0:00

All right, we're rolling. All right,

0:03

welcome. Welcome to the show

0:05

Ian Landsman. Hello invited myself

0:08

on. You invite himself on

0:10

and we're using his Riverside.

0:13

This is notes on work

0:15

featuring Ian Landsman talking

0:17

me down. He's like, your message was

0:19

something like. I'm a third of the

0:22

way into these podcasts and I need

0:24

to convict him and do this. I

0:26

got to the first two. The third

0:29

one about how you were building a

0:31

micro charting framework and I was just

0:33

like, I got to get on the

0:36

horn to this guy. If you listen to

0:38

the first two, you get what's going on.

0:40

Yes. So I think I have a good

0:42

idea. Okay. So where do we start

0:45

here? Like, all right, where I'm sitting

0:47

right now. After these podcasts, I'm

0:49

feeling pretty good. I was feeling

0:51

pretty bad and now I'm feeling

0:54

pretty good. Okay. Tell me why

0:56

I shouldn't feel good. So, so

0:58

with the conclusion, I couldn't get

1:00

to the third one. So the

1:02

third one, did you finish? I'm

1:04

gonna build a micro line only

1:07

charting framework thing. Yeah. Okay.

1:09

Alright, so here's my point. It's

1:11

going to be so sick. I forgot

1:13

a few problems with that. Okay. Just

1:15

in the first two podcasts, right? Like

1:18

it's incredibly complicated. And a hundred times

1:20

you're like, do we need that? Do

1:22

we need labels? We don't need labels?

1:25

Do we need, you know, labels? We

1:27

don't need any bars. Do we need all

1:29

this stuff? I guess here's the thing to me.

1:31

Anybody who's really using a chart

1:34

library, who actually had data to

1:36

chart in a serious application? needs

1:39

the full charting library. I

1:41

don't know. So I'm looking

1:43

at stripe and they literally have

1:46

a line that you can offer

1:48

the points. Oh God. And get

1:50

a little tool tip of what

1:52

data is at what points. And

1:55

they have two labels in the

1:57

x-axis that says the beginning time.

1:59

and the end time and that's

2:02

what they have for like most

2:04

of the app. Well they have

2:06

bar charts, I've seen bar charts,

2:08

there's line charts, so there's multi-line

2:10

line charts, multi-line easy, got it,

2:12

okay. Listen, the only things, ticks,

2:14

and labels, which you said ticks,

2:16

or maybe not gonna have it,

2:18

I can't do his ticks. It's

2:20

a big one. I literally can't

2:22

do this. Writing writing stuff. In

2:24

the SVC is this a problem?

2:26

It sounds like it was maybe

2:28

a problem. I have text and

2:31

text. Yeah, yeah, like dates and

2:33

times, numbers and things. The generating

2:35

of it's not necessarily a problem,

2:37

it's the spacing of it. It's

2:39

not, they're not divs. They don't

2:41

adjust to each other. They're just

2:43

painted over each other. Right. And

2:45

I say to put them on

2:47

the X-axis. Yeah. You have to

2:49

figure out what you're going to

2:51

do with all that. There's a

2:53

lot of stuff in it. No,

2:55

I don't, because I'm not doing

2:57

ticks. What do I do with

3:00

a chart with no ticks? So

3:02

a lot of charts have no

3:04

ticks. They have the beginning and

3:06

end date or kinds. Men and

3:08

Max, why? The people who need

3:10

such a ridiculously simple charting thing

3:12

are coin base, stripe, lemon. No,

3:14

they're not. Stripes are not like

3:16

that. What are you talking about?

3:18

We're on it right now, we're

3:20

pulling on it. I was just

3:22

in there looking at this. But

3:24

I will look again. There was

3:26

bar charts. You always see multiple

3:29

kinds of charts. I do actually

3:31

see these bar charts. Yeah, I

3:33

just saw them. I literally saw

3:35

them before I got on the

3:37

call. If you go on reporting,

3:39

go on reporting. Show me. Okay.

3:41

Let me to it. Yeah, I

3:43

can't show you, but just tell

3:45

me. Morning, I was just in

3:47

there. Okay, so if you go

3:49

reporting overview, there's bar charts with

3:51

numbers, year over year, smaller, your

3:53

percentage difference. Reporting, overview, gotcha, okay.

3:55

Okay, now you click that little

3:58

box that expands it and you

4:00

get a line chart over the

4:02

bar chart, you get tooltips over

4:04

everything, you get ticks, you get

4:06

labels, you get interactivity when you

4:08

click it. Okay, okay. A lot

4:10

of stuff in there. Yeah. y-axis

4:12

to go with your left-hand y-axis?

4:14

So yeah yeah yeah okay right

4:16

right there are there are things

4:18

there are some things like that

4:20

that I think particularly on stripe

4:22

when I see something and it's

4:24

like I could mostly do that

4:26

but then oh like on stripe

4:29

they have they'll show like a

4:31

progress line Where like they'll put

4:33

a lot like render half a

4:35

line it extends. It's like yeah

4:37

estimated continuation or stuff like that

4:39

like that stuff is like oh

4:41

That's that last bit and I

4:43

wouldn't even know how to do

4:45

that in a normal charting library

4:47

like so All right, let's look

4:49

at this bar chart I think

4:51

bars are within my future I

4:53

think I could do them right

4:55

now if I thought about it

4:58

hard enough Their ticks are these

5:00

ticks on this bar chart this

5:02

so bar charts use categorical data

5:04

categorical x-axis so you all are

5:06

going to render every single one

5:08

and if there's not enough room

5:10

you can so the way that

5:12

I'm building it let me get

5:14

this I don't know if this

5:16

helps you at all because I

5:18

know I know what you're saying

5:20

I know what you're thinking this

5:22

is good but this I'm going

5:24

to just completely run you off

5:27

yes so mine is going to

5:29

be driven by like SVC's that

5:31

you deal with as a developer

5:33

and it's not that bad it's

5:35

like you sort of build your

5:37

own SVC So you would say

5:39

template because all these other libraries

5:41

are basically doing that except just

5:43

removing you one like Recharts have

5:45

you have used recharts at all?

5:47

No, I don't know about recharts.

5:49

So basically you say like chart,

5:51

like that's my, this is what

5:53

Chadsey does as well, like so

5:56

you got like chart component, then

5:58

inside of it you'll have line,

6:00

and online you can specify stroke,

6:02

stroke width, stuff like that. Right.

6:04

That's just a wrapper around a

6:06

path, literally an SVC path, and

6:08

all those attributes get passed down.

6:10

And that's fine, but it's just

6:12

like an unnecessary removal of like

6:14

one level, where so mine's gonna

6:16

be like. line, but it's literally

6:18

a path element. So you can

6:20

add tail-in classes, you can add

6:22

whatever attributes you want. So what

6:25

this does is give you like,

6:27

now you're in the SVC. So

6:29

you can use any CSS you

6:31

want, you can add to the

6:33

SVC. Like you can add your

6:35

own deaths at the top of

6:37

like linear gradients and transitions and

6:39

whatever you want to apply. So

6:41

what this means and why I

6:43

told you this is because if

6:45

you're using my thing. If you're

6:47

rendering bars, and there's not enough

6:49

room for the labels of the

6:51

x-axis ticks, you can rotate them.

6:53

And then they'll just be rotated

6:56

and they'll fit nicely together. It's

6:58

like you just render it and

7:00

you go, oh, that doesn't look

7:02

right. Rotate. And so I have

7:04

to do a lot of work

7:06

in this. Yeah. Like way more

7:08

than with chart JS or equivalent

7:10

library. So the problem, I mean,

7:12

so chart, this is the other

7:14

thing, is like, what's the alternative.

