0:00

Welcome to syntax today. We have the state

0:03

of front end survey results and Scott and

0:05

I are going to go through it. Some

0:07

interesting stuff in here in terms of like

0:09

what frameworks are people using? How are people

0:11

writing CSS? It's a

0:13

really great survey because it kind of gives

0:15

you a glimpse into what

0:18

are people actually using and what are people

0:20

moving towards. And it gives a there's like

0:22

almost 7,000 people have taken or 6,000 people

0:24

have taken this chance

0:28

from 139 countries. I

0:30

think it's a pretty good look at

0:33

where the industry is at right now. So we're going

0:35

to go through the result

0:37

and sort of talk about our

0:40

surprises and what we think about the

0:42

future of web dev. My

0:44

name is Wes with me as always

0:46

Mr. Scott Tolinski. He's doing

0:48

great today. Probably not as great

0:50

as possibly could be. But

0:52

we want to talk to you real

0:55

quick about century. Century is the error

0:57

exception performance insight tool. We've

0:59

heard us talk about them a million times.

1:02

Basically you got a website that's

1:04

buggy, slow, you got issues. Century

1:07

is going to give you insights into why that

1:09

is and help you fix it. Check it out.

1:11

century.io forward slash syntax word.

1:14

One thing I really

1:16

like about this survey is

1:18

that it feels like it's out of our bubble a

1:20

bit. Some of the state

1:23

of you know, reactor, whatever those those

1:25

tend to feel like a

1:27

survey that's representative of my

1:30

bubble in web dev. So

1:32

it's nice to see information from what

1:34

could be considered like a wider breadth

1:36

of the industry. Yeah, I

1:39

would survey is always going to

1:41

be like

1:44

skewed in some direction. But

1:47

for this one, it seems to

1:49

skew way heavy Asia Europe, like

1:52

out of the 6000 people that replied 3000 from Europe 1200

1:54

from Asia, and then only 1000 from North America.

2:00

So I think that is is pretty

2:03

good and like fair point. I never

2:05

heard of the survey before, which

2:07

is good. Maybe maybe it shows a

2:09

little bit of a different look. So

2:12

let's start with the framework aspect of

2:14

this. This is talking about what frameworks

2:16

have you used in the last year.

2:18

And also the results of the survey are kind of

2:21

nice where they're relatively simple. It's just like, did

2:24

you use it and like it? Did you use it

2:26

and dislike it or the opposite? And

2:29

the most used and liked framework

2:31

is react coming in at 70

2:33

percent and then 15 percent use

2:37

and dislike, which you can kind

2:39

of expect that. I think that's not very that

2:41

surprising. But Svelte is Svelte's got

2:43

what? Twenty five percent used and like out

2:45

of all respondents. That seems really high in

2:47

view. Forty four point eight percent use and

2:50

like, which is great. I like

2:52

this interface where it shows the like

2:54

the darker outline around because you can see

2:56

that like very clearly at a glance that

2:59

Svelte has the highest percentage of want to

3:01

learn in the future, which is great. I'm

3:03

going to point that out, which is what

3:05

do people want to learn? Like,

3:07

what are people using now? Reacts felt

3:09

in view. But what do people want

3:12

to learn? HTMX pretty high as well.

3:14

Svelte extremely. I think what is felt the highest

3:16

one in want to learn? It is. It is.

3:18

It is. And the second. It's

3:21

great. Yeah. Oh, yeah. We use

3:23

it. You know, actually, you know what? One

3:25

thing I, you know, turn into a little

3:27

Svelte love sesh here. But the one thing

3:29

I really enjoy most recently about the transition

3:32

to Svelte five was that the migration assistant

3:35

worked on the first try and

3:37

only gave me like five

3:39

files to touch after the fact, which

3:42

is pretty substantial given the changes. Yeah,

3:45

it worked really super well. It was

3:47

maybe the smoothest migration I've done on

3:50

anything. And let's face it,

3:52

I've done a handful of migrations. What's

3:55

also interesting here is that HTMX is number

3:57

two for want to learn, but

4:00

also some of the highest for

4:02

not interested and no opinion and

4:05

only seven percent for used in

4:07

and liked so htmx gets like

4:09

the hype award right it has

4:11

yeah because it's the second most

4:13

wanting to learn thing but in

4:15

reality but then again it's

4:17

new so seven percent is is probably a big

4:19

jump up from where was in the past it

4:22

is it is wild to see that

4:25

something that most people weren't aware of like

4:27

two or three years ago is is now

4:29

this high meanwhile something

4:31

like Phoenix which is fairly mature even

4:34

like alpine JS is only like seven and

4:36

1.8 percent so yeah

4:39

I'm going to remember one yeah I

4:41

think like linkedin is is one of

4:43

the biggest users of ember I wonder

4:46

who else is any big out other

4:48

sites on ember yeah I

4:50

did you ever learn ember do you ever pick it up slightly

4:53

I touched upon it it was kind of

4:55

like during the like angular you

4:57

know we're all kind of coming off of backbone

5:00

looking at angular and then ember was was one

5:02

of the good big come ups from it we

5:04

had dinner with that someone worked at linkedin at

5:06

the after the meetup and he

5:08

was saying embers great but it's a little

5:10

bit different and like

5:12

there's not a ton of documentation

5:15

on like bugs and stuff outside of

5:17

luckily inside of linkedin they can they

5:19

can reference all those people because they

5:21

use it quite a bit internally man

5:24

I did that my very first one of

5:26

the first tutorial courses I ever bought was

5:28

from I don't know if you remember peep

5:30

code yes yeah yeah

5:33

I bought a I bought a

5:35

a course from peep code on ember and I

5:37

took it on the airplane to somewhere

5:40

I forget where maybe even did to

5:42

yeah forget I took it on the airplane

5:44

somewhere I did the whole tutorial course and

5:46

like left being like all right you could

5:48

do ember yes I could see how this

5:50

works yeah that's I learned Express on

5:53

from peep code as well he had a really cool

5:55

tutorial about like making pies and most recent

5:57

pies and in web sockets and

6:00

I remember watching it on an iPad at the

6:02

gym. There's this is like one, you know, like

6:05

the, like those core moments in your like web

6:07

development history. You just remember where you were when

6:09

you learned something and

6:11

I was one of them. That's a weird, weird

6:13

remember that I have. Yeah. Let's

6:16

talk about rendering frameworks, um, which is

6:18

like, basically you have like reactants felt

6:20

in view, but like what frameworks do

6:22

you use on top of it? Most

6:25

used next. And then basically

6:27

half of that. So next is

6:29

52% half of that is nuxed

6:31

and Astro coming in and

6:34

then Svelte kid at 17% remix only 12%, which I'm

6:36

always surprised at the, the

6:40

numbers behind remix. You know, I always think

6:42

that they would be much larger, but just

6:45

goes to show that like when something big

6:47

catches on, it's quite

6:49

slow to move on to something else. Cause

6:52

that's an Astro. I mean Astro kind

6:55

of also showed up. It was

6:57

one of those ones that you felt like,

6:59

how are they going to differentiate themselves at

7:02

first being a static site platform, then just

7:04

like the, the whole thing being you can

7:06

use any front end framework to me, that

7:08

wasn't super compelling until, uh, it's really evolved

7:10

into like, it just, uh, maybe one of

7:12

the smoothest, uh, full stack

7:14

frameworks to work in. And again,

7:17

Svelte kid has the highest want to learn

7:19

in the future, which is pretty great for

7:21

that team. I think it's a good product.

7:23

Yeah. I don't know too much there.

7:26

Gatsby with the lowest. That's also not shocking

7:28

given that we're gas, these trajectory has been,

7:30

I'm surprised it's even at 9.4% of

7:33

people wanting to learn it. But Hey, let's

7:36

take a look at, at state management, because

7:38

outside of the view and Svelte world

7:40

state management picture is a little bit more

7:42

complex as we've, we've talked about a lot

7:44

of times on this show. And

7:47

so a lot of these are react specific

7:49

options except for, uh, Penia,

7:51

I thought it was Pina, but at Pina,

7:54

uh, I think we had the same conversation

7:56

last time. Either way, the highest most used

7:58

in like is. it the

8:00

react context API straight to the dome.

