0:05

Hey Hey everyone, welcome to another

0:07

episode of Ruby of I'm David

0:09

I'm and today on our panel

0:11

we have on our panel we have John Epperson.

0:13

Hello? And Luke Sutter's? today we are

0:15

talking with a special guest, Yvonne Jeanco.

0:17

Hi guys, Hi for thanks for inviting

0:20

Yeah, so do you so do you

0:22

mind telling us a bit about

0:24

who you are, some of

0:26

the things you're working for and

0:28

all that good stuff? that good

0:30

course. Of These days these days I'm freelancer,

0:32

also consulting from time to time, But

0:35

I started to use to use

0:37

Ruby and Rails in 2006,

0:39

a since then I was

0:41

doing a lot of

0:43

different stuff. I was two agencies,

0:46

founder of two agencies, a

0:48

organizing all the different a lot,

0:50

and all the different stuff.

0:52

up to lately? have you been

0:55

up to lately? and some

0:57

consulting, some consulting, also a

0:59

decent amount of teaching, of

1:01

course. And we had you

1:03

come on to talk about

1:05

one of the talks that

1:07

you gave at that you gave at

1:09

Abstract, called less abstract Rules effects

1:11

of expressing rules in

1:13

or of object oriented

1:16

visual way. Yeah, in a visual way. In

1:18

In pictures, right. Would you

1:20

mind giving us a highlight us

1:23

of that talk? Sure. talk?

1:25

The whole thing started

1:27

as a conversation. I

1:29

I guess a year ago it's another

1:31

conference in in Petersburg and

1:33

me and some other people

1:36

who are also involved in teaching

1:38

in teaching rubies. We had we had a discussion

1:40

that sometimes it's very

1:42

hard to explain. some some

1:44

abstract concepts to the

1:46

newcomers. And there was was one

1:48

rhetorical question, which was said,

1:50

It sounded like, like, Wouldn't be

1:53

great if we could

1:55

illustrate, let's say, dependency injection

1:57

with just one picture picture

1:59

so that everyone can... understand it

2:01

and understand it in the correct

2:03

way. the correct way. So yeah, that

2:05

was just a rhetorical question

2:08

as I said, I

2:10

but later later I was

2:12

teaching another pack of students.

2:14

And I was drawing drawing

2:17

picture to them and those pictures

2:19

were like pictures of

2:21

little dudes dudes those those

2:23

picture were meant to be

2:25

classes. So I was was playing in

2:27

OOP them. them. And then I

2:30

this this question I thought I

2:32

thought, what could be like

2:34

the next steps to illustrate

2:36

more stuff of the OOP in a way?

2:39

And I started drawing I started

2:41

to those dudes and I

2:43

thought, all right. and

2:45

I thought, all right, let's methods

2:47

in our code, and

2:49

then our code. more and

2:52

more and more and more and

2:54

more and to develop to develop

2:56

of of language. to

2:59

express what's going on in

3:01

your code. code. So yeah, something like this.

3:03

I like this. I mean,

3:05

your pictures are awesome. you. Did

3:07

you you. Did you come to a

3:10

conclusion, you did you measure that this

3:12

works? So in my talks, I my talks,

3:14

I was just excited that it

3:16

all makes sense. of the

3:18

of the fact that it's still all

3:20

all weird. I was And I was

3:22

excited about it and wanted

3:24

to present this this to basically to show

3:27

that. But I I think it was helpful.

3:29

for me in several ways. First

3:32

of all, I was able to

3:34

explain. some concepts of

3:36

my students. to my Secondly,

3:38

I was using it

3:40

in my my freelancing as

3:43

well. as well. So found

3:45

it useful when you have some.

3:47

you have some complex case

3:49

where you where you don't

3:51

know how to model it correctly or

3:53

you you just have some. complicated

3:55

set of classes when you're

3:57

solving some specific problems.

4:00

What I did I did I, let's

4:02

say I had five or six or six

4:04

glasses I I didn't like

4:06

the design. design. I just. I just decided

4:08

to draw all of them. all of using

4:10

this visual language. And

4:13

from that, I started

4:15

gaining from that I started method

4:17

is duplicated in several

4:19

places in several so on

4:21

and so on and

4:23

so on and so on. it was

4:25

useful was in my my

4:27

well. as well. I'm not

4:29

sure how to share it

4:31

with more people and how

4:33

to make it more useful, but

4:36

that I had some attempts in that

4:38

as well. I thought it

4:40

was a really interesting idea because

4:42

a lot of people can do

4:45

it, you know, just the of

4:47

passing the arguments, but I arguments, but they

4:49

don't really have an understanding of

4:51

it. And I think getting that

4:53

visual model in your head really

4:55

goes a long way. way. Did

4:57

you find that people reacted well

4:59

to it? well to it? people usually

5:01

like it a lot because

5:03

it's a very surprising way of

5:05

thinking about the code for

5:07

them. the code for them. So yeah,

5:10

people are really excited, but

5:12

I don't think many

5:14

of them find it very

5:16

practical so far. I I

5:18

it's it's of the drawing of

5:20

the rails the mountain of learning

5:22

rails. of learning rails. Yeah. This is,

5:24

yeah, I wish you should tell about

5:26

it, yeah, this is a interesting

5:28

thing. thing. What is

5:30

at the top? What

5:32

is at the top of the of

5:34

the rails So I would say it's

5:36

not a say it's not a rails mountain,

5:38

it's a cycle you know, not

5:41

a cycle, but well, let's

5:43

say say life every

5:45

rails developer. So it's also

5:47

model I developed for

5:50

another talk in Ukraine. talk

5:52

in Ukraine. I I simply spotted

5:54

that many many of developers

5:56

go through the same

5:58

steps. steps and I... I was

6:00

on many of them as well.