7:16

I mean a little bit, not

7:18

in serious anger. I mean, high

7:20

charts a lot. Okay, high charts.

7:22

Chart JS a little bit. High

7:25

charts is a paid thing. Okay.

7:27

Charts and charting libraries suck because

7:29

there's 10 million configuration options. A

7:31

lot of configuration. But you don't

7:33

even understand. You can simplify. See,

7:35

to me, this is the whole

7:37

thing. Where does Caleb provide value?

7:39

Okay. It's the attraction and design

7:41

and simple application. Yes, give me

7:43

the API, give me, make it

7:45

awesome. Like, don't, like, do you

7:47

building a charting library? is zero

7:49

value, there's already a million charting

7:51

libraries. They all have 6,000 closed

7:54

issues. When are you going to

7:56

work on the 6,000 closed issues

7:58

that are going to come up

8:00

inevitably once you release this thing?

8:02

And people are like, it doesn't

8:04

render in this weird way, right?

8:06

Like, all that stuff. The date

8:08

stuff was one thing. I always

8:10

think it's the date thing too.

8:12

But this is even worse. This

8:14

is high five times better. No,

8:16

I think this is five times

8:18

better. No, because I can't. No,

8:20

because it doesn't have the charts.

8:23

You're not going to build, you

8:25

just said you're not going to

8:27

build all the charts people need.

8:29

Are you going to build a

8:31

scatter plot? Are you going to

8:33

build a pie chart? No way.

8:35

Okay, these are the types people

8:37

you use in real applications. When

8:39

you need those, you put in

8:41

D3 or whatever. Like, I'm not

8:43

even going to be able to

8:45

make a wrapper around any chart.

8:47

It's going to look like the

8:49

rest of flux. And obviously it's

8:51

going to open up like there's

8:54

some way to pass in the

8:56

like config stuff. If you want

8:58

to get down to the like,

9:00

I need to set my chart

9:02

JS or whatever library, like I

9:04

need to get deeper in there,

9:06

great, you could pass in the

9:08

config. But the like passes data

9:10

from Laravel into it, past, you

9:12

know, the feaming default to looking

9:14

like flux. Like I think the

9:16

base case could be magical without

9:18

you building a chart library. Which

9:20

is a ton of work. And

9:23

it will never ever ever ever

9:25

ever be as good as these

9:27

other chart libraries. Right. Because you're

9:29

not going to put that kind

9:31

of work into it. It's not

9:33

going to be as powerful as

9:35

these charting libraries. But that's what

9:37

you need in the chart library

9:39

though. But it will be simpler.

9:41

But nobody wants that. Who wants

9:43

a simple chart library that goes

9:45

one thing? I do. That's no.

9:47

This is Ian. This is where

9:49

we are fundamentally misaligned. Are you

9:52

ready? Yes. I'm ready. I want

9:54

a minimal charting library. I always

9:56

have any web app that I've

9:58

been in and building. I go,

10:00

uh-huh. I need a chart at

10:02

the top of this table to

10:04

show revenue over time. Yep. That's

10:06

almost always what it is. Maybe

10:08

usage over time, maybe memory over

10:10

time, but it's something over time

10:12

and it's usually money. And I

10:14

go, what am I going to

10:16

use? Okay, well, pull in chart

10:18

JS. I have to learn the

10:21

words data set and I have

10:23

to get in like to tick

10:25

callbacks because I don't want. when

10:27

you go to chart j s

10:29

and you pull in the default

10:31

it is loaded and looks like

10:33

not what you want exactly same

10:35

thing with return same thing with

10:37

d3 then you need one little

10:39

configuration and now you're like in

10:41

the whole thing again you know

10:43

I guarantee you are capable of

10:45

making that clean and easy for

10:47

me I hate it I make

10:49

it clean and easy for me

10:52

because I want to be able

10:54

to reach in there and get

10:56

all the bag of tricks and

10:58

you're never going to build those

11:00

bag of tricks which means that

11:02

I know I will always have

11:04

to use a different charting library

11:06

I will never use fluxes and

11:08

I think a lot of people

11:10

will not be able to use

11:12

fluxes because as soon as I

11:14

need literally anything that's not the

11:16

ultimate most simple use case it's

11:18

a flux charting library is totally

11:21

used. Okay then let me what's

11:23

both I'm gonna I'm gonna steal

11:25

man you or at you or

11:27

at I'm going to at least

11:29

go there. And I want you

11:31

to go here. I'm going to

11:33

go there though. Okay. I don't

11:35

know what to use. Because of

11:37

the library or something? Yes. Yeah,

11:39

I mean, I don't know what

11:41

to use either. I mean, there

11:43

is no great option here. You

11:45

know, if it was in react,

11:47

I would use recharts. Recharts to

11:50

me, feels good. It's good enough.

11:52

It's D3 under the hood. It's

11:54

a fairly lightweight wrap around it.

11:56

they have these transitions you might

11:58

need whatever but and you and

12:00

they have all the extra configuration

12:02

that is like a little weird

12:04

like you know that you might

12:06

want to customize so so right

12:08

there why don't you just do

12:10

that it's order yeah but poor

12:12

Oh, I think at that point,

12:14

I think that's what I'm doing.

12:16

I think I'm building it. But

12:19

you're not wrapping in D3 and

12:21

giving us all the fun different

12:23

charts and stuff, right? Oh, so

12:25

I am oscillating back and forth

12:27

between using D3. But yes, on

12:29

the last episode that you heard.

12:31

It's no D3. This is the

12:33

thing about D3. And that, like,

12:35

if you, even just on recharts,

12:37

they say, I think on their

12:39

home page. So they say, okay,

12:41

built on topics SVG elements with

12:43

a lightweight, don't throw the ticks

12:45

in my face. It's so funny.

12:48

The most like basic charting feature.

12:50

The ticks, right there, checks inside

12:52

the box. Recharts. built on top

12:54

of SVC elements with a lightweight

12:56

dependency on D3 sub modules. So

12:58

I'd have to dig into their

13:00

actual repository to see what they're

13:02

doing. But my guess is they're

13:04

kind of doing what I'm doing,

13:06

which is like, I don't know,

13:08

let's just see if I don't

13:10

know, it's too long for me

13:12

to actually look at one of

13:14

these bar charts that TSX. Yeah,

13:16

I don't know. They're probably doing

13:19

what I'm doing, which is using

13:21

the D3 math. Some of it

13:23

is really valuable and some of

13:25

it is just like you could

13:27

have like three functions built by

13:29

clients that are going to do

13:31

this. So anyway, what I'm saying

13:33

is, I can't... I mean, this

13:35

recharts, recharts has 2,800 closed issues

13:37

and 440 open. Chartjas has 6,800

13:39

closed issues and 400 or something

13:41

open. These are huge, gigantic libraries

13:43

that do a lot of stuff

13:45

and have a lot of weird

13:48

edge cases. The scope that they

13:50

like take on is insane. But

13:52

your but the super simple scope

13:54

is just so not like this

13:56

it feels not aligned with flux

13:58

to me. I guess that's my

14:00

problem. Like I don't view flux

14:02

as like this is the ultra-

14:04

simple, you know, components for your

14:06

really stupid, simple projects. Don't play

14:08

anything realness. This is not for

14:10

real projects. This is for your

14:12

very simple, don't, nothing complicated, super

14:14

easy. I don't know, dude. I

14:17

don't know. I mean, you know,

14:19

no, no, no, no. No, it's

14:21

definitely not it. It's, like, with

14:23

the date picker. Take the date

14:25

picker as not. The date picker

14:27

does everything you would get out

14:29

of any other date picking library

14:31

that you could find. Yes, exactly.