8:02

That's the highest using like I

8:06

get it. Yeah, it's simple. But I still would

8:08

want maybe it's a little something more more, you

8:10

know, it just goes to show that like,

8:13

Come on, react, give us give something

8:15

better, you know, the context API. Yeah,

8:18

it's okay. Set state, put it in

8:20

context API. But there's all these like

8:22

gotchas and downsides to it. But obviously,

8:24

that's what what people are actually using

8:27

to do. And then Zoo stand Redux,

8:30

Redux toolkit coming in at

8:33

a fairly close tied second.

8:35

And then mobx

8:37

only at 8% Joe

8:39

tie at 11%. So Zoo

8:41

stand quite a bit more popular in

8:43

Joe tie. Yeah,

8:46

that's actually not that Zoo stand came first.

8:48

And it was kind of one of the

8:50

first alternate ones. I am surprised that mob

8:52

x is so low given at one point

8:54

felt like it was between

8:56

Redux and mobx as being the two

8:58

major choices in the space and redux,

9:01

obviously, the most popular one. But you

9:03

think after being around for so long,

9:05

and having an approach that's more similar

9:07

to like, what we're

9:09

doing in solid and you that

9:11

mobx would, yeah, be a little

9:13

bit higher. Can I quiz you on

9:16

the next one? Don't look at the other libraries

9:18

or have you? Okay, so I'm gonna quiz you

9:20

on the next one, which is what

9:22

other libraries have you used in the

9:24

last year. So I'm gonna I'm gonna

9:26

rattle off the six libraries to have

9:28

here. And you have to guess, which

9:31

is the most used

9:33

and liked and

9:36

the most not interested.

9:38

Okay, the what

9:40

is the most used and

9:42

like library between low dash

9:44

jQuery, emmer, RxJS, Ramda and

9:46

underscore. Jeez, jeez,

9:49

that's a quite quite a

9:52

bunch of things. So the most used

9:54

and liked out of all of those. Yeah,

9:57

gosh. Oh,

9:59

man, That's a tough because. Or

10:02

maybe we'll ask this, what's the most

10:04

used? Not

10:07

liked or disliked, but just used entirely.

10:10

Yeah, jQuery. That's not actually.

10:12

So jQuery is a close

10:16

second, but low dash coming in at 70%

10:18

of used. Okay,

10:23

I can see that kind of. Maybe we're finally

10:25

starting to see the like, the

10:28

slow on use of jQuery. I know

10:30

that like it's been gone for what,

10:32

like 10 years or so, but people

10:35

still use it. People have existing applications

10:37

and that's a frustrating

10:39

thing sometimes about the bubble is people

10:41

don't realize that like you pick tech

10:44

and you work with that tech for,

10:46

for many, many years, especially when it's

10:48

such a low level choice, like writing

10:50

all of your JavaScript in jQuery

10:53

or picking a library like react.

10:55

Yeah, I'm going to say for the

10:57

other one, it's probably for the least

10:59

used or whatever. It was like least

11:02

the most not interested,

11:05

most not interested. I'm going to say emmer

11:07

only because it feels too niche. Yeah, I

11:09

think emmer has the highest,

11:11

no opinion because I don't think anyone

11:13

even knows what emmer is. Okay. Then

11:16

what about since, since low dash was the highest,

11:18

then what's the, the underscore

11:20

we'll say underscore. Yep. It's an

11:23

underscore. Yeah. So 30% of

11:25

people are not interested in learning underscore. Which

11:28

makes sense. I mean, very close under

11:30

that as jQuery though. I don't

11:32

even know what I'm using. I would use low dash for

11:34

today anyway, so that I couldn't just hit myself.

11:37

Yeah. I have like the opinion

11:39

of low dash is not bad. You

11:41

just reach for the methods when you need it.

11:44

And I almost always,

11:47

when I run into somebody using

11:49

low dash, I think I could,

11:51

I could write that

11:53

in some of the newer methods or

11:55

whatever. And there are like some

11:57

weird use cases where you're like, you have to

11:59

like pick. and whatever, but even that there's

12:02

all these new JavaScript methods

12:04

that allow you to intersect

12:06

arrays, um, finds, find which

12:08

items overlap each other, find

12:10

which items are not

12:12

included in either. Like there's very

12:15

few use cases unless you're

12:17

doing some pretty heavy data stuff

12:20

where, where I think you even need it. And

12:22

that's not to say it's bad. Lodash

12:24

is fast and it's, it's well tested and all of

12:26

that stuff, which is great. But when

12:28

I see it used, I often think, huh, I

12:30

don't think you need it to pull that in. That's

12:33

how I feel too. And I do reach for

12:35

just, uh, people who, who we've talked about just

12:37

on this show. Um, we'll link it

12:39

up in the show notes, but just as a good alternative, cause

12:41

you can just grab one or two, um, methods

12:43

or anything that you need straight out of just

12:45

instead of bringing in a big

12:47

old platform or something. Okay. Well,

12:50

that's not true because Lodash for the

12:52

longest time you've been able to load

12:54

just the method you've needed from. Oh

12:56

yeah. Yes. Totally.

12:58

Yes. But I, I almost

13:01

do as well as the one time I do reach for

13:03

it is debouncing and throttling

13:06

where I'll, I'll install just

13:08

debounce or just throttle. And

13:10

because like, I hate writing that myself.

13:13

It's it's too complex. Yeah.

13:15

And I don't honestly don't know what the sizes are.

13:17

I think what attracted me to just is the fact

13:19

that the sizes of all of these things are copy

13:21

pastable. Yeah. Yeah. They're,

13:23

they're published by their own

13:25

self. They don't have to, it's

13:29

not like part of a library where it

13:31

reaches in for other utilities. So it's very

13:33

copy pastable. Let's talk about data.

13:35

I think data is an important one, how we're

13:37

bringing in data to our applications. This

13:40

one's shocking to me, 73% the highest

13:42

used in like there's Axios. I gotta say

13:44

why, what are we doing here? That

13:47

to me is kind of wild. I just, I

13:50

just never have needed it. I don't get

13:52

it. I know you used it and liked it West

13:54

in the past, but I

13:57

say we're God, we did an entire

13:59

show on. Why do people still use

14:01

Axios, which is episode number four 53

14:03

go to syntax out of

14:05

femme four slash four 53. Why do

14:07

people still use Axios over fetch? And

14:10

we detail why people

14:12

still do it. There are some use cases,

14:14

which is like interceptors and

14:16

expired JWT tokens and validation status

14:19

and all that.

14:21

But again, I, I

14:24

don't find myself reaching for Axios all that

14:26

often. Um, but

14:29

it's higher than native fetch.

14:31

I know what's up with

14:33

that. Yeah. Yeah. Then native

14:35

fetch isn't hard. I get it that there's

14:37

some niceties from Axios, but like, is it

14:39

that hard to implement those that you got

14:42

to bring in a library for it? Yeah.

14:44

Yeah. I think there even is some, there's

14:47

like an Axios like library that gives

14:49

you the pieces that people use. Like

14:51

people want to set their off headers

14:53

once. Yes. And or people don't want

14:55

to have to double Jason, you know?

14:57

Yeah. Especially with TypeScript and like the

14:59

answer to that as well, just write

15:01

a wrapper function, but Axios kind of

15:04

is that wrapper function. So I think

15:06

there are alternative libraries that sort of

15:08

give you that Axios wrapper

15:10

without the baggage of

15:12

the whole library. Yeah. Word. Isn't

15:15

that actually, no, that's, is that K Y is that what

15:17

that library is? K Y. Yeah.