6:02

So the first step first love you love

6:04

rails, when you believe in rails in

6:06

rails way, and you in love, fell

6:08

in love. on you Then

6:10

later on, you some getting some

6:13

problems and you fixed them with

6:15

some with some magic. And after several

6:18

attempts to fix stuff with

6:20

might you might get disappointed

6:22

in rails way. And then then you

6:24

switch to another step

6:26

where you start learning

6:29

about all the alternative

6:31

approaches approaches like solid, dryer, be, I

6:33

don't know, trailblazer, know, service

6:35

objects and so

6:37

on, so on, of course, patterns.

6:39

Then you can of course You

6:41

can, of course, find where you

6:43

feel where you feel comfortable.

6:46

because like your projects over

6:48

a specific size or like

6:50

you just found just

6:52

found your way

6:54

of managing complexity

6:56

in web applications. But many people

6:58

they decided that, all right, right,

7:01

something is wrong with the something

7:03

might be wrong with

7:05

the way we develop

7:07

applications. they started to leave, and they

7:09

to leave. that they should try

7:11

that they should try other

7:13

approaches. And this is other

7:16

approaches. four, like the this is the

7:18

top of the mountain, is the final, the

7:20

top of the mountain is when you

7:23

in to learn from all those

7:25

approaches in different ecosystems. And

7:27

you, well, on the stop level, it's level,

7:29

it actually doesn't matter for

7:31

what technology you use. you

7:34

use, you can return to Ruby and

7:36

you can return to Rails with all this

7:38

knowledge. rails with you've broken the

7:40

matrix. and you've broken the to

7:42

do stuff. now So

7:44

that's the final step. that's

7:46

the for me getting

7:48

up to that me broken

7:50

that matrix's broken place. It originally

7:52

came from me learning rails

7:54

I learned Ruby. Ruby. That That was

7:57

probably one of the biggest mistakes that I

7:59

made. that just I just dove. right into rails without

8:01

understanding how Ruby classes

8:03

how to create true plain

8:05

old Ruby objects. And so

8:08

I would have fat

8:10

fat controllers with super

8:12

fat models, things were

8:14

not maintainable, they were

8:17

not manageable, I had

8:19

zero unit tests on

8:21

my old applications applications and I

8:23

thought that, man, Rails isn't really

8:26

all that intuitive. It's not

8:28

all that great. So it really

8:30

really was, I'm maybe I'm just

8:32

a bad programmer. which is kind

8:34

is in of rails disciple way. So for

8:36

So it me, it was getting

8:38

up to that peak of

8:40

the mountain by actually

8:42

learning Ruby. up on some up

8:45

on some Ruby techniques

8:47

like or just just creating

8:49

Ruby objects instead of

8:51

just relying what Rails

8:53

provides. So So a better

8:56

Ruby developer made me understand

8:58

a lot more and become a

9:00

better Rails developer essentially. a better Do

9:02

you think essentially. Do

9:04

you think that faster

9:06

with Ruby Ruby and

9:08

it's so good. good

9:11

at accelerating you through the early

9:13

stuff. For me, I would say, was

9:15

a I was a pretty horrible

9:17

developer for a long time, Mostly

9:19

because I was because I was just

9:21

trying to do things by

9:23

myself. I wasn't including other people

9:26

in my work and that

9:28

inherently limited me on what I

9:30

was able to do. seeing a

9:32

pattern here that you I feel like I'm

9:34

seeing a pattern here, that you are

9:36

taking these abstract concepts and putting

9:38

them into pictures. part, like, the most part,

9:40

getting good seem to be getting good

9:42

reception of it, like connecting connecting with

9:45

this kind of stuff. feel like I feel

9:47

like it's kind of analogous to, you know that.

9:49

that one dude that you know you know that like

9:51

makes good analogies all the time all the

9:53

of thing. of I feel like it's

9:55

kind of like that, kind of like that kind of

9:57

a person tends to be able to be able to.

10:00

maybe it's a skill of a teacher. I

10:02

don't know exactly where I'm going with this,

10:04

but I feel like there's something here, which

10:06

is I think what you're coming to tell

10:08

us. like there's maybe I don't have a system. I

10:10

I haven't measured it yet, coming to tell the

10:12

beginnings of one. hey, Is that where

10:14

you feel like you are in the process I do you

10:16

feel like you're further measured it yet, I

10:18

do a lot of

10:20

teaching one. Is that where you feel lot

10:23

of my relatives are teachers,

10:25

or so I feel like you're further along?

10:27

something this. And yeah, these things I'm things

10:29

I'm talking about here and

10:31

there, they are kind of

10:33

a part of a bigger

10:36

system that I have in

10:38

my mind. in And also, I've

10:41

tried to write a

10:43

to write a book. call Painless

10:45

Rails, but yeah, it's yeah,

10:47

it's already two years after

10:49

I started writing it. right

10:51

now postponed state, but state.

10:55

But like the bigger, the

10:57

goal for me is to

10:59

make development using

11:02

joyful again. Because I'm

11:04

a freelancer I I'm a

11:06

freelancer, I I usually... switch

11:09

from from project to

11:11

project, and I see

11:13

that a lot of

11:15

people, they inventing their own own

11:17

bicycles. I mean mean, remember

11:19

those times when we

11:21

had we had like single rails way

11:23

when we all on the

11:25

step. one of this mountain. We all

11:28

were enjoying the enjoying

11:30

the process. And then

11:32

we started more approaches

11:34

approaches all grew our all

11:36

grew, our projects and bigger and

11:39

more complex. We need it. other

11:41

some other methods to manage

11:43

complexity, we but we ended

11:45

up having dozens of of

11:47

different approaches there there is

11:49

no single unified way of

11:51

doing that that now. I I

11:53

think ideally we should

11:55

have one unified

11:58

approach again. would

12:00

fit to 90% of all of all

12:02

projects. Of Of course,

12:04

like the domain models lot of

12:06

specifics is different, is

12:08

but I'm still pretty

12:10

sure that for 90 %

12:12

of 90% of 90% of for

12:14

90 % of functionality, of

12:17

90 % of all

12:19

Rails projects web web projects

12:21

on Ruby. Ruby. we can

12:23

develop a single unified way.