14:33

Yeah, exactly. That will not be

14:35

true of charts, right? It will

14:37

not be true of charts, right?

14:39

It will not be true of

14:41

charts. It will not be true

14:43

of charts. It will not be

14:46

true. So that's the difference. So

14:48

you took two months and built

14:50

dates, pretty much. with the other

14:52

day things out there. I'm sure

14:54

there's some edge cases, whatever, blah

14:56

blah blah. But it's like 90%

14:58

of the other ones. You are

15:00

not going to be 90% on

15:02

this chart library in six weeks

15:04

to edit these big chart libraries.

15:06

Tell me. Unless you are. I

15:08

mean, maybe it will be. It

15:10

seems very unlikely. This chart is

15:12

going to be done by the

15:14

end of the week. Line charts,

15:17

multi-line line charts with coverable tooltips

15:19

and basic ticks. Zero dependency and

15:21

cumulative weight of like two kilobites

15:23

will be done by the end

15:25

of the week. Won't have ticks.

15:27

Right, won't ticks? Do I have

15:29

anything? This is the thing I

15:31

need you to at least acknowledge.

15:33

The majority of web apps and

15:35

any time I've needed charts and

15:37

web apps, I have only needed

15:39

what I have just described to

15:41

you. What I could have done

15:43

in a week. I build, right,

15:46

that's 100% not true. We have

15:48

like five different types of charge.

15:50

We have really complicated charge. We

15:52

have a heat map. You have

15:54

a multi-area area one. Yes. We

15:56

have three lines with a ball.

15:58

underneath like so I don't I

16:00

have not built the real app

16:02

that's just like here's a single

16:04

line line chart or whatever that's

16:06

an ultra simple line chart I'm

16:08

not I've not personally worked on

16:10

something like that. So and what

16:12

you're describing heat maps and whatever

16:15

like there's no that's not like

16:17

a off-the-shelf component lab you know

16:19

I'm not gonna offer heat maps.

16:21

Like what if you have to

16:23

build it in this scenario? That's

16:25

my point. I need why you

16:27

don't need to build it all?

16:29

I need a junk drawer. Like

16:31

what I need is I want

16:33

this charting library, but I also

16:35

want as soon as you reach

16:37

the end of it, you go,

16:39

I need D3. And then there's

16:41

some way for you to just

16:44

use D3 or whatever or chart.

16:46

I mean, you could just use

16:48

chart. You could use any of

16:50

these things and I could provide

16:52

like. Yeah, if you want to

16:54

hear a simple one, right? But

16:56

then you also have some type

16:58

of Caleb, nicely integrated chart JS

17:00

thing, that's probably only going to

17:02

take you another week or whatever,

17:04

right, to come up with a

17:06

reasonable API and the components and

17:08

whatever to pass things around and

17:10

let you get at the underlying

17:12

config if you need to, makes

17:15

it look like the rest of

17:17

flux and react to its beaming

17:19

and blah blah blah blah. Then

17:21

you have that then you have

17:23

like basic charts and advanced charts

17:25

advanced charts. I'm shipping you over

17:27

here charts That's a good a

17:29

good way to but with a

17:31

helper with a little helping hand

17:33

so that you don't have to

17:35

spend three days figuring out how

17:37

to like make chart jas look

17:39

like the rest of your completely

17:41

flux app like Caleb's already done

17:44

that part. It all works nice

17:46

together and if you need all

17:48

these other feet on that I'm

17:50

not recommending chart jas, but Whatever

17:52

one of them right obviously there's

17:54

multiple ones But there's an integration

17:56

there that feels Organic if you

17:58

have a more advanced need. There's

18:00

not a lot of good stuff

18:02

outside of react so it's like

18:04

hard to just pick one like

18:06

we could use chart j s

18:08

it and so this is my

18:10

fear is is like if you

18:13

look at filament charts they I'm

18:15

pretty sure they use chart j

18:17

s and the experience of using

18:19

it is like you just have

18:21

it's so much configuration and whatever

18:23

like somebody who needs the thing

18:25

just needs the thing and they're

18:27

gonna eat whatever shit they need

18:29

to eat so that's 100% true

18:31

but it's like my answer I

18:33

can't part of that's each chart

18:35

It doesn't make any sense. Like

18:37

you should do that yourself. You

18:39

need this. Well, you're a web

18:42

developer, just do it, you know,

18:44

like. Sure. But I feel like

18:46

that's not the promise of flux,

18:48

right? I don't know. I don't

18:50

think the promise is like everything

18:52

you'll ever need ever in a

18:54

web app. It's like, you know,

18:56

I don't know. Maybe it is.

18:58

But like, I don't know if

19:00

it's reasonable for me. Like if

19:02

you go to Shadsey, I mean,

19:04

This doesn't apply. If you go

19:06

to Shadsey-N, like, they're showing you

19:08

simple bar charts, simple line charts,

19:11

like, very simple, out of the

19:13

box, what you need. But of

19:15

course, this whole argument, the points

19:17

moot because you can just configure

19:19

recharts. I mean, you go on

19:21

their charts, there's all kinds of

19:23

stuff. There's pie charts, there's all

19:25

different colors, there's stacked bar charts,

19:27

there's a stacked bar charts, there's

19:29

negative. Bar charts, there's like hundreds

19:31

of charts in here. I, radar

19:33

charts. Yeah, at the very top

19:35

navigation click charts. There's radar charts,

19:37

there's radial charts. I was just

19:39

in the chart component. Yeah, yeah.

19:42

Yeah. Yeah. This is what I

19:44

want. I want this. All right.

19:46

Let me build it for you.

19:48

Yeah. Go with me here. Yeah.

19:50

First chart. We have to just

19:52

unfortunately describe this. listener. First chart.

19:54

This I could show, oh well,

19:56

I guess, I don't know if

19:58

we're going to put the video

20:00

out. No, no video. No video.

20:02

That's just, yeah, that's fine. So

20:04

first chart, two lines, both have

20:06

area underneath it and categorical labels,

20:08

Jan Fab March, April, May, June,

20:11

and then horizontal, horizontal lines. Like,

20:13

yeah, and then the tool tip

20:15

with the values and the colors.

20:17

This is, this is basically done.

20:19

This to me is what I'll

20:21

have, MVP ready. Bar charts, I

20:23

gotta figure out bar charts. All

20:25

right, so this one, but this

20:27

one's doable. Like if I decide

20:29

to do bars, this is very

20:31

doable. Pie chart, this is where

20:33

I had a, so I had

20:35

a talk with Claude and Claude

20:37

thinks I, that line charts are

20:40

like 80% of the use case

20:42

like that I need, or maybe

20:44

70%, bars are like 20% and

20:46

then this pie stuff is like.

20:48

Like, and then there's a lot

20:50

of like people like there's warnings

20:52

I guess like it's not always

20:54

a good practice even to use

20:56

a pie chart like I don't

20:58

like this angle you're going on

21:00

here. People find it like confusing.

21:02

But see here's the thing that's

21:04

not that's not your issue because

21:06

are you making an educational site

21:09

for the best way to people

21:11

to chart or are you giving

21:13

developers what they've been a tool

21:15

that they can use to build

21:17

what they've been told to build

21:19

or that they think they need

21:21

they need to build or that

21:23

they need to build or that

21:25

their customers are asking help spot.

