasy book series by james islington it's very nice uh he just released the second

book in a trilogy and i can really recommend it if you if you're into fantasy so yeah that might

be my next book yeah

it's

right it's very good it's very good so kevin if i understand

you

correctly, that means that you will turn the Svelte

Playground in a progressive web app?

Well, that was kind of one of the reasons why I wanted to try it

or that I picked this thing up again.

Because

I kind

of want to be able to do SvelteKit

like the Playground, right?

The same kind of experience of not having to install web containers

and all this crap that you have to install.

It takes forever and it breaks half the time

depending on the browser that you're using and stuff.

But there's like this one big thing is that you...

It's hard to tie together like updating the files

to actually building it and deploying it

because you have to actually build it to get it to adapt, right?

So you can't do it in like a dev environment, if that makes sense.

or at least I don't know how to uh maybe maybe some magician uh that knows V really really well

can can figure that out I don't know but yeah so those are my three picks a lot of picks this week

right that's crazy all right uh thank you Ilja for joining us uh it was a lot of fun I I had no idea

about all this NGO stuff.

And it's

kind of opened another world

as to how, like there's this other thing

where you produce stuff for NGOs

that's very different to how you would build

for other people.

And of course I

made myself the floating head.

Sorry?

I made

myself the floating head.

Interesting.

I think we're losing it.

My

stomach

is

growling.

Yeah, yeah.

All right.

Thank you, everyone, for listening.

We will talk to you next week.

Bye.

Bye.

Bye.