15:19

I'm not, as you're describing this,

15:22

I was looking at it and I was thinking, I'm

15:24

pretty sure that's this K Y del it's from a

15:26

prolific package. So

15:29

how's Cindy Sora delightful HTTP

15:31

requests. It is a simple

15:35

HTTP client based on fetch specifically. So

15:39

I do think that's what you want.

15:41

Yeah. Yeah. And but nobody knows about

15:44

it. 66% of people, no opinion on this

15:46

thing. So give K Y a check. I, I

15:48

don't feel like I need it or want it.

15:50

But if you're reaching for Axios,

15:52

K Y might be a good option. As

15:54

far as other ones we tend

15:56

to encounter here, I'm pretty stoked on the fact

15:58

that Tanstack queries. so high at 43% if

16:01

used and liked. Uh, when we talked to

16:03

Tanner, he certainly made me like a

16:06

believer in all the stuff he was doing. So yeah,

16:08

it's, it's nice to see that doing, doing

16:11

well, Apollo client kind of holding on there

16:13

a little bit. 25%

16:15

is not nothing, but yeah. Um,

16:18

I haven't heard about Apollo in a little while.

16:20

I don't know what their status is to be

16:22

honest. Yeah. I, that's a great question. I don't

16:25

know as well. It's, it's been a while since

16:28

I've, I've used Apollo sort of. I think

16:30

like all things, GraphQL went a little bit

16:32

more enterprisey, uh, cause that's

16:34

sort of where the use case for a lot

16:36

of GraphQL stuff is. Hosting.

16:39

Can you, can you guess the

16:41

top hosting providers? Oh,

16:44

top hosting providers. Keep

16:46

in mind, like this is like JavaScript primarily

16:49

front end people. Okay.

16:52

Well, one, I think Netlify is going to be

16:54

on there. I think for cells going to be

16:56

on there. Yeah. Digital ocean's going

16:58

to be on there. Cloudflare and AWS,

17:02

I guess. Yep.

17:05

You can digital ocean, not on there. Um, surprisingly.

17:07

So number one, for cell 36.2% and then AWS

17:09

32% under that. So

17:14

more people using for cell over AWS,

17:16

which makes sense. I'm just surprised.

17:18

I keep, I know I keep saying this, but like,

17:21

I'm so surprised Amazon has not

17:24

made a Vercel like experience.

17:26

And every time we say this, people come out and

17:28

say, yeah, but there's AWS.

17:31

What's the, their equivalent. And

17:35

why you say this was it's so funny. I went to

17:37

like login to pay my bill on AWS the other day

17:39

and I got hit with this login. Um,

17:41

here's the new sign in UI and somehow

17:44

the new sign in UI not only looks

17:46

like the same as the last one, but

17:48

it's somehow just like slightly worse. I don't

17:50

even, I like, don't know what's going on.

17:52

And then I went to like, go through

17:54

the billing experience. And I was just like,

17:56

how, how have they managed

17:58

to make this? Such a maze

18:01

how have they managed to make just finding

18:03

my where my credit card information is like

18:05

so many clicks and Such

18:07

an amazed to get to I really truly

18:09

don't understand it is shocking Even

18:12

Google though Google's a developer platform site and

18:14

the Azure's developer platform. Yeah, they're all just

18:16

I've already read We're able to get to

18:18

the cup a couple of times. But man,

18:20

they're just they're rats nests of

18:22

yeah, I'm just stuff Yeah, the

18:25

hilariousness that I'm a forgetting

18:28

What AWS is like

18:31

continuous integration thing where you like push a git

18:33

commit and it rebuilds it and deploys it for

18:35

you The fact that I'm

18:37

forgetting what that call is called, which I'm sure you're

18:39

screaming at it right now I know by the way,

18:41

I can what is it? Amplify

18:44

the amplify. That's it. Okay,

18:46

and The fact

18:48

that I was searching is you can

18:50

more AWS Deployment

18:52

thing or AWS vault all

18:54

this versatile alternative and that did not come

18:56

up once. Yeah Yeah,

18:59

they just have so many products and

19:01

like come on you gotta think They

19:04

could put a couple Hundred

19:06

people on this and build it. It

19:08

is kind of wild that DigitalOcean didn't show up I feel

19:10

like at some point they would be very high on this.

19:13

Yeah just given

19:16

Yeah, given where they've

19:18

been in the past. Yeah. Yeah DigitalOcean's

19:20

app platform is is specifically

19:24

like what we're talking about which is like a Platform

19:27

as a service, right? It's not you don't get

19:29

a Linux box and login and install node you

19:31

just like set up some config file and you

19:34

give it your git repo and it

19:36

will build and That I've

19:38

used that a couple times, but it never really

19:40

took off As

19:42

much as I thought it would Cloudflare pages though is

19:46

Again, that's cloud flares equivalent to that. It's

19:48

higher than I thought way. It would be as well 10% Yeah,

19:51

70 80% of people using

19:53

continuous integration right now I don't

19:55

know what the other 20% of the

19:57

people are doing like are you FTP and dragging

20:00

and dropping, what's the alternative to continuous

20:03

integration in these days? FDP-ing, RSync,

20:06

I don't know. Logging in and Git pulling and

20:08

then running. Drag, well

20:11

like, you laugh, but

20:13

if you sign up for ADDS

20:15

serverless functions, like their

20:17

initial interface for deploying a serverless function

20:19

is upload a zip file of

20:22

your serverless function with your node modules in

20:24

it. Like that's their interface for

20:26

that. And obviously there's way better ways to do it,

20:29

like that's your getting started. But

20:31

the CI, GitHub actions, 70%. Like

20:35

remember back in the day, everyone

20:38

used CircleCI or Travis? CircleCI, Semaphore,

20:40

Travis, there's a whole- 7% Travis,

20:42

pretty low. I used

20:44

a bunch of these and I never liked any

20:46

of them. Would you wait? Why it makes sense

20:48

that- They're all painful. That Google or GitHub actions

20:51

were able to just do it. It

20:55

is one of those things that we talked about,

20:57

especially in the co-pilot episode that we just recorded,

20:59

where if you can

21:01

put things closer to Git and

21:04

GitHub, where people are already

21:06

using GitHub, then GitHub

21:08

will take over. And I think this is a prime

21:10

example of it. It's not like GitHub actions are the

21:12

most friendly, fun thing to work on in the world,

21:15

but they certainly do the job. It just goes

21:17

to show you just make it free, make it

21:19

available, people are gonna use it and just take

21:21

over that platform entirely. I haven't

21:23

reached for anything else. I do a

21:25

ton of GitHub actions on most projects.

21:27

So definitely myself included into

21:29

that. Have you

21:31

used Microfrenance? This is kind of

21:33

an interesting one because the whole

21:35

Microfrenance thing is I keep hearing

21:38

about it. I keep talking to

21:40

people about it at conferences that

21:42

are from larger companies, but

21:44

it's not something I've dipped into. And I think

21:46

that this is a show that we need to

21:48

have a guest on for because

21:52

the whole idea of microservices

21:54

was very popular. And

21:56

now we have this idea of Microfrenance where you have

21:59

different teams. building different components

22:01

that then come together. And

22:04

then there's this whole, this whole Webpack

22:06

module federation thing where you

22:08

can share code between them. I think that

22:10

we need to, we'll ask you this, dear

22:12

audiences, who do we talk

22:14

to about Microfrontas? I've had a lot of

22:16

people reach out about the topic, but it's

22:19

almost always that they're from a company trying

22:21

to hawk their thing. And

22:24

that's sometimes fine, but I'm always

22:27

hesitant to do that. I'm

22:29

always hesitant to do that as well. Yeah, the Microfrontas

22:31

thing seems to me a problem or

22:34

a solution to solve a problem that we,

22:36

you and I just will never face because

22:38

we don't have many different

22:40

large code bases. 3,000 JavaScript

22:42

developers on staff, yeah. Like you see

22:45

tons of components, different skill levels, different

22:47

departments. I mean, that's just not a

22:49

problem we're going to hit. So it

22:52

does seem like kind of a hairy solution to

22:54

me. I feel like it's a better solution to

22:57

just get everybody on page ring the same thing,

23:00

but man, that's probably not practical. It

23:02

probably is, but you're not ever going

23:04

to do that in fast

23:06

enough where people can get their work

23:08

done. I know, it's such a

23:11

dumb guy view of it. Or why are we just all used

23:13

to the same thing? Yeah. All 5,000 of us. Oh,

23:16

great idea, yeah. Let's just like

23:18

scrap every piece of tech that we've built

23:20

our entire business on for

23:22

the last 10 years. I could do

23:24

that. Next one, package managers.