21:27

My customers want certain types of

21:29

charts. Maybe it's not the best

21:31

type of chart, right? But maybe

21:33

that's not my job to tell

21:35

them you're being an asshole and

21:37

you're being stupid. Why are you

21:40

using that kind of chart? No,

21:42

we don't have that kind of

21:44

chart because it's stupid. Why are

21:46

you doing that? I'm just going

21:48

to give them the kind of

21:50

chart they want. So like, that

21:52

I think is where to me,

21:54

like that should be your mindset

21:56

in my opinion. Oh man, there's

21:58

no bar chart that's stacked. Like

22:00

and... So you can then have

22:02

a conversation with them about why

22:04

it's dumb and Claude said that

22:06

you shouldn't use pie charts or

22:09

whatever, right? Or you can make

22:11

the sale and say, I'd be

22:13

like, yeah, we got pie charts.

22:15

Give me $600 now. Great, I

22:17

got your $600. You have

22:19

pie charts. Everybody's happy, which

22:21

is what I'm thinking is the

22:23

better option. Okay, so let's say,

22:26

let's do it. We're gonna do

22:28

it. I asked Claude what are

22:30

the non-react open source charting libraries

22:32

so D3 number one that's kind

22:34

of out like do you think it's

22:37

like a lower level thing right is

22:39

that correct level like you can't just

22:41

be like line me you know

22:43

you really have to build up

22:45

like the SVC yourself you're still

22:47

creating yeah the problem with chart

22:49

JS well one of the many

22:51

problems is it uses canvas and

22:53

not SVC okay which is like kind

22:55

of fine but makes it just

22:57

way heavier and whatever. Why do

22:59

I care? Flexible. You might not

23:02

care. I don't care. I 100%

23:04

don't care. All right. I could

23:06

imagine caring less to be honest.

23:08

Plotty. Plotty. Plotty. J. Plotty. J.

23:10

Plotty. J. Plotty. J. Ploty. J.

23:12

I mean, I guess I'll just

23:14

have to look into some of

23:16

these and see what's going on.

23:18

But I bet high charts, but

23:20

that high charts, but that's a

23:23

commercial license. Yeah, I've heard people

23:25

use high charts. I've definitely heard

23:27

of that. Whoa, dude, yeah. I

23:29

don't know. Or the recharge thing,

23:31

is there a way to, is there

23:34

a way to port it? Could you

23:36

and call do some work on porting

23:38

it? I don't know, it depends how

23:41

it's built, right? You initially said it's

23:43

a thin layer around the three. So

23:45

if it's actually a thin layer around

23:47

the three, maybe you can just port

23:50

it. Did I say thin. It's probably

23:52

a sick layer. A sick layer around

23:54

the three would be something different. Again,

23:56

I don't love that option because then

23:58

you are still maintaining it. I'll prefer

24:00

you to be maintaining the API, the

24:02

flux API to it is the thing

24:05

Kayla maintains. Not Kayla maintains a whole

24:07

charting library. Yeah. All right, go with

24:09

me. Tell me. I'll come to you

24:12

now on this. Come to me. Tell

24:14

me the world where this works and

24:16

where this is a good idea. Give

24:19

it to me. Where, which part? Your

24:21

side. Give me my side. Give me.

24:23

Yeah, give me even just a pro

24:25

of the light library. Okay, see, I

24:28

think a lot of things you think

24:30

are pros, I definitely don't think are

24:32

pros. Some of that is just built

24:35

into a thing, like developer, there's some

24:37

things that are just differences between you

24:39

and I, and that sort of like,

24:41

like a segment, the developer community of

24:44

like two groups of people. Right. And

24:46

I am firmly in this group and

24:48

proud. So like, that's a part. So

24:51

we're not going to change that. So

24:53

I guess the thing I like, if

24:55

you're saying that. If you're saying the

24:57

ultralight. thing is going to be done

25:00

next week or whatever, right? Then I

25:02

think that's definitely a benefit. That's probably

25:04

the biggest benefit to me. Because I

25:07

think you going off for two or

25:09

three months while you have like competitors

25:11

and all this stuff, you've got to

25:14

be out there selling, right? So if

25:16

you disappear to build a chart library

25:18

for three months, I think that's a

25:20

disaster. So I think in that regard,

25:23

it's like, hey, it's out next week,

25:25

it's for lines. Maybe you find out

25:27

that that truly is enough and nobody

25:30

really wants anything else. And fine. And

25:32

when they do, you just say go

25:34

obviously, you can just implement whatever you

25:36

want, just load up chart jazz or

25:39

whatever you want to do. And I

25:41

think that's fine. I guess to me

25:43

that wasn't like, you know, it's fine.

25:46

It's not exciting. It's not like there's

25:48

going to be a page on flux.

25:50

It's like, oh, would you click charts?

25:53

Like when I click charts on Shad

25:55

CNN. I'm like, damn, this has everything

25:57

I could want, whatever I need to

25:59

build, this thing has it. Whereas when

26:02

I click. on Flux, there's going to

26:04

be a line chart that might not

26:06

have ticks, and that's going to be

26:09

the whole thing, right? And so that

26:11

is not going to make me feel

26:13

like, oh, this is no matter what

26:15

criteria comes up in a year from

26:18

now, when someone says, hey, we got

26:20

this new feature, we got a little

26:22

chart, and I'm going to be like,

26:25

oh shit. Now I can only use

26:27

line charts. So now I got to

26:29

bring in another package to do a

26:32

different chart to do a different chart.

26:34

this chart is going to look totally

26:36

different than this other chart. So I'm

26:38

going to figure out, can I make

26:41

them look the same? I don't know,

26:43

like you've got this custom methodology thing,

26:45

how am I going to figure out

26:48

how to back change that into the

26:50

make it look like this new chart

26:52

yes thing I have to do or

26:54

whatever? So, you know, I think just

26:57

in terms of like overcoming objections, it's

26:59

hard when you're not going to have

27:01

a really flushed out offering there. If

27:04

this is important, maybe it's not important.

27:06

Yeah, yeah, yeah, yeah, yeah. Maybe you

27:08

shouldn't have anything. I mean, that's that's

27:10

another thing too. Like, this is a

27:13

thing that people don't think about, but

27:15

you could think about it the other

27:17

way. It's like, maybe you shouldn't have

27:20

charts at all. Right. Is it better

27:22

to have nothing than to have something

27:24

that's weak? Right. Potentially. I don't know

27:27

if that's true in this case, but

27:29

it might be. big item that people

27:31

want included. It's tricky. I agree that

27:33

that this is foolish in like if

27:36

I'm thinking about like flux as a

27:38

business and where time goes and where

27:40

users of flux want time to go,

27:43

even what I want to be doing.

27:45

It's like I don't think I want

27:47

to be like fiddling with like charting

27:49

bugs all day. I think it was

27:52

that I started on this, I think

27:54

I started on this path thinking like,

27:56

let's just see what I can do

27:59

with charts, totally open mind, and got

28:01

into it and realized like, there's nothing

28:03

good and to. do the good thing

28:06

yourself you end up pretty much doing

28:08

everything yourself anyway with D3 and then

28:10

I sort of found some things that

28:12

I was like excited about was like

28:15

wow I could the thing that you

28:17

know that speaks to me that's different

28:19

thing that I always wanted was a

28:22

super lightweight SVC charting library like how

28:24

hard could this be like a few

28:26

kilabytes give me a kill me with

28:28

the kilowites dude it matters to me

28:31

it doesn't matter it's irrelevant So, all

28:33

right, I mean, Chuck Jass, I just

28:35

saw, is tree shaking. So, in theory,

28:38

you know, only ship the kilabytes that

28:40

you use. So, I mean, right, but

28:42

with Charles, that's, it's gonna be a

28:45

lot. Like, I'm sure it's 100. Okay.