12:25

imagine if we could

12:27

measure if we could measure a different

12:29

approaches. like a different approaches,

12:32

like with trailblazer, we

12:34

ratio of, I don't

12:36

know, of, I don't know, five, and some

12:39

another metric of, I

12:41

don't know, 10. know, ten.

12:43

for another, like, when you'll rail, we

12:45

have like those numbers

12:47

of seven and six.

12:49

six. And we could just

12:51

compare those approaches. approaches. And

12:53

let's say, and this, approach is

12:55

optimal. is optimal. And it

12:58

it sounds pretty, like, fantastic,

13:00

I would say, I still I

13:02

would say. it's I still think it's

13:04

doable. Man, I can explain it.

13:06

take some time. gonna take some time. So

13:08

from my perspective, I think that right right

13:10

there, that motivation to sort of like of

13:13

like get. the I don't I don't know that there's,

13:15

at this I look at this as sort

13:17

of being a thing that thing that there's a

13:19

perfect thing out there all we're all trying

13:21

to go get it, right? all We're all

13:23

trying to seek it. And I feel

13:25

like that is sort of like the motivation

13:27

that drives drives like half of the things that

13:29

you see out in the world. the You

13:32

know, somebody's upset about rails.

13:34

know, now we have have what Sinatra and

13:36

army I mean I mean, they're even

13:38

on your mountain, right? Like

13:40

they're so popular. popular like they're exactly competitive

13:42

much, but. but but it's competing

13:44

against pieces of it, right? like we have these

13:46

we have these things where somebody's like,

13:49

I'm going to try and seek that

13:51

one thing, right? Or, right or you

13:53

know, in in world, we have this, we

13:55

have like we have we have have you somebody's

13:57

gonna come out with another. with another you

13:59

know, framework. I mean, angular is is

14:01

kind of like sort of

14:03

on the and embers basically kind of

14:05

gone. know, we're You know, we're not even

14:07

talking about the the generation before that, that

14:09

are Like that are just gone now.

14:11

Like people are always like to recreate this

14:13

bicycle to achieve that one place. that

14:15

I don't know that we I don't know that

14:17

that stop that cycle. I I

14:20

think there's always going to be

14:22

more churn in the front the

14:24

back the That end. in my wasn't

14:26

my conclusion, but you're welcome to

14:28

make that conclusion. conclusion. So let me

14:31

explain what I mean. I mean, and

14:33

And why I think it's at

14:35

least at least theoretically. So

14:37

first of all, this all,

14:39

this approach. in pictures

14:42

of in pictures of

14:44

little dudes. our code

14:46

we ways to analyze our code

14:49

coverage, have right now? So

14:51

we have have coverage, for example,

14:53

we have some complexity with

14:55

Flock or some other stuff with

14:57

Ruby Critic, but those are

14:59

just numbers. just numbers. say

15:02

you have a class of

15:04

500 lines of code. code, and

15:06

right, you have a metric that

15:08

your class has your of something

15:11

like, I don't know, of something like,

15:13

I don't know, 15. And doesn't give you

15:15

like a lot of information a

15:17

can work with. And

15:19

you can of And

15:21

I think of the tools give us...

15:24

very low level hints on

15:26

code quality. And we

15:28

tend to think that tend to

15:30

think that Rubikov about code quality.

15:32

quality think those are just are

15:34

just about code style they can't show

15:36

you that something is wrong.

15:38

something is wrong but they give

15:40

you a high high-level hints

15:43

and now let's say let's say

15:45

we take the same class

15:47

of 500 lines. to express

15:49

it to express it with a picture

15:51

of little dudes we can have like like Here's

15:53

our guy and this guy, all

15:56

all the are methods, let's

15:59

say we have. seven methods so

16:01

our guy has seven arms

16:03

and we arms this we see

16:05

like this. my talk I was showing like the

16:07

my talk, I was showing too long then

16:09

if the method is too long,

16:12

then the arm is kind of

16:14

long and wavy. you And

16:16

from that picture, you already see that

16:18

it's not very have good to have such

16:20

an arm. It's not very convenient. and

16:23

for for another method,

16:25

you have have seven have seven

16:27

arguments there. there and arguments are

16:29

seven fingers. are seven fingers

16:31

fingers is not with

16:33

seven fingers is not on,

16:35

and so on, and so on.

16:37

and now you have this and now

16:39

you have the properties of

16:42

the the properties one little

16:44

compact picture, in and your

16:46

brain is picture and your brain

16:48

is actually works with

16:50

visual objects much

16:53

better when with text.

16:55

and it takes you less

16:57

time to understand of

17:00

this picture. And actually there

17:02

is a thing

17:04

called channel faces and it's used, you can

17:06

you can Google for it, and this

17:08

is a technique when when

17:10

people faces with

17:12

different with different expressions

17:14

to show multidimensional data. For example,

17:16

you have a table with

17:18

a lot of columns. with a and

17:21

you have to and you have to, like

17:23

I don't know, know, it's something.