23:27

NPM, no doubt, 56%. Easy

23:30

peasy. I'm surprised that yarn is still at 21%.

23:33

I felt like yarn had such a big fall

23:35

from grace when they had their whole like V2

23:37

thing came out. I felt like that was going

23:39

to be like an Angular 2 moment for them,

23:41

but it seems like they're holding on. PNPM, that's

23:43

what I use. It's at 19.9%. I

23:46

like PNPM. There's occasionally some like little weird

23:49

things with it, but for the most part,

23:51

PNPM handles a lot of stuff really well. And I don't know,

23:54

I'm going to, I don't want

23:56

to speak out of ignorance here. But

23:59

in NPM, workspaces, do

24:02

you have to define all

24:04

of your packages in

24:07

their version as workspace, you

24:10

know, version? Because

24:12

one feature that I've been really liking

24:14

in P and PM workspaces is

24:17

to treat all

24:20

packages, like check for the packages

24:22

first in the workspace without having

24:24

to put workspace in the version.

24:27

You just put the version like

24:29

normal in your package.json and it

24:31

will check the workspace first, even

24:33

without the word workspace in the

24:35

version. Meaning that like I can

24:38

use the same dependency, like line

24:41

of code to pull from both

24:43

NPM when it, when it's somewhere

24:45

else, but also from my workspace.

24:48

If it's within my workspace itself

24:50

in my mono repo. Yeah. I

24:54

don't know the exact answer to that, but

24:56

I do know that the P and

24:58

PM one is a little bit simpler.

25:00

I recently went and tried

25:03

to deploy my entire course platform

25:05

to Dino deploy just to

25:07

see, can you act like, so

25:09

they're like, we have full node support now, blah, blah, blah,

25:11

you know? And I was like, all right, bet

25:13

let's try like this huge node app that

25:16

I've been working on for 10 years. That's

25:18

a massive mono repo and there's

25:20

multiple workspaces or there's, there's yeah,

25:23

multiple workspaces in it. So

25:25

I use PMPM to manage my workspaces

25:28

and obviously deploy it to Dino play. They're going

25:30

to use NPM or they're going to use their

25:32

NPM equivalent. And the thing

25:35

I had to change was that I use

25:37

like workspace colon star in my package, Jason,

25:39

which is really nice because it just finds

25:41

the folder that all top level folders you

25:44

want to add a new package, you just

25:46

add another one. I had to go change

25:48

it to explicitly stating all

25:50

of the workspaces in there. Oh,

25:53

you did. Yeah. Yeah. But aside

25:56

from that, I don't think I try

25:58

not to do anything. that

26:00

is specific to your package manager because I

26:02

want to be able to run it with

26:04

NPM or with PNPM. That's part of the

26:07

reason why I'm get whenever a

26:09

project uses yarn. Like I was trying Redwood

26:11

JS out the other day and it uses

26:13

yarn and like it wasn't installing

26:15

properly and I was like, oh, like I have

26:17

the wrong version of yarn installed. And then, but

26:20

yarn also now comes with, um, what's

26:24

the thing? It comes with node. What's that called

26:28

in, in node where node now

26:30

ships package managers as well.

26:32

I remember that that was a whole

26:34

big controversy when they added that to,

26:37

to node. It was like the yarn

26:39

team specifically was really pushing for it,

26:42

which I can understand

26:44

why it's their, their product. They wanted

26:46

alternatives. And by the way, the, the

26:48

PNPM package setting for people

26:51

who are wondering, because I mentioned this

26:53

is a setting called link workspace packages.

26:56

If enabled locally available packages are

26:58

linked into node modules instead of

27:00

being downloaded from the registry. Oh

27:03

yeah. Yeah. And so it's basically

27:06

like, again, you just use the versions

27:08

and it automatically does linking instead of

27:10

needing to put workspace in the version,

27:13

which is nice. All right. It's called

27:15

core pack. So cord pack is an

27:18

experimental tool that helps with managing versions

27:20

of your package manager. So this is

27:22

interesting because your

27:25

package manager can specify which version

27:27

of your package manager to use.

27:29

So if you have one node

27:33

installation or you have one project that needs

27:36

like a specific version of yarn to run

27:38

and another one that needs a different version of

27:40

yarn to run, you can use

27:42

different versions of yarn and PNPM on

27:45

a different one without having to do

27:47

the whole switcharoo every single time. And

27:49

I know everyone's saying just use Docker.

27:51

No, I will not use Docker for

27:53

that type of thing. I think I

27:56

get a little bit closer to using Docker

27:58

every day. West. I think I think that

28:00

is, that's my future here. I'm gonna be

28:02

one of those guys like a year from now, it's like,

28:04

I love Docker. That's so funny, because

28:07

I have like nine Docker images running

28:09

on like servers. I have a local

28:12

running, but like on my own computer,

28:14

I just don't feel like

28:17

doing that. I'd rather go to several places

28:19

to start my database itself, to boot

28:21

up my server and then boot up my UI. Rather

28:24

just run off those processes individually.

28:27

I want to download the like 47 gigabyte Docker file every

28:31

time I want to run something. Docker's

28:33

crazy. I got it. Let's look at run

28:35

times. We talk a lot about run times,

28:38

but 96% of people are using Node. Oh,

28:42

because this is a multiple choice. I'm wondering like how

28:44

people are using Bun and Node. Okay, so Node

28:47

at 96%, Bun at 10%, which

28:50

is the second highest. Dino, it kind of a shocking

28:52

2.6. Dino just works y'all.

28:54

It is really nice and smooth.

28:56

I think Bun is just like a

28:58

little extra sprinkles on top, but Dino

29:00

just works to me, you know? Yeah,

29:03

I think the Bun is a bit

29:05

more, it's kind of like the HT Max answer,

29:07

is it's a bit hyped. I also, I think

29:09

a lot of people are using Bun for their

29:12

tooling and their build as well, just

29:15

like as an alternative to PNPM or

29:17

Yarn or something like that, right? So

29:21

you're seeing quite a bit of that as

29:23

well. I'm surprised to see that so much

29:25

higher than Dino though. Which

29:27

is just from, like

29:30

I haven't heard of anybody actually running Bun

29:32

in production just yet. I've

29:35

seen people use it, like even on Cloudflare

29:37

pages, you can use Bun to

29:39

build your website and then

29:41

deploy it to Cloudflare pages. So you

29:43

don't run your website in Bun, but

29:45

you can use it as the build

29:47

tool. Yeah, let's talk about build tools,

29:49

TypeScript, 96% or 90% of people have used

29:52

TypeScript. Not

29:54

surprising given, you know, I guess 90%

29:56

is a lot, 90% is a lot. Everybody.

30:00

Yeah. Yeah. So if you're

30:02

not on TypeScript and you're one of those people

30:04

that's holding out, you are part of the nine

30:06

point four percent. It is funny. Anytime we like

30:08

post things on YouTube, we occasionally get comments from

30:11

people that are like TypeScript is a waste of

30:13

time. And it's like, OK, buddy, it's that's you.

30:15

You're one of the people here who are

30:18

in that nine percent. Yeah. 94

30:20

percent of people used

30:22

and liked TypeScript used and

30:24

liked 94 percent. Used

30:27

and liked. But like the

30:29

amount of people that are still using jQuery

30:32

and everybody is 90 percent on

30:34

on TypeScript is is amazing.