28:47

That's what the matter. Okay. So, I

28:49

get it. Yes, yeah. So you have

28:51

that. So you have the light, it's

28:54

light. But you really aren't like, like,

28:56

like, I thought the whole thing would

28:58

be rocket science. And then I opened

29:01

it up and went, a bunch of

29:03

this is very simple. I'm a major

29:05

fan. I came across in the podcast

29:07

because it's like I can't even do

29:10

ticks is kind of where we're at.

29:12

That doesn't seem like, that's like the

29:14

most basic part of the choice. The

29:17

rocket science. You're like, you're like, we're

29:19

already at rocket science for the ticks.

29:21

Like, there's a lot of other things

29:23

in there though. I was pretty breezy

29:26

until I got in the ticks. Like,

29:28

like, but even just the things we

29:30

first talked about, like the X-axis labels

29:33

labels. Yeah. Now I have to figure

29:35

out the spacing of the exact labels.

29:37

What if the data is variable? Again,

29:40

our charts, it's always variable. It's not

29:42

just like always 12 months. It's going

29:44

to be my time, it might be

29:46

48 months, might be five years, might

29:49

be one year, might be whatever. So

29:51

now I have to come up with

29:53

all the math to figure out like,

29:56

well, how wide is this thing? chart

29:58

config chart dot label I would like

30:00

to know how much configuration is about

30:02

like in some of these shad examples

30:05

like is he saying somewhere like in

30:07

this is all recharts raw are there

30:09

parts of it that are like yeah

30:12

tick formatter so it's a callback that

30:14

receives the day okay so he's there's

30:16

okay so here's a component called x-axis

30:19

for recharts okay And it's got the

30:21

key, the data key, you know, that

30:23

you're using, tick line if you're rendering

30:25

that or not, axis line if you're

30:28

rendering that. Okay, tick margin. Also find,

30:30

like, this is all stuff that can

30:32

be configured, done in an SVC, like

30:35

you can literally just size the SVC

30:37

differently. I know you don't want to

30:39

get in SVC. Sure. Min-Tik gap. Yeah,

30:41

that's where it's where it's like. So

30:44

start to get into like it's doing

30:46

so much aware of some stuff I

30:48

really want I shouldn't mess with the

30:51

recharge and see what happens if you

30:53

put like that that to be like

30:55

64 like doesn't how to you know

30:58

Yeah So then I've used in the

31:00

past two there are like you can

31:02

it'll drop off labels and things if

31:04

it needs to with some kind of

31:07

smartness, which don't always work great to

31:09

be fair. They are definitely downside there.

31:11

But what else is out here is

31:14

like a formatter. So the tick formatter

31:16

is a callback that accepts the date

31:18

object and then you are in charge

31:20

of converting a date object into English

31:23

and you would use JavaScript international date

31:25

time. Right. Formatter and whatever. So that's

31:27

like, yeah. I mean, the hard part

31:30

about my thing. is like, how would

31:32

you even, like if I create, if

31:34

I had a chart JS or something,

31:37

where am I giving you that configuration

31:39

ability? You know, where are you writing

31:41

tick formatting code? You're in the API

31:43

guy, I'm not the API guy. You

31:46

know what I'm saying though? I do.

31:48

It's like with this, you're already in

31:50

job. script so you can just pass

31:53

a callback you're already in react with

31:55

backendy stuff and live wire stuff it's

31:57

like you can't write that in PHP

31:59

because that's on the back end and

32:02

you don't on your chart to have

32:04

to like call back every time it

32:06

needs to play well label. But wouldn't

32:09

you for you send it in formatted

32:11

that's how I've always handled it. So

32:13

have you interesting have you formatted the

32:15

labels of the like do you generally

32:18

send in the labels to a chart

32:20

formatted that's what I used to do.

32:22

Yeah, that's what I would do. Generally

32:25

speaking, I mean, I don't know if

32:27

I've done it that way every time,

32:29

but that's usually how I would think

32:32

about it. I guess I'm sending, I'm

32:34

giving the chart library the exact stuff,

32:36

right, and formatted and ready to go,

32:38

and then it's charting. The part where

32:41

it gets tricky is if you have

32:43

like a thousand data points within two

32:45

months or something, you don't want the

32:48

ticks to be on a time interval

32:50

and display the time, which you can't.

32:52

No, I had, you know, like you

32:54

can't format that I had unless you

32:57

knew how many ticks would be rendering.

32:59

That's where it's true. They do it

33:01

stupidly, right? Like it's just like, well,

33:04

we can only show one and every

33:06

10 ticks. Right. So you would just

33:08

do one every 10. Like we're not

33:11

like relabling it necessarily to be grouped.

33:13

That you could do that then have

33:15

some grouping or whatever, but that would

33:17

be a separate chart. Into your heart,

33:20

Ian. Yeah, do it. What if you

33:22

have to take this at face value?

33:24

What if I could create a line

33:27

charting library and bar charting library that

33:29

did all of these things all that

33:31

automatic tick stuff all that formatting stuff?

33:33

Space stuff line multi lines areas bars

33:36

stacked bars horizontal grids vertical grids vertical

33:38

grids tooltips the whole thing what if

33:40

I could do that? How would you

33:43

feel about my approach? I mean, I

33:45

still would not prefer it. because the

33:47

end of the day if I want

33:50

a scatter chart and this doesn't offer

33:52

a scatter chart now I'm in the

33:54

place where I have to use two

33:56

different chart libraries I have to try

33:59

to make them look the same that's

34:01

always a disaster it's gonna be impossible

34:03

right so like I don't love that

34:06

I mean it's probably if you got

34:08

it more full-featured like that right like

34:10

it's you're probably capturing more of your

34:12

core use case group right like now

34:15

instead of like Maybe lines only is

34:17

good enough for like, I put the

34:19

number much lower than Claude and you

34:22

do, I mean, I think it's probably

34:24

like 30 or 40 percent, but if

34:26

you get bars and it does all

34:28

the magic stuff that you kind of

34:31

want, you know, then maybe it is

34:33

like 70 percent or something like that.

34:35

And so maybe that's enough. So I

34:38

think that gets you closer at least.

34:40

It sounds like it's like annoying that

34:42

because if you have a more what

34:45

I would call a more sophisticated or

34:47

real use case than like not real

34:49

is not appropriate right but like a

34:51

sophisticated use case which I think a

34:54

lot of people do when you're getting

34:56

into charting things you're already at a

34:58

certain level of app I think in

35:01

terms of yes not the like every

35:03

admin builder you've ever seen that's like

35:05

we have a little chart in the

35:07

top here like yeah rarely you see

35:10

that in a real apps but you

35:12

always see it in admin builders right

35:14

but in real apps you see the

35:17

40 sections single line yeah the little

35:19

single you know here's three pills at

35:21

the top of the chart and it's

35:24

a little thing like that's fine but

35:26

like in real apps you rarely see

35:28

that news case but in but you

35:30

do see like here's the reporting section

35:33

in the reporting section there are lots

35:35

of reports that show you all different

35:37

stuff. I'm never gonna, that's usually, they

35:40

have different types of charts in them,

35:42

like that's what's in the reporting section.