17:25

like showing some some death

17:27

rates some information about

17:29

health in different

17:31

states. And for example,

17:34

death rate is expressed with a

17:36

size of a size of eyes and

17:38

don't know, something else is

17:40

expressed with with the and

17:42

so on and so on. So, so

17:44

and this so on. So and this of

17:47

our brain, of our brain which

17:49

faces. faces. And with

17:51

this, you can get

17:53

very efficient. efficient. found this I

17:56

close close the external faces yeah, And

17:58

what I found. I found... that

18:00

if you use this in in

18:02

Dynamics. so So let's

18:04

say we this picture

18:06

of of 500 lines class and then

18:08

we get a pull we get

18:11

a pull request, which changes this class and

18:13

maybe some others. some we

18:15

can have another

18:17

picture, another picture. is this arm

18:19

is gone, this arm got

18:21

fingers. and this

18:24

arm becomes longer. so

18:26

on and so on and so on. So.

18:28

on. Now you can like, very

18:32

you can

18:34

see differences, what high

18:36

level differences were

18:38

added to your added to

18:40

your system. is... So that would

18:43

be kind of would be

18:45

of of part of

18:47

some kind of continuous

18:49

integration system. Yes, Yes,

18:51

correct. Actually, so two links

18:53

I've sent to you, to

18:56

the the one which... called dudes hub. This

18:58

is a This is a

19:00

prototype which is developed

19:02

by my student. does that,

19:04

which already does that,

19:06

actually. it to your repository and it

19:08

to your posting you and it will

19:10

start little you pictures of those little

19:13

dudes whenever you create Does it work on

19:15

it work on like as It looks

19:17

like it has a on a cat

19:19

cat file. it should work. it should work. I

19:21

didn't have the I didn't have the chance to play

19:24

with it a lot, and I I said, it's

19:26

a prototype, but it should

19:28

work. work. Okay, awesome. See here,

19:30

see here. it Not sure, an

19:32

it looks like an

19:34

app. link right, in link it in

19:36

the stuff or whatever to play

19:39

with. Oh to play with it. Oh, yeah. So the, the,

19:41

is the last one called one called

19:43

DUDT and DUDSP is, yeah, it's

19:45

an app. Basically, it

19:47

should be deployed first

19:49

deployed first to, uh, yeah, but I but

19:51

I think I should just do that

19:53

so that you could be able to

19:55

plug it to it to your Apple. So they

19:57

together or something. or something?

20:00

Unity is a gem that

20:02

you can plug to

20:04

your repository repository I believe it's

20:06

a common line, you

20:08

can say, utility and you can

20:10

say, these classes

20:12

as those. as

20:14

of little pictures you will

20:16

get that. dudes and you will

20:19

get that. Gotcha. the, is

20:21

the. Yeah, it's a library.

20:23

Yeah, it's the core it's

20:25

the core of the whole thing. actually

20:28

draws all those. this can This

20:31

can potentially change

20:33

a way of thinking

20:35

about and of complexity. it

20:38

requires a course, it requires

20:40

a lot of more research and

20:42

I don't know, marketing, I like it

20:45

so it I like it. like what

20:47

it sounds like what you're arguing. here please

20:49

correct me me if I'm wrong. Sounds

20:52

like what you're kind of making an

20:54

argument for of that an you for

20:56

is that if you express

20:59

your code quality in this this

21:01

visual kind of language.

21:04

basically can. two things. One,

21:06

I think two things on

21:08

noise, two, cutting down on

21:10

noise bandwidth because we're using more you're

21:12

increasing bandwidth to we're using more parts of

21:14

our brain to help kind of handle

21:16

this or whatever. you're you're communicating with more

21:18

bandwidth. more Correct, yes. I mean,

21:21

I the amount of information

21:23

you can you can. in

21:25

one place is also important.

21:27

Like, remember Like our brain can

21:29

handle only can handle five to

21:31

seven things in one things in

21:34

one like bite. Like here, of the

21:36

compactness, can you can decent amount

21:38

of decent it in one

21:40

place, and on one picture. in

21:42

one and I think on

21:44

one picture. Yeah, and I think that where

21:46

I like glance the at glance of

21:49

where things where things are at. So I

21:51

guess one of these little little people

21:53

or or as it looks looks like

21:56

they're called, If if one of

21:58

them has 10 different r that

22:01

would essentially be like

22:03

10 different public methods

22:05

that that class has.

22:07

which in a of cases

22:09

if you're creating a poro would

22:12

not be ideal. So right

22:14

off the bat, you can say,

22:16

wow this class has some

22:18

serious complexity because you have 10

22:20

different entry points to be

22:22

able to interact with this class.

22:24

Maybe we should simplify it

22:26

so it only has one arm

22:28

at most two and then

22:30

create other smaller classes to handle

22:32

whatever the other methods, public

22:34

methods were that this class was

22:36

originally housing. Correct Yes.

22:39

That's a good example.

22:41

I did similar thing in

22:43

my talk in at

22:45

RubyConf. I just took another

22:47

talk of Sandy Matz

22:49

where she she was doing exactly that.

22:51

She took one huge class and

22:53

step by step, she split

22:56

it into. something like

22:58

15 smaller classes. and I

23:00

basically illustrated the whole

23:02

thing. So, and yeah you

23:04

can see it in

23:06

the talk itself. I mean.