30:36

Right. So there's. Does

30:39

that mean like I would have thought

30:41

it would the nose would be a bit higher because

30:43

of legacy stacks, you know, just just

30:45

simply can't move it over to TypeScript.

30:47

But maybe people are just

30:49

starting to opt into it's a

30:51

writing the new pieces in TypeScript and and

30:54

then consuming the old JavaScript versions as

30:56

they need it. Yeah. 2.2 percent

30:58

of people used and liked flow.

31:01

So if you're one of the people who held

31:03

on to flow, then

31:06

yeah. Good luck there. I

31:09

do hear occasionally there's some nicer things about

31:11

flow, but yeah, not nice enough for me

31:13

to to jump on that bandwagon here. That's

31:16

a little bit too small. Fifty three

31:18

percent of people said that TypeScript overtook

31:20

JavaScript and became the new front end

31:23

standard. I believe that JavaScript will turn

31:25

into something like TypeScript. Sixteen percent of

31:27

people agreed with that statement. What do

31:29

you think about that? I

31:32

whether or not it will turn

31:35

into something like TypeScript. I do

31:37

believe that we'll

31:39

get some sort of types as

31:41

comments as that that proposal was

31:44

where we can write types and it

31:46

will run as JavaScript. And

31:48

we did get TypeScript running support within

31:50

Node. It's in all of the other

31:53

JavaScript run times. It kind

31:55

of feels like JavaScript is going

31:57

to turn into something like TypeScript at some point.

32:00

Yeah, yeah, I don't know that we'll

32:02

ever get like actual types as part

32:04

of the engine. Actual. But

32:06

from an authoring perspective, yeah, I

32:08

think we're we will get

32:10

that where we can we

32:12

can write types in JavaScript and then you

32:14

can simply consume that via anything.

32:16

Like you should be able to run a.

32:21

Typed JavaScript file in the browser without

32:24

having a syntax error, you know, meaning

32:26

like the engine should be able to

32:28

strip out those types of for it

32:30

runs it. And then I

32:33

think that we'll see once we get that

32:35

we'll start seeing some tools

32:37

that are like TypeScript

32:41

ish, you know, like a new typed

32:43

version of we'll see tools built

32:45

on top of that to sort of formalize what

32:47

that looks like. And maybe we'll all standardize

32:50

on on one way going forward.

32:53

Yeah, I think TypeScript might be something else. We'll

32:56

see. We'll see. There's

32:58

also one point nine percent of people

33:01

who are still holding on to everyone

33:03

will soon forget about TypeScript. Yeah, that's

33:05

that's going to happen. That's about as

33:07

likely as I just going away next

33:10

week. Yeah, you know, that's that's not

33:12

going to happen, folks. Browser technologies. Eighty

33:14

two percent people using the fetch API,

33:16

which I guess other people are still

33:18

using Axios. Hey, before

33:22

before you even go on from that, I don't know if you

33:24

noticed, Wes, but in

33:26

the front of each one of these section,

33:29

there's a little there's an expert who gives

33:31

a little an expert. Yes. Yeah. An expert.

33:33

And actually, one the one before this was

33:35

Daniel Roch. Shout out to Daniel. He was

33:38

at our syntax meetup in SF. So it

33:40

was nice to meet him. But guess who

33:42

wrote something about fetch? It was me. I

33:44

wrote something here. Yes, fetch. It's both surprising

33:47

and unsurprising that fetch has eighty two percent

33:49

usage rate. On one hand, it's amazing how

33:51

quickly we've standardized on fetch,

33:53

which is simple enough that most users

33:55

don't need external dependencies. That's me putting

33:57

in my my axi stick right there.

34:00

On the other hand, we don't usually move that fast. Uh,

34:03

that was my, my little, uh, well, I had some paragraphs

34:05

actually. They tightened it up for me. I'm not a gifted

34:07

writer. So their staff

34:09

looked at it and gave it a little editorial,

34:11

but this is my words for the most part,

34:13

so read my little sentence here and sing if

34:15

you want. Yeah. Other

34:17

things noted a lot of these APIs that

34:20

got usage, even though they're, they're still under

34:22

50% or a lot of the things that,

34:24

well, I think we're going to see more of

34:26

because they're born native app like features, whether that

34:29

is like the storage API,

34:31

even just using service workers

34:34

or caching APIs, web workers,

34:36

index DB file system API full

34:38

screen full screen. Yeah. These things

34:41

are going to get more ubiquitous. They're going

34:43

to get more available. The one that is

34:45

kind of crazy is that 3.5% are using Houdini, which.

34:49

What part of Houdini are they using? I don't

34:51

know. No one's using Houdini like maybe with it

34:54

making a demo, but I disagree. I

34:56

think that people are using at property occasionally

34:58

now and at property is like probably one

35:00

of the, is that part of Houdini as

35:02

part of Houdini? Yeah. Okay. So maybe we

35:04

can explain what app property is. Do you

35:06

want to give a stab or do you

35:08

want me to? Yeah, no,

35:10

I'll give a stab. It's basically, you

35:12

can think of it as like types

35:14

for CSS variables. So typically when you

35:16

define a CSS variable, you're just saying,

35:18

here's the variable, here's the value. And

35:20

that doesn't tell CSS or the browser

35:22

a whole lot about what that value

35:24

is. And with this app

35:27

property rule, what you're essentially doing is

35:29

you're taking like four lines of code

35:31

to define a single

35:34

variable as in here's what the variable

35:36

is, here's the type of the variable

35:38

is a number or what, and

35:41

then here's a default value for it.

35:43

Now the cool thing about that is

35:45

because you know what type it is,

35:47

it makes interpolating those values possible by

35:50

the browser, it's the thing

35:52

that makes it. So the

35:54

classic example is I can, I can

35:56

with app property, I can animate a

35:58

gradient. That's. That's what's cool

36:00

about it. But yeah, I think for the most

36:02

part, it just types for your variables. The

36:05

other good use case or that I've used app

36:07

property for is in a

36:10

keyframe, if you want

36:12

to, instead of changing a property

36:14

in a keyframe, like for example,

36:16

a rotate value, if

36:18

you want to update a rotate

36:21

value in a keyframe, you have

36:23

to say transform, rotate, whatever. The

36:25

downside to that is if you

36:27

have other transforms, then you

36:29

you're overriding those by by just

36:32

simply putting transform, rotate, whatever. And

36:35

what you can do with app property

36:37

is you can just update the variable

36:39

in the keyframe. So you could say

36:41

dash dash, rotate value 20 degrees. And

36:44

then wherever that variable is referenced by default,

36:46

it would just snap because it doesn't know

36:49

what the type of the CSS variable

36:52

is. But with that property, you can

36:55

say this is a a length or

36:57

a percentage or a degree or something

36:59

like that. And then the browser will

37:01

know, oh, this keyframe updated a variable,

37:03

which is a degree. I can animate

37:06

degrees. So then it will actually

37:08

give yourself the the smooth animation or transition

37:10

that you're looking for. Yeah. Counterpoint,

37:13

I just use the transform properties like rotate

37:15

as a property straight. Is that

37:18

is that in all the browsers now? Oh, yeah. Just

37:20

for a while, I remember talking about it on

37:23

this podcast like like five years ago. Oh,

37:25

it's big and it has been. But

37:28

like that's that's great if you are straight

37:30

up doing it. But if you are using

37:32

like that variable in a calc or something

37:34

like that, you know, I want to I

37:37

want to turn this to 12 degrees, but

37:39

I need to calculate something else based on

37:41

that value, then it gets a little tricky

37:43

because you simply just want to update a

37:45

variable and anything that is then reactive to

37:47

that variable does it. So let's look at

37:49

the update. Rotate

37:52

it for it came into five, believe

37:54

it or not, it came into Firefox in 2020. It

37:58

came into Safari in 2021. and

38:00

it came into Chrome in 2022. It has

38:02

94% global usage.