35:44

Yeah, dude, it's tricky. I hear what

35:46

you're saying, I'm having, you know, a

35:49

big part of this is the, this

35:51

is just the like, um, builder, addictive

35:53

thing, you know, you know, you like

35:56

to be in full control of your

35:58

code and what you're shipping. Yep. And

36:00

I gave that up with the text

36:03

editor because I feel... I felt like

36:05

there was a wall of like something

36:07

I'm unwilling and unable to do. Yeah,

36:09

this is very similar to that. I

36:12

feel like this is very similar. That

36:14

is a very complicated area. If you

36:16

really were to go all the way

36:19

down in there, that is very, very

36:21

complicated. Yep. And this feels like that

36:23

as well. To me, maybe that's not

36:25

true. I've never tried to build a

36:28

chart library. Anything over lines and bars

36:30

feels that way to me. That's the

36:32

line. falls off a cliff and yeah

36:35

but man yeah reporting I mean I

36:37

guess I hear what you're saying I

36:39

now I don't know bro don't you

36:41

think you're gonna have a ton of

36:44

bugs and stuff even just through lines

36:46

and bars but sophisticated lines of bars

36:48

if you go sophisticated lines and bars

36:51

yeah a lot of work in there

36:53

don't you think it would be these

36:55

things would be all over the place

36:58

there's gonna be things like I could

37:00

see people's data sets being like so

37:02

Bazaar that like, you know, I don't

37:04

know, one of them is like way

37:07

off the chart and it messes up

37:09

the whole chart and they're like, yeah,

37:11

definitely going to have that, you know,

37:14

negative values is something that's negative values,

37:16

you have to account for. Yep. Did

37:18

you like quadrants? What about quadrants? I

37:20

want like, I guess that would be

37:23

more for like a scatter. You never

37:25

have a line with quadrants. Would you

37:27

have a bar with quadrants? Probably not.

37:30

That's a good edge case thing, so

37:32

probably not. But negative, definitely, below the

37:34

X axis for sure. And that is

37:37

where it gets weird and bad. All

37:39

right, so this, this is, so, I

37:41

mean, when I think about that stuff,

37:43

yeah, it's, it's like I want to

37:46

tackle all that stuff, but it'll take

37:48

me three months. It's not a week

37:50

long thing. If you build a thing

37:53

with D3. You end up having to

37:55

do all that stuff yourself. Like, you

37:57

end up having to just like account

37:59

for that. I wonder what chart JS

38:02

does. Maybe I just forced myself to

38:04

spend a few hours using all of

38:06

these libraries and trying to break

38:08

them and seeing how bad their APIs

38:10

are. Yeah, I don't know, man. I mean,

38:13

I guess which way the other thing is like

38:15

you could look at it more

38:17

like that's where again, like my

38:19

upside, the kind of positive takeaway

38:21

I had of like what you're

38:23

building right now if you just

38:25

ship lines is the idea of like,

38:27

then you can find out. Sort of

38:30

at least to some degree like Hey,

38:32

everybody's like I need

38:34

more than this or everybody's like

38:36

this is fine I do think I

38:38

Don't know I think there's the

38:40

people you interacted right who

38:42

are going to tell you whatever this

38:44

is great, but I do worry I

38:46

guess that if I'm like looking at

38:49

Shad Sien and I'm looking at flux

38:51

and I got the charts page like

38:53

you just have no story for

38:55

that which maybe isn't a problem

38:57

right but maybe it is I don't know,

38:59

but it feels like it could be a problem.

39:02

No, absolutely. No, a big part of

39:04

it. At the end of the day,

39:06

it's tricky, man. This is the classic,

39:08

like, I've used this example a

39:10

hundred times, but when, like,

39:12

before, Bootstrap had themes, and I

39:14

would use, like, Edmund LTE and these

39:17

other themes, like, they kind of looked

39:19

like crap, they were really narrowly to

39:21

configure and deal with, but they did

39:23

everything. And then I remember using

39:25

bootstrap themes and feeling like this is

39:27

such a breath of fresh air. It's

39:29

like there's so much more that I

39:31

had to do. It was like clean and

39:33

minimal and simple and just didn't feel like

39:36

shit. But then I also remember feeling like

39:38

this doesn't do anything. It's just displays for

39:40

things. And I have to like, oh, if

39:42

I want to add that, I have to

39:44

build that. Like I don't even know how

39:46

to build that. So it's a balance of

39:48

like making something that feels. And that's

39:50

where Shadsey is doing it

39:52

perfectly well with charts, because

39:54

they're basically just giving you

39:56

configuration for recharts, which is

39:58

a good library. But I just

40:01

don't know of a library that

40:03

feels good like that. But maybe

40:05

chart chess isn't as bad as

40:07

you think it is. I don't

40:09

know if that's true or not.

40:11

I mean I've used a lot

40:13

and I have in Alpine components

40:15

I have a chart JS integration

40:17

and whatever and it often takes

40:19

me like I mean they used

40:21

it in pulse I'm pretty sure

40:23

and I remember helping Jess with

40:25

a lot of that chart JS

40:27

stuff and just like kind of

40:29

seeing how to wrangle it to

40:31

get it to work to get

40:33

it to work. It's at 50

40:35

lines, it's so much config, it's

40:37

gross. Anyway, I guess I just

40:39

think when I think about like

40:41

the big picture, yeah, why is

40:43

somebody going to use live wire?

40:45

Why, okay, there's no, part of

40:47

the big negatives was there was

40:49

no component library, right? Now there

40:51

is a first party component library

40:53

for live wire. Yeah. I guess

40:55

I feel like there is a

40:57

sort of. implied there's something implied

40:59

there to me that it's going

41:01

to be fairly feature complete yeah

41:03

at some point right and so

41:06

like so now I'm comparing should

41:08

I use react should I use

41:10

live wire should I use view

41:12

what are the different component libraries

41:14

everybody's on the component library tip

41:16

right nobody wants to build components

41:18

anymore I wanted it's table stakes

41:20

whatever my JavaScript reactive JavaScript solution

41:22

is I want there to be

41:24

components. So I don't have to

41:26

do that and I can get

41:28

right to my business stuff, right?

41:30

So like, I guess that's just

41:32

where I come from. It's like

41:34

when people look at all this

41:36

and they're making a choice on

41:38

a project, I just feel like

41:40

if flux, the more feature complete

41:42

flux is then the higher likelihood

41:44

they choose live wire and flux

41:46

over inertia and react. So this

41:48

is where this is where it's

41:50

different. This is what makes this

41:52

stuff hard. Is in react. you

41:54

have radics, base ui, react area,

41:56

recharts, you have, I mean I

41:58

literally, you could... just have the

42:00

react ecosystem which is incredible and

42:02

capable and polished every other command

42:04

palette built for them everything their

42:06

drop-downs their many everything is all

42:08

built but listen but listen well

42:11

hold on let me finish my

42:13

spiel finish your spiel out of

42:15

the gate you just have the

42:17

react ecosystem which is incredible and

42:19

capable and polished every other ecosystem

42:21

doesn't have that including view but

42:23

They have some semblance of that.