23:08

kudos to Sandy Metz for like kind

23:11

of basically doing this for

23:13

so long. like and giving

23:15

good talks right on it. But

23:17

yeah I I definitely am coming

23:19

around to the thing here that

23:21

having a visual would make communicating

23:23

about that stuff a lot faster. because

23:25

I think one of the things that I've definitely

23:27

encountered. is that while

23:30

I have spent a lot of

23:32

time. looking at talks like Katrina Owen

23:34

is also really good at this, right? But

23:36

like watching talks And

23:39

reading about stuff and I feel like

23:41

I'm pretty good at identifying these things,

23:43

you know And I'll comment on them

23:45

in and things like this. I feel

23:47

like it's a very hard thing even

23:50

for programmers who you know we were

23:52

like, ah, programmers are smarter than the average bear

23:54

and therefore they should be able to. to

23:56

get this stuff, I think

23:58

it's a hard concept. for everyone. Maybe

24:01

visual language would help overcome some of

24:03

those barriers. I know for me

24:05

it definitely would. And

24:07

just take the example, this kind

24:09

of like off topic, but

24:11

related, but like a Kanban board,

24:14

a Kanban board usually has

24:16

three columns To do, doing, and

24:18

done. And that's great for single

24:21

-flow tasks, tasks that don't

24:23

require anything other than a

24:25

to -do, doing, or done. You've

24:27

the most important column, which is

24:30

the won't -do column. Yeah,

24:32

Yeah. yeah, if

24:34

But anyways, my

24:36

problem with that kind of

24:39

style. is that you don't have a

24:41

great visualization. Okay. so this story

24:43

that we have now developed this

24:45

feature on, it's done. What is

24:47

done even mean? Does it mean

24:49

that it has gone through the

24:51

proper code reviews that it has

24:54

past a QA checkpoint that

24:56

it's been deployed, know, what

24:59

does that actually mean? So

25:01

in a CIECD kind of

25:03

world where, or a more

25:05

proper development life cycle world,

25:07

you have many more steps

25:09

than just to do, doing

25:11

and done. And so I

25:13

like my boards actually more

25:15

visual where you do have

25:17

several different columns, each column

25:20

representing the next step for

25:22

that story. So you're able

25:24

to that. see at a glance okay

25:26

all of my work is done

25:28

there's no more expectations for me the

25:30

sprint why don't that pull something

25:32

else in or learn something or do

25:34

something else so that way there

25:36

is no confusion of oh the story

25:39

is done but it's still actually

25:41

waiting for a review and waiting to

25:43

get deployed to our staging environment Something

25:46

like that. So I see the

25:48

visual aspect that we're talking

25:50

about here with the little dudes

25:52

being very helpful in not

25:55

only as a, for new developers,

25:57

but for seasoned developers to

25:59

see At a glance, what is

26:01

the current situation? Where are

26:03

things currently at? How healthy is

26:05

our actual code? actual What's the status

26:07

of our actual story? stories?

26:10

Just that at a simple

26:12

glance view, having knowledgeable

26:15

actionable information. Yeah,

26:17

I think it can be also

26:19

useful if you want pass the

26:21

Pass the knowledge about some

26:23

system you're developing. developing and like, especially

26:25

some part of some of some of

26:28

the system, of the system. you

26:30

can draw it in

26:32

this way. it in this way and

26:34

first of all, you you

26:36

this picture and then you

26:38

explain what it means like, what

26:40

it means. It's this way you

26:42

can get this more way to

26:44

a more things. way to explain

26:47

I guess you can hope that can

26:49

hope that people understand it

26:51

more that's more hope that I work

26:53

this way. work this way. I

26:56

I definitely am really into this image thing. thing. I

26:58

think I also agree with you, Dave, right? you, Dave,

27:00

right? I agree. I I have a small

27:02

critique, I but I must agree the the board

27:04

thing. board thing. would say because my my

27:06

board, I think everyone has a combo

27:08

on board that board make a trade off

27:10

decision between having enough columns to actually

27:12

capture all the states that you care

27:15

about, all know, because my combo on

27:17

board is like You columns my combum board is know,

27:19

seven columns big on. I'm

27:21

focused on. Well, did you get get the

27:23

seven? All All right, so I have the like

27:25

the needs column, right? right the actual backlog.

27:27

know then you have in then you have

27:29

in progress, in I have an which column,

27:31

which means that, else is somebody else is

27:33

supposed to thing at this thing and do,

27:35

you know. and then do something ready to there's

27:38

a ready to deploy column. a completed

27:40

and then of then of course box right because

27:42

you you can't forget the ice box or

27:44

seven. Okay well well that visually, I

27:46

have a large screen here and that

27:48

visually takes up the whole screen. If

27:50

I'm on my laptop screen, screen. I think

27:52

the the column is hidden, you know. hidden,

27:54

you know, that's not good. It's a wide board. ban board.

27:57

Right, but anyway, so I'm not trying

27:59

to get to get too... down in nuts, my

28:01

point is is is visual information

28:03

being. being from this.

28:05

this, but but I think it's a

28:07

stretch. prop, like, this is a This is a

28:09

problem where the illustration is stretched

28:11

too far. far, right? And I think think

28:13

that what is making me really

28:15

comfortable about this code thing so far,

28:18

I least so far right, can't think

28:20

of a way right in which

28:22

this I think Yvonne, I think if

28:24

on like you've captured most of the problems I

28:26

can think of so we're just talking

28:28

about about. a a high bandwidth communication

28:30

solution for something that's a problem

28:33

right now. It's just an easy

28:35

sell, I think. right I

28:37

mean the potential is

28:39

enormous. the potential is and

28:41

I really liked I really in

28:43

your talk bit you talked about

28:45

the exceptions. talked and how

28:48

you visualize exceptions.

28:50

visualize exceptions as a Yeah,

28:52

basically things go wrong

28:54

go wrong and you should should understand

28:56

me, though. me Most of the

28:58

things the things I... haven't even tried

29:00

tried to, it was was not. like

29:03

they really really drive you

29:05

when you you get basically

29:07

of your universe, like you

29:09

basically think, all right, and how this thing

29:12

will look like. thing will look

29:14

like. And it's, it's,

29:16

it's obvious now, now, like,

29:18

something wrong, the should want me. That

29:20

was really evocative. And

29:22

what immediately occurs

29:25

to me occurs to some

29:27

kind of monitoring

29:29

system combined the concept of

29:31

the... regurgitating exceptions

29:33

with the can

29:36

assign the regurgitations

29:39

can assign the you can see to

29:41

the dudes so you can see

29:43

which classes are being physically

29:45

sick. I did just look through

29:47

the repo. There's nothing

29:49

about failing the guy in the guy throwing

29:52

up. not not implemented

29:54

yet. Well my my weekend

29:56

gone. In some ways, some

29:58

ways this reminds me of a gym

30:00

that I played around with a

30:02

while ago, several years ago, ago, rails

30:04

ERD, but But this has a

30:07

much better at a glance of

30:09

communicating. data

30:11

in state of of your Rails application. So I

30:13

I think it's really cool. I

30:15

may check this out a bit further.