38:05

Yeah, you can use that now. Yeah,

38:07

you can use it. You can just

38:09

straight up use the rotate value, which

38:11

is good because it's annoying to even

38:13

if you're not using a keyframe, it's

38:15

annoying to have to write transform rotate,

38:17

rather than just rotate to do that,

38:20

folks. Progressive web apps that

38:22

most people think the popularity will slowly

38:24

increase. I agree with that. The APIs

38:26

are coming in slowly to

38:29

all browsers and everything like that. You

38:31

can now install apps on your home

38:33

screen and the correct ways on many

38:36

platforms now. So yeah, I think PWAs

38:38

will continue, but I think it's more

38:40

or less like the feature set

38:42

of PWAs are going to become

38:44

more just generally used rather than

38:46

people focusing on building a PWA

38:49

progressive web app. Like for instance, I

38:51

mean, like now the stuff I do

38:53

always has the progressive web app stuff

38:55

about it. It has a service worker,

38:57

it has a manifest file, and an

38:59

icon and a theme color. It

39:02

works offline. It does all the stuff that

39:04

a progressive web app should do, but I've

39:06

never sat down and be like, let me

39:08

progressive web appify this thing. Yeah, yeah, you're

39:10

just using features that will be handy.

39:12

I'm throwing stuff into the cache. So

39:16

I hope so. We built a

39:18

little PWA for the syntax. At

39:21

GitHub universe, we had a little

39:23

iPad on the table and

39:26

it was just looping through the syntax shorts

39:28

that we had and had a big syntax

39:30

logo and some of the links. And it

39:32

was meant for like people walking by were able to

39:34

be like, oh, syntax, I've heard of that. And then

39:37

they would sort of come into the booth and then

39:39

we'd sell them on Sentry. But

39:42

I built just a little web

39:45

app for that. And then I used some of the

39:47

features like turning the tinting the URL

39:50

bar and being able to bookmark it

39:52

and putting a icon

39:55

on there. One thing I didn't do is I wanted

39:57

to figure out how to store the dot

40:00

TS files of a HLS stream I want to

40:02

figure out how to store those locally in the

40:05

browser so that if the Wi-Fi

40:07

were to go out they would the video

40:09

would keep looping hmm. I got something for

40:11

us Yes, this is gonna be

40:14

good. What's don't

40:16

look at the styling tools one. Yeah, okay so

40:20

Given the options which you could imagine

40:22

most of them, you know plain CSS

40:25

Panda Sass style

40:27

of components stylus tailwind those

40:29

types of things Which

40:32

is the most used and liked? And

40:36

then actually I'll give you like give me one two

40:38

three Most

40:40

used and liked ways

40:43

to write CSS. Yes, I'm

40:46

going to say vanilla number

40:49

one vanilla tailwind

40:51

and I

40:55

Want to say sass because like

40:57

nobody hates sass. I'm

40:59

gonna be I'm gonna be That

41:02

was really good Wes really you got the

41:05

you got the you got all three of

41:07

them correct You got the order wrong. It's

41:09

plain CSS then sass then tailwind Which

41:12

men props for nailing all of those? I mean,

41:14

it's easy to think about Potentially

41:17

any of these other hyped ones, but in

41:19

reality like what else is there CSS modules

41:21

is the next highest one? Which is pretty

41:24

predictable that it would be CSS modules and

41:26

then after that style components Shout-out

41:29

to max who was also at

41:31

our syntax meetup. Yeah author

41:33

of style components He's great to

41:35

talk to but yeah stylus sadly

41:37

enough three point five percent 56%

41:41

of people have no opinion on it. They don't even know what it

41:43

is 30% of

41:45

people not interested Yeah, rest

41:47

in peace stylus This

41:49

is a really good one to just look

41:51

at like awareness of like what is out

41:54

there right like we've talked panda CSS on

41:57

This this quite a bit, but like I

41:59

think it looks like almost nobody even knows

42:01

what that is. And

42:04

yeah, like Tailwind's huge, SaaS is huge.

42:06

And the SaaS one is interesting because

42:09

like we did it, again, we did a show

42:12

on like, can you drop SaaS? And like almost

42:14

all of the features of SaaS can

42:17

be done in plain CSS, you know, Nest

42:19

and Scobee. In fact, they're even better in

42:21

plain CSS because they run in the

42:23

browser. Yeah, and soon who knows

42:25

once we, once slash if we get

42:27

mix ins and functions,

42:30

that's all she wrote right there. I

42:32

guess maybe loops, but yeah. It's

42:34

gonna be, yeah, even like there

42:37

are some features, maybe we should

42:39

like do an update as to that because

42:41

there are now style queries that have come

42:43

out. There is now a proposal for mix

42:45

ins and functions in

42:47

CSS. Miriam Suzanne's working

42:49

on. We should, let's do

42:52

that episode again because when we recorded, we

42:54

said there is nothing but hopefully one day,

42:56

but now we have proposals for those last

42:59

few bits. And

43:01

especially with the style

43:03

queries, which is part

43:05

of container queries, but it allows you to

43:07

essentially do if statements in CSS. If

43:10

this Boolean is turned on, then apply all

43:12

of the CSS. Man, got

43:15

a lot, right? Yeah, got a lot

43:17

of it. Testing, who is

43:19

responsible for testing your software?

43:22

Developers and QA, mostly developers,

43:24

mostly QA. So 10% of

43:26

people saying mostly QA. So

43:28

yeah, that's more than I

43:30

thought. Remember we talked about that on a previous episode

43:32

is I didn't think that many

43:35

people had dedicated QA. And

43:37

it turns out that quite a

43:39

few people still have dedicated QA. Yeah,

43:42

QA big. People often are using

43:44

unit tests primarily, then end to end,

43:46

then integration. I personally do most things

43:48

in end to end just because of

43:51

make sense in a browser context. To

43:53

me personally, that said, I know unit

43:55

testing is like the

43:57

standard for just like, oh, I'm gonna learn testing.

43:59

and learn how to test a function. And

44:01

I'll still do occasional unit tests, but mostly it's

44:04

now just running the app

44:06

in something like Cypress, which has 42% usage

44:08

here, and

44:12

Playwright 36% usage, which

44:15

I thought Playwright at this point would

44:17

start to overtake Cypress for some reason.

44:19

I couldn't tell you why. And I

44:22

do wanna say I am

44:24

happy we're using Playwright on the syntax

44:26

site because, well, not just because of

44:29

this, but the, oh,

44:31

one of the, I don't know if it's the creator of Playwright

44:33

or one of the guys who works on Playwright did

44:37

a code review of our PR. We'll audit. And like updated

44:39

our things. It was like, your tests will run faster

44:41

this way. It was like, man, that

44:43

rules, thank you. That's so cool, thank you.

44:45

That's great. Yeah, I actually had somebody

44:47

email me the other day asking like, how

44:50

do I test this? And they

44:52

had like a user uploads

44:54

a PDF and the PDF is parsed and then

44:57

the parsing values gets fed through this next thing.