42:25

Right. The vanilla, JS, back-end rails,

42:27

Python, Jango, not put, yeah, whatever,

42:29

Jango and Phoenix and Laravel have

42:31

none of that. We have old

42:33

crusty J-query libraries. Nobody's building vanilla,

42:35

you know, that's just, you don't

42:37

have that, that horsepower horsepower. So

42:39

I rolled up my sleeves and

42:41

wrote. a drop-down menu and went

42:43

to the ends of the earth

42:45

on a safe dragover and like

42:47

every little piece like that. And

42:49

I think, because it's like, I'm

42:51

better off, that's the story of

42:53

flux, like I've done this, every

42:55

component you could have said what

42:57

you're saying right now about. But

42:59

I went the other direction, found

43:01

out that I can maintain giant

43:03

piles of complicated JavaScript, just like

43:05

anything else, you get the big

43:07

bugs up front and then, you

43:09

know, and then you have a

43:11

long tail of small stuff small

43:13

stuff that most people don't even

43:16

encounter. Yeah, I don't know. You

43:18

know what I'm saying? Yeah, I

43:20

don't know. See, so it gives,

43:22

I guess to me, it's like,

43:24

there's two angles there. One is

43:26

that like, Shad embraced all these

43:28

external libraries, right? And I get

43:30

what you're saying that they are

43:32

superior libraries. That's the point I'm

43:34

making is they're superior. Right. Because

43:36

Shad can throw together a bunch

43:38

of components based on top of

43:40

a bunch of open source. pretty

43:42

high quality to very high quality

43:44

libraries right you couldn't do that

43:46

and i i i don't think

43:48

I would have said that for

43:50

all things like I don't think

43:52

I would say you select to

43:54

because I think that would be

43:56

janky maybe you could but I

43:58

bet when I said that I

44:00

bet at that point I don't

44:02

think you would have said that

44:04

knowing the complexity involved I don't

44:06

know about for select to the

44:08

complexity on select to is steep

44:10

like very steep so select two

44:12

is like unmaintained though see that's

44:14

a difference like I don't think

44:16

that's true of like chart JS.

44:18

Do you know what I'm saying?

44:21

So that's there's places where you

44:23

can hedge a little bit like

44:25

the WhizzyWig. This is a maintained

44:27

thing, the tip tap or whatever,

44:29

right? And chart JS is a

44:31

maintained thing. I mean, it's got

44:33

65,000 GitHub stars. So it's not

44:35

like, oh, I'm pulling this library

44:37

that's like 800 people depend on.

44:39

So. It's not, like you're not

44:41

going way out there and taking

44:43

a huge risk. And you could

44:45

always work backwards, I feel like.

44:47

It's like, hey, if it doesn't

44:49

work out, there's a too clunky

44:51

ultimately, then as time allowed, you

44:53

could always swap it out or

44:55

build something in addition to it.

44:57

But I don't know. I just

44:59

think there's a couple of big

45:01

spots where like using libraries is

45:03

going to give you an advantage.

45:05

You're a one man show, you're

45:07

in there, you're hacking everything, you're

45:09

maintaining all this other JavaScript. That

45:11

is stuff where you couldn't, there

45:13

aren't good alternate solutions. Even the

45:15

date stuff, you know, maybe you

45:17

could have, maybe you couldn't have,

45:19

but I think, you know, you

45:21

want me over to like your

45:23

solution being superior there. I think

45:26

that makes sense. Because again, the

45:28

date stuff, I think there is

45:30

a lot of stuff that's not

45:32

maintained and things like that. You

45:34

don't want to use anything like

45:36

that makes sense. But I don't

45:38

think that's the case here. I

45:40

hear what you're saying. I hear

45:42

what you're saying, Ian, and what

45:44

am I going to do about

45:46

it? I don't know, but I

45:48

will say that I was pairing

45:50

with Josh. handly and he said

45:52

like he said like so I

45:54

guess this is the announcement we're

45:56

doing a flux 2.0 and we're

45:58

bundling a stuff bunch of stuff

46:00

together and I was like let's

46:02

just see if I can bang

46:04

out you know some charts here

46:06

real quick and so yesterday I

46:08

was pairing with him and he

46:10

said yeah he's like well you

46:12

know you need to like announce

46:14

flux too and and like we

46:16

should put a date on the

46:18

calendar for when to do that

46:20

and he's like just don't do

46:22

it Don't say charts until after

46:24

the end of this week. You

46:26

know, don't say anything. Don't, yeah,

46:28

don't be committed to it. Yeah,

46:31

like don't publicly say flux 2.0

46:33

will include charts until you're sort

46:35

of done with this week that

46:37

it's going to take you to

46:39

get it done. You know, yeah.

46:41

Okay, so let me just, I

46:43

got to ask you one more

46:45

thing. Yeah. If because, because when

46:47

we end this call, I am

46:49

left with actual decisions to make.

46:51

If I build a minimal charting

46:53

library that does lines, maybe bars,

46:55

will it, even, is there a

46:57

world where I build it? It's

46:59

in there. It serves a good

47:01

amount of people. And then I

47:03

can upgrade it. without like kicking

47:05

myself in the face or looking

47:07

bad or it being a bad

47:09

experience like is there something here

47:11

like should I call it flux

47:13

simple chart or something right like

47:15

that seems like the obvious solution

47:17

like basic charts simple charts fast

47:19

charts svg charts svg charts like

47:21

if these you ship is or

47:23

this which is like the basic

47:25

line charts just to get people

47:27

rolling do the little simple charts

47:29

at the top of your table

47:31

whatever fine And then you can

47:33

decide for advanced charting, what does

47:36

that look like? Is that? Yeah.

47:38

Are you just going to forever

47:40

tell people? to pull in something

47:42

different, do you want to make

47:44

it more integrated? Maybe you can't,

47:46

maybe it, maybe there isn't a

47:48

good way to make it more

47:50

integrated and it is best to

47:52

just let people do it on

47:54

their own. I mean, Younger Kill

47:56

would have been like, I got

47:58

my next, my next big like

48:00

thing. I'm going to be a

48:02

charting guy because there's a giant

48:04

hole. There's like this giant gap

48:06

between like high charts and everything

48:08

else in the video jazz world.

48:10

There's nothing simple. I mean people

48:12

do pay for these charting libraries

48:14

like it's a thing I've paid

48:16

for it for years. I think

48:18

there needs to be a simple

48:20

free open source charting library that

48:22

is limited in scope and I

48:24

bet it would take off. It's

48:26

like lines, bars, basic stuff. No,

48:28

are you telling me for serious

48:30

there's not 50 of those out

48:32

there that has to be. Not

48:34

that I'm aware of. But maybe

48:36

it this supports your point of

48:38

like. anything like that inevitably won't

48:41

be popular because you're gonna go

48:43

I need it to support you

48:45

know infinite values or whatever it's

48:47

so rough when you build a

48:49

chart is always a disaster right

48:51

like it's yeah all this data

48:53

it's of heavy queries it's complicated

48:55

queries it's all this stuff right

48:57

just on the back end now

48:59

I gotta display it and people

49:01

want to be able to click

49:03

the chart and have it go

49:05

to a different page and show

49:07

you detail views all the stuff

49:09

it's like to not be sure

49:11

that when I need a different

49:13

chart that this library can do

49:15

both charts, it feels like a

49:17

big ask. Now again, I don't

49:19

know if people are doing very

49:21

simple things and they know they're

49:23

only going to align charts fine.

49:25

But I can't really imagine any

49:27

B to B app feeling that

49:29

way. And I don't know, just

49:31

there is always just more complication

49:33

with these things. The reason they're

49:35

big in huge libraries is because

49:37

it is a very complicated problem.