30:17

a bit I do have some

30:19

problems problems of of the things

30:21

that struck me as odd

30:23

me the methods with too

30:25

many lines of code in

30:27

them were depicted as as arms. So

30:29

the So the methods of

30:31

the arms are ready. And

30:33

when they have too they have too

30:36

in them, they become in them, they

30:38

become not exactly

30:40

like this. exactly

30:43

like is method is

30:45

then then you have a

30:47

long arm. arm and your arm

30:49

swollen if there is

30:51

too many conditional logic

30:53

in it. in it. So it means...

30:56

like you get it, right?

30:58

right a complexity score. It's

31:00

kind of, it's kind

31:03

of so because here

31:05

it's used here it's used

31:08

to to the nature

31:10

of complexity of is

31:12

what we this is

31:14

what what I came to so

31:16

I think. with this

31:18

this approach you can show you can

31:21

show that there are

31:23

two types of complexity. One

31:26

One is internal, where you

31:28

where you have arms, arms,

31:30

long arms methods, and many

31:32

methods and too many

31:34

internal methods. This

31:37

is all internal complexity

31:39

And external complexity is

31:41

where where you have small classes,

31:43

but you have small classes, but you have

31:45

tons of them. the point I was

31:47

making in the point I was

31:49

making in the end

31:51

of my talk, after I

31:53

demands talk, that is that complexity

31:55

actually didn't go away. away. It just

31:58

form. the form. So...

32:00

And so it was all it was

32:02

all internal complexity. It was all in

32:04

one single class. And now we

32:06

have 15. we have 15 classes. the

32:08

complexity is now external, is

32:10

so we managed we in

32:13

this way. but sometimes way.

32:15

But you can have a lot

32:17

of lot of but it's still

32:19

hard to deal with those classes. with those

32:22

it means So it gone

32:24

too far with far your OPE

32:26

or like basically So that's

32:28

the idea. idea. And I think that

32:30

think are saying when we are saying

32:33

that we in like the good code, the whole

32:35

question is the whole question

32:37

is about the balance between

32:39

internal complexity and external

32:41

complexity. So sometimes it's okay to

32:43

have to conditions. Sometimes you have

32:45

to have to into into

32:47

polymorphic classes. but that's

32:49

that's up to you. decide. Yeah,

32:51

I think I think that's That's kind of

32:53

what what I was. Hinting at a little

32:55

bit earlier, earlier, right? Like, I think

32:57

it's very difficult for people to actually

32:59

recognize. to actually and know where

33:02

they've put their complexity, right? right?

33:04

are really quick to recognize, you know.

33:07

you fat controllers and fat models,

33:09

but like but the hood, the

33:11

reason why fat controllers or

33:13

fat models controllers us is because us

33:15

is because of, you know, Okay, I have a

33:17

ton of conditionals in all of

33:19

my methods, all or I have a

33:21

million lines in every method, right? I'm

33:23

basically writing every code inside my OOP

33:26

writing imperative code and maybe

33:28

there's a maybe there's a there's

33:30

a different way to put that, right? Like maybe this

33:32

method is super hard to test or something, right? Like

33:34

there's a lot of reasons why you a want to

33:36

move it out. way to feel like. right? We're

33:38

like, look, if we can write easy

33:40

is super then. to test, or is pretty good code

33:42

and I can maintain it or whatever and

33:44

that's sort of like why way to measure. measure, that

33:46

our complexity isn't out of sorts. But yeah,

33:48

But I think the reason why this tool

33:50

is awesome. The way that I'm way that

33:52

I'm understanding from you, it's awesome awesome because

33:55

it makes it really quick to be

33:57

able to communicate to somebody else. else. This

33:59

is where are all the, this is how

34:01

all the pieces lie, right? Like, Like, okay,

34:03

there's a bunch of skinny long arms or

34:05

there's a bunch of fat long arms or

34:07

there's just fat arms or whatever. Like I

34:10

suddenly am giving you a ton of information

34:12

about where you chose to put your complexity. Exactly

34:15

right. And when I'm

34:17

talking about the optimal way

34:19

of developing using some approach,

34:21

like let's say Rails, I'm

34:24

actually talking about this. So

34:26

we can quantify the way

34:28

we work, like let's say

34:30

if we, like every

34:32

requirement we get from the customer.

34:35

we We can quantify in

34:37

a way like, here should

34:40

add a new entity,

34:42

like here we We

34:44

have a fork in business logic.

34:46

Here we have something else and

34:48

something else and something else. And

34:51

let's say we've done that. And

34:53

also, we can with this

34:56

visual approach, we can

34:58

quantify different approaches like

35:00

trailblazer approach, like dryer

35:02

bee approach, Hanami and

35:04

Rails and I don't know

35:06

Rails know plus service objects.

35:08

For every specific approach, you can

35:10

say that, all

35:12

right, in Rails, I

35:14

just threw another the in

35:17

my controller, and I'll be

35:19

fine. And you

35:21

can visually see that all

35:23

right this arm got got

35:25

bigger or got more Fetter

35:27

and with another approach.

35:29

with Hanami I've had

35:31

it. one repository, one service,

35:34

one class for my action.