45:00

And they're like, there's no way for me to

45:02

like unit test some of the functions just written

45:04

in such a way. And

45:07

I think that unless you were to

45:09

do a bit of a rewrite and sort

45:12

of modify everything, the NTN testing is

45:14

maybe your best bet to have some

45:16

sort of security in that this thing still

45:19

works when you finish making

45:21

updates to it. Yeah, yeah, for

45:23

real. And they often have like

45:26

test recorders, which is nice, where you can just

45:28

click around the site and it'll record your actions,

45:30

right? That at least like the navigating parts of

45:32

using your thing. And that way you just have

45:34

to write the expects

45:37

or whatever, I expect this to do that. Code

45:39

editors, 75% using VS code, about

45:43

18% using JetBrains. I

45:46

also, the other day, JetBrains

45:48

announced the ability

45:51

to have a free access

45:53

for non-commercial use, which

45:55

I think is huge because like a lot of

45:57

times you'll say, why don't use JetBrains? I don't wanna pay for

45:59

it. You know, I think that's a

46:02

big move for them for you can now

46:04

get a web storm PHP storm All

46:06

of these storms the jetbrains IDE series for

46:08

free if you're not commercial So if you're

46:11

just doing a side project you want to

46:13

check it out You can get it for

46:15

free, which I think that's really exciting

46:17

to see that Three percent

46:19

on vim not even a

46:21

percent on sublime anymore poor sublime I

46:24

know hey and guess what it's over

46:26

for vs code cursor has officially killed

46:29

vs code by not showing up in this

46:31

chart at all Did we have

46:33

somebody leave a comment like too late cursors

46:35

already got it or something? Yeah. Yeah I

46:39

was so I was tweeting all of the stuff

46:41

that was announced at GitHub

46:43

universe and like I'm no I'm

46:46

no github stan. I'm no vs code stan. I

46:48

like it. I'm here for whatever's the best I'm

46:51

gonna get up star so I have to like

46:53

which yes got us as vested interest. I

46:56

got no vested interest but People

46:58

like too late. I was like brother. I

47:01

think you're underestimating the

47:03

fact that Like you have

47:05

owns all this stuff, you know Yeah, the

47:07

heavy hammer the the number of poor souls

47:09

that have you use Microsoft teams Because

47:12

they use outlook for their email is extremely

47:14

high Yeah For

47:16

real for versus slack it might not be the

47:18

best but like I don't I'm rooting

47:20

for cursor by working for VS code as well That's

47:23

I don't think it I think it's way too

47:25

early in this code editing game to even decide

47:27

a winner Yeah for real

47:30

and and again, it's not even on this chart.

47:32

So yeah, yeah browser

47:34

code editor Code

47:36

pen coming in high code sandbox

47:38

stack blitz JS fiddle 10% this

47:40

surprised me I Would

47:43

expect to see? GitHub code

47:45

spaces on here. I would

47:47

expect to see maybe Google

47:49

IDX on here You

47:51

know like the hosted VS code versions rather

47:54

than this but it

47:56

seems like people Are

47:58

maybe not using those? Although like I heard

48:01

the other day Shopify has all of

48:03

their devs working in the browser

48:05

in VS code, right? It's just like

48:08

an IT breeze is you have to

48:10

worry about access and having having code

48:12

on the local machines, all

48:14

of that stuff. But it seems like

48:16

from these answers like code pen, you're

48:19

posting cool CSS demos code sandbox. You're

48:21

posting your broken code stacklets. You're

48:24

posting your demos and

48:26

jazz fiddle. You're posting a 10 year

48:28

old stack overflow answer.

48:30

Accurate. Very accurate. It's

48:33

deadly accurate. Yes. Perfect.

48:36

And replit. It's one of these things

48:38

that I keep seeing. It

48:40

seems to be getting really popular. She's big

48:42

in the AI space. I need to give

48:44

that a fair shake because I feel like

48:46

it's becoming extremely popular outside

48:49

of my bubble. And

48:51

those are always interesting to me. It's when

48:54

people who are learning web development do

48:56

something different than what people who are in

48:59

web development, because that might be the next

49:01

thing, you know? Yeah, totally. A couple last

49:03

ones. Build tools. Veet by

49:05

far. Take it the cake. No surprise.

49:07

Veet rules. Big fan of

49:10

Veet. Second place is ES build. Wow.

49:12

Third place is web packs. A web pack.

49:15

Really feeling the heat now. Linting

49:17

tools. Yes. Lint. Everybody's

49:20

using ES lint. Everybody's using prettier. Everybody's

49:22

using and liked style lint. We use

49:24

style lint and I like it. It

49:26

keeps web class in check. Highly recommend, yeah. Yeah.

49:29

I'm a big fan. I've also been using markdown

49:31

lint a lot lately. It

49:34

does a good job at both formatting my markdown

49:37

nicely, but also telling me when I'm doing

49:39

bad things, like not nesting the levels

49:43

properly. The

49:45

poor thing, go back to the last one, which is web pack,

49:47

is almost as many people

49:49

use and dislike web pack

49:51

as use and like

49:53

web pack, which is a bit

49:56

of a shame, but it's just like a web

49:59

pack has become such a. huge thing and the

50:01

API is not easy to

50:04

sort of reason about. Whereas like

50:06

with, with V you almost don't even

50:08

need for a lot of projects. You

50:10

don't even need config. You just

50:12

use it. And then if you do need config, like

50:15

it's not hard. Like I have 120 line

50:17

V config, which is pretty

50:19

complex and it, that was, it was a joy

50:21

to build. I was really happy about that. Yeah.

50:25

Yeah. Word operating systems, majority of people

50:27

on Mac. Yes. What's up. That's kind

50:29

of interesting to me. Like doesn't, I

50:33

think most people are not on a Mac. If you were to,

50:36

to actually gauge, I don't know. What

50:38

would you think? Are you thinking if

50:40

you took every developer building JavaScript, do

50:43

you think there'd be mostly Mac? Ah,

50:47

yeah. I mean, I'm so biased here. I

50:49

can't give you a straight answer because I'm

50:51

going to just say, yes, that's what I

50:53

like and prefer. And I think it's a

50:55

smoother experience for me, but yeah, I mean,

50:57

you often hear people being like, who even

50:59

users Mac? I mean, a lot of people

51:01

using it. So, um, I

51:03

wonder if we were to pull our audience, I bet it would

51:05

be like 72% Mac,

51:09

but if you were, I bet if

51:11

you were to pull the wider audience,

51:13

you would get a bit more heavily

51:15

windows. Let's wrap it

51:18

up with the future trends in your opinion,

51:20

which of the following trends will gain popularity

51:22

and which will die. So the

51:25

highest trend of popularity is

51:27

performance. That's great to see

51:29

a response of ness. Wait,

51:32

it's no high. Oh no. Developed user experience

51:34

is 90%. Whatever

51:37

that means. User user experience

51:39

is that how is that different from development? It's

51:41

like how it is for the people using your

51:43

application. Sure. Yeah. User experience.

51:46

We want them to be faster, smaller, you

51:48

know, quicker apps. I guess that all falls

51:50

under user experience, right? Responsiveness.

51:53

That's I I'm hoping that doesn't

51:55

mean like responsive CSS. I think

51:57

everybody's writing responses that probably mean.

51:59

means like, like what I'm

52:01

reading for this performance is high, responsiveness is

52:03

high, server rendering is high. I think

52:06

that what that means is that we're

52:08

focusing on how do we make these

52:10

apps fast and feel good. So

52:13

what, what will lose popularity?

52:15

GraphQL 50% of

52:18

people think GraphQL will lose popularity,

52:20

which I think I agree with

52:22

not because it's bad. Again, it's

52:24

just not the tool for absolutely

52:26

everybody. I think it's clear where

52:28

the use cases of GraphQL shines.

52:31

Micro front ends about

52:33

tied 30% think it

52:35

will gain and 30% think

52:38

it will go down. It's a pretty good

52:40

split. Yeah. Missing on here. Uh, local

52:43

first stuff. Yes. Where's the

52:46

local data stuff on here? I think

52:48

that kind of falls into several of

52:50

these things just individually, but nothing specifically.

52:52

I think that's something they should be,

52:54

uh, should be tracking here. Cool.

52:56

All right. Well, that's an interesting

52:58

look at where we're at with our, the

53:01

whole web development scene right now. I always

53:03

love going through these and just getting a

53:05

good feel for where everybody's at. Yeah. Yeah.

53:07

Especially again, outside of our bubble,

53:09

I think is important here. So, um, cool

53:11

survey. I was really stoked to be a

53:14

part of it. So, um, shout

53:16

out to all of the folks who put this

53:18

survey together and including me in it. Cause it

53:20

was, uh, nice to do this kind of thing.

53:22

So shout out to state of front end 2024.