49:39

And I don't think that there

49:41

is a like easy simple solution

49:43

if you want to provide the

49:46

same feature set or a similar

49:48

feature set. Of course, you could

49:50

just not have all the features

49:52

and then it's just a line

49:54

on a page. Great. Like, I

49:56

mean, you could do that in

49:58

each. You don't even need that.

50:00

You don't need anything. Like you

50:02

could make a line with some

50:04

CSS or whatever. But you know, at some

50:06

point, like you wanted to do all these

50:08

other things, people would be like, oh, inside

50:10

the tool tip. I want to be able to

50:12

put a link in there. I want to be

50:14

able to show more data inside a tool tip,

50:16

not just. Don't just give me the

50:19

data that the point is, I want

50:21

to be able to do other data

50:23

inside that tool tip with details and

50:26

whatever, right? Like there's all these things

50:28

when you get into the chart what

50:30

people want to do, depending on. But

50:32

like I, there's no world where I'm able

50:35

to offer a nice API with chart

50:37

or whatever, where you can put your

50:39

nice details inside a tooltips and not.

50:41

Oh, come on, I believe in you,

50:43

Caleb, I believe in you. No way,

50:46

no way. There's no. I pretty much

50:48

need to offer some sort of, so

50:50

what, I mean, what filament does

50:52

is, you figure out all in PHP. Yeah.

50:54

And it's just, but it's a lot, like,

50:57

it's, it's a very, I mean, yeah,

50:59

I don't want to, I like that, I

51:01

like that a lot. I like that a lot.

51:03

I want that a lot. I

51:05

want that, I just want to

51:07

be, I just, this is the

51:09

problem with a live wire, right?

51:11

I just want to stay in

51:13

PDPP land. If there's a chart

51:15

object or whatever that I have

51:17

to populate with data and then

51:19

the data points can have objects

51:21

which have details and whatever, blah,

51:23

blah, blah, blah, that sounds amazing.

51:25

I love everything about that and

51:27

then you're spitting out a bunch

51:29

of JavaScript gobbledygook that's 500 lines

51:31

long that talks chart JS or

51:33

whatever library is. That to me sounds

51:35

optimal. For the advanced use

51:37

cases, definitely overkill. For

51:40

a simple use case. Yeah, I just have

51:42

this 50 points of data. It's

51:44

a date and a value and

51:46

that's it. Nobody wants anything else.

51:48

Nobody wants to click it. Nobody

51:50

wants to do anything. Yeah, fine. Like,

51:52

that's that's simple. All right,

51:54

I'm gonna finish burning my week. If and

51:56

then at the end of the week, I

51:58

will decide where you're at. if this is

52:00

a giant mistake and I'm gonna hop

52:02

on a like a screen share with

52:04

you at at some point and see if

52:07

I can show you something that you

52:09

are like let's do it flows me

52:11

away I'm not sure that you'll be like

52:13

let's do it unless I'm sure that

52:15

you'll be like let's do it unless I'm

52:17

sure the line you're coming up it's

52:19

gonna be awesome but it's just the problem

52:22

is when I need a scatter chart

52:24

and this chart see anything didn't I Oh

52:26

yeah, there's scatter chart, but it's, you

52:28

know. Oh, scatter chart. I'll make you

52:30

a scatter chart, are you kidding me? Yeah.

52:32

Don't do it. Not on a grid?

52:34

I know, I know, the scatter chart's one

52:36

of the easier ones. Let's go, let's

52:38

do a scatter chart. Not really different than

52:41

a line chart, really. It's sort of

52:43

the same thing, but without the lines. Let

52:45

me build you. Like a Cartesian if

52:47

it's a Cartesian grid thing I'm in no

52:49

radar charts. No None of that soon

52:51

as we get polar. I'm out Until

52:53

I crack open that black box realize it's

52:56

humanly possible to learn it and then

52:58

want to do it No heat charts you

53:00

give me the heat charts. No way

53:02

can't do it won't do it. Is that

53:04

thing called re chart? Do they have

53:06

any of this stuff? No, they do have

53:09

the treat which is like a heat

53:11

chart. Yeah, they do have it They

53:13

have the tree map. I don't know what's

53:15

involved. Style it that way though, but

53:17

high chart. You must be able to. H.R.B.

53:19

dude. Oh yeah, they do. They do

53:21

have it. I mean, I mean. The thing

53:23

with all these charges, I don't think

53:25

any of these things are too hard for

53:28

you, right? Like I think you could

53:30

make a pie chart. Like I feel

53:32

very confident. You can do it. No, I

53:34

believe it. But the problem is like

53:36

the 8,000 million edge cases of like and

53:38

all the little sub things when people

53:40

want to show a certain way. That's where

53:43

all the trouble is. Not in like

53:45

can you draw a circle and divide it.

53:47

Like I feel like. Yeah, you could

53:49

definitely draw a circle and divide it

53:51

up into the parts and put a number

53:53

in it. Like, not gonna be a

53:55

problem. But when you wanted to magically fly

53:58

together with a nice animation and you

54:00

wanna do, you know, all the stuff, you

54:02

want the chart segments to be clickable

54:04

and what happens when they're clicked and blah

54:06

blah blah blah blah. All that stuff.

54:08

Don't do it all. And I think

54:10

you can do it all. But I don't

54:12

know if it's the best use of

54:14

your time to do it. All right. This

54:17

is good. You've at least you've at

54:19

least like shaken me and I stated my

54:21

game up for what truly what you've

54:23

done is woken me up from my just

54:25

like, you know, Ivy drip of like

54:27

problem solving. We're selling software here Caleb.

54:29

We're selling software. That's what we're trying to

54:32

do. Sell software. Honestly, it is literally

54:34

if you were like talking to somebody about

54:36

the mechanics of like a business thing

54:38

like this and they didn't. Most people wouldn't

54:40

understand that like you're selling software. But

54:42

you could just be selling someone else's software

54:45

that they gave to you for free

54:47

and are letting you sell. They're letting

54:49

you sell it. You're literally like you can

54:51

just have millions of hours of work.

54:53

You can literally have hundreds of thousands of

54:55

hours right here. Like and you can

54:57

just capitalize on it. It's so much. That's

54:59

sort of interesting thing too about the

55:02

idea of like the P. P. abstraction layer.

55:04

It's like you can then even most

55:06

likely fix some of the rough edges. the

55:08

final config right like now you're massaging

55:10

things to do what you want to

55:12

do potentially and things like that as well

55:14

so that's kind of interesting but yeah

55:16

yes you can get this hundreds of thousand

55:19

hours of labor and eight thousand bugs

55:21

they've already fixed and all that stuff it's

55:23

totally free you can ship it but

55:25

just you're what your value is you're giving

55:27

me a nice API I don't have

55:29

to think about all that stuff too

55:31

much unless I really need to and then

55:34

I can dig in there and that's

55:36

fine but I'm getting the kaleidified API Let

55:38

me make this simple line chart thing.

55:40

Everybody's gonna love it. nobody's going

55:42

to want anything

55:44

else. to be else. to want

55:47

anything else. to want

55:49

they're going to carry

55:51

you around at going to

55:53

carry you around at their

55:55

party libraries and see

55:57

if any of them

55:59

aren't the worst thing

56:01

in the world.

56:03

But they're probably all

56:06

our thing in the world, but

56:08

don't want to use.

56:10

all our except charters, which I don't

56:12

thanks. want you, brother. Okay,

56:14

Ian. Thanks. right, to you.

56:16

Appreciate you. Say it.