35:36

And I see that

35:38

I've added another kind of

35:40

complexity. I've added three

35:42

more classes. And this

35:45

way you can compare. So you can

35:47

see that here we added internal

35:49

complexity, here we added external complexity. And

35:52

now with this approach we can measure.

35:54

for every typical feature we're

35:56

developing. think, of course, we

35:58

can, we can, we should. really simplify

36:01

this, but for

36:03

theoretical weights, it it should

36:05

be enough. this way, this way,

36:07

we amount of amount of

36:09

internal complexity, this amount of

36:11

external complexity, and we

36:13

can now compare those numbers.

36:15

those those numbers are. those

36:18

if we can. if we those

36:20

numbers, those numbers, we we

36:22

can find this balance.

36:24

We could find that optimal

36:26

approach optimal approach which would add...

36:28

optimal amount of internal complexity

36:30

and the optimal amount of

36:32

external complexity. This is what

36:34

I was talking about This is

36:36

I was talking that when

36:39

I think that it is

36:41

still is to find an

36:43

optimal way of managing

36:45

complexity in Rails applications in

36:47

whatever web applications. web applications

36:50

in bit. and Ruby. Well,

36:52

any language in any language because it's

36:54

its usage to usage to

36:56

the language itself. Sure, assuming

36:58

we put a value

37:00

to like a dude versus a fat a

37:02

fat arm, for example. we have

37:04

we have those values, then we should be able to compare them.

37:06

to compare them, right? what we're getting

37:08

at here. we're getting at yeah, you

37:10

basically you all the trade -offs you're

37:13

doing. trade-offs you're doing, changing this

37:15

thing for that thing. that This

37:17

is what you get. what you a fascinating

37:19

idea. I still think the think

37:21

the complex methods, the arms are

37:23

big, big. people might like that

37:25

because they like to have

37:27

big arms. arms so they think

37:29

that that these aren't swollen methods

37:31

of muscular. kind of muscular

37:34

Well, if you if you few

37:36

more segments to the

37:38

term, you will not like

37:40

not much. much yeah go look at uh

37:43

go get look at the go get, look

37:45

at the images on duty

37:47

or whatever. They look a lot

37:49

less like swole arms it like more

37:51

I don't know, have or like,

37:53

like those images those diagrams of of

37:55

in like biology class of of like

37:58

I don't know, know the she's wrapping

38:00

around your nerves or whatever that's called.

38:02

don't look like called. You don't Yeah, of

38:04

course, spool I said, it's a prototype.

38:06

It's a very simplistic It's

38:09

a very but still

38:11

able to express the initial idea, I

38:13

think. Oh, the initial I'm not critiquing. I'm

38:15

saying that I don't think you have

38:17

to worry about people confusing them for

38:19

swall arms. I think we're good. think. Oh, yeah,

38:21

I think I'm gonna add this to

38:23

a project I'm saying that I have a project have

38:25

a project out there right now that

38:27

I'm kind of working on on. And you

38:29

know, I'm running I'm running and Rail's best and

38:31

Rails Best Practices on it. I mean, it's a you

38:34

know, the mean, it's a pain in the

38:36

butt to test all three. to probably need

38:38

the dude's hub saying, but yeah, I bet we could, I

38:40

bet you could do this. you could do

38:42

this. be great. Yeah, and please be great.

38:44

some feedback. yeah, if give me some

38:46

feedback. you and see what you're doing

38:48

online, if people want to follow you

38:50

and see what you're doing online, where should

38:52

they go? N, N, N, N, N, N, N, N, N, N, N,

38:54

N, N, N, N, N, N, N, N, N, N, N, N, N,

38:56

N, N, N, N, N, N, N, N, N,

38:58

I guess. N, N, N, And yeah, I think

39:00

it's a good I think If a

39:03

good point. If you wanna

39:05

see other stuff I was

39:07

working on, it's can see some other

39:09

projects you can see some

39:11

other projects of mine. And probably,

39:13

of that's gig thanks on of

39:16

course, my I was working I

39:18

was working on as well.

39:20

ahead and let's go ahead and move on

39:22

to some you want do you want to

39:24

start us off? off? My pick, because

39:26

I've been doing it all

39:28

week, all week, is is Linode, the the

39:30

hosting company. I've been

39:32

watching some of their videos from last

39:34

year. from was, I

39:36

think, their 19th anniversary. their

39:39

And the founder was

39:41

pointing out that was

39:43

is an older company. older

39:45

Amazon web services for

39:47

cloud hosting. So check

39:49

them out. hosting. So check them out,

39:51

Linode Linode Linode.com.

39:54

I do my first one is sort

39:56

of a mixed PSA. PSA

39:58

slash know. know rediscovery, whatever. whatever,

40:01

back when I was like

40:03

super poor college you know,

40:05

working out was too

40:08

expensive. you know, I

40:11

did push -ups and sit -ups like all the time.

40:13

That was the same that I did. And, you

40:15

know, now. that you have to be

40:17

at home. you know, just like reminded,

40:19

I've been doing my push ups and sit again.

40:21

I've actually been doing it for a

40:23

few months, but I'm but just like

40:25

of how awesome a choice that is

40:27

since I can't go out to work

40:29

out anywhere else. So yep, That

40:32

was pick number one. I've got to ask,

40:34

how many push -ups can you do? How

40:36

many can I do or how

40:38

many do I do every day? Cause

40:40

those are completely different numbers. Well,

40:42

I want to note both the numbers.

40:44

Okay. Well, I can do about.

40:47

50 or 60 in one go. Which

40:49

is, I don't know, it's God. you're

40:51

going end up looking like one of Ivan's method

40:53

people. I feel like if I looked like one

40:55

of Ivan's method people would be harder for me

40:57

to do those push and sit -ups. But,

41:02

But, you know, I do I

41:04

do 20. in two sets every

41:06

day. So because you

41:08

don't want to go to your limit. whatever.