53:25

Give it a check. Let us know what you

53:27

think. How accurate

53:29

do you feel like this is

53:31

if you're out there and you're listening to us

53:33

talk about this? If do you feel this is

53:35

accurate? If you do, or don't leave

53:38

a comment below, best place to comments on YouTube

53:40

while you're there, feel free to give this thing

53:42

a thumbs up a like, a follow, a share,

53:45

um, send it to your grandma, do all

53:47

that stuff. So thank you so much. And,

53:49

uh, Spotify also now has, has comments. They,

53:51

Spotify, Spotify, Spotify, they, Spotify had comments for,

53:53

I dunno, probably about six months, a year

53:55

now, but you could never reply

53:58

to them. So it was just like people would. write

54:00

stuff and be like, cool, I have no

54:02

idea who you are or any, how to reply to you. But

54:04

now you can actually reply to

54:06

comments on Spotify, which I think is pretty

54:08

cool as well. So if you're in Spotify,

54:11

leave a comment as well. Word

54:13

sick. Cool. Well, let's get into

54:16

sick pics and shameless plugs.

54:19

Do you have anything today? Yes.

54:21

I'm going to sick pick the flighty

54:23

iOS app. So I've been using a

54:26

app called flight view to track flights

54:28

for probably 10 years now. And

54:31

someone put me on to using flighty, uh, which

54:34

is a beautiful iOS app. It's one

54:36

of these ones. That's just so well

54:39

designed, has all the features, really nice

54:41

tracking on it. Gives you

54:43

all kinds of cool stats. I used

54:45

it for my last two flights. I

54:47

was like, this is fantastic. Gives you

54:49

stats on like, like historically, this flight

54:51

has always been 12 minutes late into

54:54

the game or early and it updates

54:56

it. And it puts your flight information

54:58

right on your home screen as like

55:00

one of those, like, you know,

55:02

I don't know what that's called in iOS,

55:04

but like when you're doing something, it's like

55:06

an action of some kind. Yeah. If you're

55:08

shopping for Instacart, it will like show you

55:10

where your person is or your Uber will

55:13

show up. It does that with flighty. And

55:15

I think it's a pretty nifty habits. It's

55:17

one of those like, like tap bots or

55:19

whatever. They build beautiful apps. I feel like

55:22

it's in that range of, of good

55:24

apps. And I didn't pay for the,

55:27

the paid version. I used the trial and then I, I

55:29

don't know if I would, would pay for it, but apparently

55:31

you can also just buy it

55:33

for a week, which I love. Cause if you

55:35

are going on a trip, yeah, shell out for

55:37

a week of this. It's going to make

55:40

your, um, your life a little bit simpler.

55:42

Yeah. I am going to, uh,

55:45

sick pick something from our holiday

55:47

gift guide episode that Wes brought

55:49

to my attention.

55:52

Ooh. Yeah. So you

55:54

had shared a candle warmer

55:56

lamp. I

56:00

got Courtney a candle warmer lap for lamp

56:03

for her office And what was so funny

56:05

is that was like I was

56:07

trying to like I was trying

56:09

to get it without her noticing it obviously, but it's

56:11

coming in from Amazon and like I

56:15

It was it was as at my son's soccer

56:18

practice the it's like 5 30 or Getting

56:21

close to six the package arrives and

56:23

I have five minutes to get home

56:25

In fact, I'm on my way home

56:27

when the package arrives and I'm like

56:30

I gotta intercept that package I know

56:32

Courtney's making dinner right now. So therefore

56:34

I should be good to go. I

56:37

Walk in the door. She's already got the

56:39

candle warmer hooked up and music and she's

56:41

like, hey, did you buy this for me?

56:43

I was like no Was

56:45

it supposed to be a Christmas gift? Yeah, it was

56:47

gonna be a Christmas gift for office She's

56:49

like, did you get this for you? I was like

56:51

no I got for you for your office for Christmas

56:53

And I couldn't believe it. It's like five minutes since

56:55

she's somehow not like not only opened it but got

56:57

it going. Yeah So yeah, that's great Yeah,

57:00

I love it The the one thing about it is

57:02

that it as you use it

57:04

the smell comes out of the candle But

57:07

the candle doesn't burn down as quickly as

57:09

the smell gets away And

57:11

it makes your candles last way longer than if you

57:13

were to burn them But also you

57:15

wonder like where's that going, you know, like the

57:17

wax is I guess going into the air being

57:20

burnt All candles work

57:22

though Yes. Yes, but

57:24

I have been looking into can you

57:26

just buy? candle scents

57:29

and put them into the wax once the Because

57:31

you you end up with this candle that has

57:33

no more smell in it after like a three

57:36

or four weeks of using it But

57:38

you still have wax left so you

57:40

can just I'm curious if you can

57:42

essential oils. I bet yeah You just don't know that's

57:45

a fire Yeah,

57:48

yeah, don't do that I thought so

57:50

too, but I looked it up Essential

57:52

oils have alcohol in them. They're

57:54

not the right thing for candles will catch

57:57

on fire. So all right. All right lesson

57:59

learned there Yeah, cool.

58:01

I'm gonna shamelessly plug Syntax

58:03

on blue sky we're on blue sky

58:06

blue sky dot app. What

58:08

is the URLs for blue sky? It's

58:10

it's BSK y dot app. But what

58:12

is for syntax is it at syntax?

58:16

Syntax at syntax dot FM. We'll make sure we

58:18

have a link to that in the show notes

58:20

profile syntax out of them Check

58:23

us out. We've been posting there. I've been posting

58:25

there. It's a it's a man besides

58:27

just being a Twitter

58:29

alternative right which I

58:32

think that was like the big selling point

58:34

at first It's a Twitter alternative, but there

58:36

are some legitimately awesome forward-thinking

58:39

features that They

58:41

should have figured out at Twitter like 10 years ago.

58:43

So shout out to blue sky

58:45

for being I don't know There's a lot

58:47

of neat stuff going on there. Yeah,

58:50

it's It's a great

58:52

there's a good community on there. It feels

58:55

like old internet, you know And

58:57

I know we did a show on it like a year

58:59

and a half ago and it kind of fizzled out but

59:01

now there's like a second wave happening that

59:03

seems a lot more significant and

59:07

There seems to be a lot of like

59:09

web development talk on it Which is exciting

59:11

to me because like threads also did it

59:13

But I find a lot of the talk

59:15

on threads is just why people

59:18

don't like Twitter and like I just

59:20

want to learn about CSS You know, I want to

59:22

I'm coming because there's people talking about web

59:25

development stuff here Yeah, I feel

59:27

like that conversation is really good on blue sky right

59:29

now. It's great. Everybody's here it's

59:31

a it's the vibes are immaculate

59:34

currently before all of the Grifters

59:38

and who knows what has arrived but

59:40

yeah, I did get one in my

59:42

so there's a discover feed Which is

59:44

like like an algorithm based thing

59:46

that tries to show you things that you

59:48

like and I did get

59:51

like a History cool kids or some

59:53

sort of like one of these things

59:55

that just posts like photos of interesting

59:57

things like Facebook slop And

1:00:00

I was just like oh no here it comes here it

1:00:02

comes Yeah, I know

1:00:05

We'll get there eventually Another thing

1:00:07

was block block lists are public by the

1:00:10

way. I don't know if you knew really

1:00:12

you can see who blocked you Luckily

1:00:14

nobody who I know or care about has

1:00:16

blocked me so That's

1:00:19

cool. I Guess it

1:00:21

has to be though because of the decentralized

1:00:23

nature. Yeah, there's a lot of interesting enough

1:00:25

And by the way, we're having Dan on

1:00:27

the show Yeah, Dan Abramov who works at

1:00:30

blue sky formerly of the react team event

1:00:32

or react. Yeah Really

1:00:34

really stoked to talk to Dan

1:00:36

about this. So Yeah, blue

1:00:39

sky. What's up? All right. That's it. Thanks.

1:00:41

I'm for tuning in. We'll catch you later Peace