41:11

Anyway, I should I should be jumping

41:13

it up to 25 soon, I think. and

41:15

I'm not feeling it yet. Okay, so

41:17

the second thing that I was trying to,

41:20

But my second pick is, So

41:22

I have an acre and a quarter. and

41:25

I'm backed up to some woods and the

41:27

whole area around me is very woody

41:29

and stuff. So I have lots of brush,

41:31

things that are hard to mow around.

41:33

I live on a hill. You

41:35

know, so, like, I can't,

41:37

I can't mow that hill, at

41:39

least not on a riding

41:41

lawn mower. so I do, I

41:43

have, you know, a pretty heavy duty

41:45

weed whacker, you know, and I've just

41:47

like, you know, I don't

41:49

know, I bought the thickest string that

41:52

it could handle or whatever, cool, great. But

41:55

it, it's just what, I don't know, it just

41:57

broke all the time and everything. So anyway, the

41:59

thing that I'm picking picking is bought a

42:01

head to replace it. that I could

42:03

not only use sticker stick or string

42:06

but that I bought bought a head I

42:08

cut off these cut off these eight

42:10

inch strings and just stick in in of

42:12

having having like a reel that bumps and bumps and

42:14

then like, I don't know, stops

42:16

working for whatever reason. Anyway I've been I've

42:18

been super happy with the fact that

42:20

I like replaced my head. So

42:22

I'm not saying you should do exactly

42:25

what I did, but I what I

42:27

up online looked stuff worked how the stuff came

42:29

away. came away. I bought a head, head,

42:31

replaced it on my weed whacker. wacker,

42:33

and now, instead of taking two and a

42:35

half hours, a it takes two. it

42:37

takes lot of time, a about a half

42:39

hour, but that's a half hour of me

42:41

being pissed because my string broke. hour of

42:43

like me being totally worth it. my

42:46

it's just been a better experience. So yeah,

42:48

I know, find it. Yeah, head that works

42:50

for you. a That's my pick. So

42:52

yeah, I don't know. do you have

42:55

any picks? wacker head that works for

42:57

didn't prepare anything. Cool. But

42:59

well, I would say I would

43:01

you never tried If you using

43:04

if you're into if code

43:06

and try using using

43:08

VS Code remote I switched to

43:10

this way of working like

43:12

several months ago. months I

43:15

start new projects, I no

43:17

longer set it up on

43:19

my local machine. up on I

43:21

simply spin machine. I

43:23

simply spin up another virtual on

43:25

Leno, on Scale Way, on Amazon,

43:27

Amazon stuff, set up my

43:30

stuff my stuff

43:32

there. and simply

43:34

work. VS Remote, work as it's

43:36

in my local, as it is

43:38

on my local as it is

43:41

on my local machine. tried that, I if you

43:43

never tried that, I highly recommend it.

43:45

And I'll jump in with jump

43:47

in with a few So

43:49

So the first pick is the

43:51

discount code learned from home. on a

43:53

a site that I run Drifton

43:55

which is a series of I

43:57

believe I'm up to over 230.

43:59

230. now. So learn from home. And a from

44:01

home. as far a second pig,

44:03

as far as just talking

44:05

about code quality and stuff,

44:07

a a pretty cool gym

44:10

called Reek, which will give you

44:12

a nice of of more

44:14

of a table visualization similar to

44:16

what SimpleCov does of all the

44:18

all the different complexity issues that

44:20

you might have within your application. And

44:23

a a third pick, which is kind

44:25

of more of an anti an anti-pick, it's

44:27

a is I've been really diving a

44:29

lot more into a lately. and

44:31

it's a lot of fun. and it's been

44:33

having some enjoyment doing it,

44:36

and then also a lot

44:38

of frustration. also a lot of kind

44:40

of a So that's kind of a pick,

44:43

have another episode on have another

44:45

stuff on that that you can that

44:47

about how your views have evolved

44:50

or not evolved or not evolved. Well, so I will say

44:52

I am I will say, I am

44:54

working on a pretty cool templating

44:56

thing going I'm going to start. my

44:58

new all my new applications, not the

45:01

ones that I do for Druchmaroo because

45:03

I because I but any application that I

45:05

do personally, it's going to

45:07

have the is going to have the

45:09

file built into the

45:11

new Rails app. It's going

45:13

to then also generate

45:15

all of the files for

45:17

deploying that application over

45:19

to a Kubernetes cluster that

45:21

that enabled. So all you

45:23

have to do is

45:25

deploy it to create a

45:27

whole new environment. create a whole new

45:29

environment. So that quite a bit lately.

45:31

with that quite a bit but also Cool.

45:33

give a thumbs up for Reek a I do

45:35

use it and I do love it, right? I

45:38

do love it. just hate having

45:40

multiple. code quality checks in

45:42

one tool, and I also don't want to

45:44

go out there. to go out for one either.

45:46

pay for one I think there's

45:48

another one called one called something

45:50

like that, which will actually

45:52

do all of the the changes

45:54

for you. And it's a

45:57

pretty safe one. think safe one.

45:59

I think test double makes it. Remember, it's been a while

46:01

since I've seen it. I've seen vaguely recall it,

46:03

I'm just checking checking out. It is that

46:05

is the same. I beat. That is the same. Yep.

46:07

All right, well. Yvonne, I Yvonne, I appreciate

46:09

you coming on and talking with us.

46:11

us and that's a wrap. a wrap. guys. Thank you

46:13

guys. Thanks for inviting me. for for

46:15

coming. Yeah, thanks. thanks. That was really, really

46:18

interesting.