0:10

What? Else.

0:14

Welcome friends your listening to the

0:17

Change Log Conversations with the hackers,

0:19

the leaders and the innovators of

0:21

the software world. Today we bring

0:24

you Ellie Huxtable, the creator of

0:26

a magical open source tool for

0:28

seem keen searching and backing up

0:31

your shell history. Oh and have

0:33

you heard our Ship It Podcast

0:36

is back? Baby New hosts, same

0:38

change log goodness. Get over the

0:40

reboot. Listen at Ship It.show big

0:43

various to our partners. At

0:45

Flight Io, the home of

0:47

Change log.com easily launch your

0:49

apps close to your users

0:51

all around the world by

0:53

now. How at Fi.i Oh

0:55

okay l huxtable on the

0:57

change log. Let's do it.

1:09

Will. Friends. I have some good news

1:11

for you. It is Launch Week once

1:13

again for Century and I'm here with

1:16

Rahul Job Ria from the Prime team

1:18

at Century so Ravel! Can you tell

1:20

me about the launch read this year

1:22

for Century in March or a huge

1:24

investment into our product by for which

1:27

I'm bigger, faster, better a November we

1:29

shared a sneak peek about or new

1:31

metrics offer it. For now, developers really

1:33

define custom metrics they care about a

1:35

monitor how quickly their apples responding to

1:38

that is as measured by. After be

1:40

accountable for Plus Also right because my experiences

1:42

that are committed to an hour to be

1:44

real one an alpha people can sign up

1:46

your access I will turn off for them

1:48

like when the couple days once they write

1:50

in any can get going right away. travel

1:52

or it's like we are looking more at

1:55

how do we make the products martyr. Now

1:57

I know the world is talking about Ai

1:59

and Ml. There also are solving we think

2:01

like entertaining problems. Super center seeking a more

2:03

thoughtful approach to it. We're trying to go.

2:05

what is a developer trying to didn't like?

2:07

Our goal is not to have he said

2:09

central to our both not have as a

2:11

few tablets you need keep open or goal

2:13

is to have this be a tabby open

2:16

when something is wrong to be the information

2:18

you need right away. Tell you how impactful

2:20

is your user base and if easy chair

2:22

and if he is on bc care about

2:24

here's how to fix it. So we're taking

2:26

a broader look at how developers use a

2:28

product. Where is the noise? They're. Seeing like

2:30

seeing repeat events are they seeing like things

2:32

that are not critical rice the top and

2:35

haft automatic to resolve. the bird my the

2:37

tsunami central of smarter with artificial intelligence. To

2:39

give you a more prioritized you are the

2:41

issues that really matter. seek and solving quickly

2:44

and move on and not be distracted by

2:46

random rates. Clicks that are you know just

2:48

goes to she's Those are two major things

2:50

coming out of, like thinking about more like

2:53

to find an emergency care about have suffered

2:55

in always like organizer issues to developers who

2:57

actually saw this false faster and the we're

2:59

also. Working on a few features for our

3:02

mobile developers like centuries, a platform that works

3:04

by any technology was there. For mobile developers

3:06

there's always been this site wait a second

3:08

there was an air for hold on a

3:10

good the of this device and see if

3:12

I can recreate it. I carry created a

3:14

kilometer like the stack Trace my gets his

3:16

desk me something they'll just fix it for

3:18

Shirley's and hopefully those ballot because it's go

3:20

up and across a user it's go out

3:22

But this was solidly this idea like I

3:24

still need to figure out like where does

3:26

that bag of used old devices for someone

3:28

running insults I was thirteen. On an I

3:31

phone eleven somewhere. So we're giving them

3:33

the ability where preteen ability for mobile

3:35

developers taxi see what happens on an

3:37

end user session. So that way there's

3:39

no question about the problem or the

3:42

latency issue and building up more performance

3:44

capabilities. Latency? Exactly how faster Office with

3:46

one thousand and three de Fazer with

3:48

my are talking about. Aside from core

3:50

platform announcements and integrations and cool partner

3:53

says we're working with guess the big

3:55

investment and machine learning are percent of

3:57

okay since you must be happens more

3:59

to. 18th to the 22nd. Check the

4:01

show notes for a link to the launch

4:03

week page. We'll be showing off new features,

4:05

products. You can tune into their YouTube channel

4:07

or Discord daily at 9 a.m. Pacific

4:10

Standard Time. You can delete a scoop

4:12

or if you want to get swag

4:15

along the way, enter your email address

4:17

at the page. We'll link up to

4:19

get swag all along the way. Or

4:22

join the Discord, whatever

4:24

works for you. edessentry.io.

4:26

That's s-e-n-t-r-y.io I'm

4:29

sure the link is up somewhere. Or check the show notes

4:31

for a link. While you're at it, use

4:33

the code CHANGELOG to get a hundred bucks

4:36

off a team plan. Again, use the code

4:38

CHANGELOG. Go to century.io You

4:58

can find the link in the description. You

5:00

can find the link in the description. You

5:02

can find the link in the description. Hello,

5:07

we are here with Ellie Huestable, creator

5:09

of A2N. Ellie, welcome.

5:12

Hi, happy to be here. Happy to have

5:14

you. This is a really cool tool. This

5:17

is one of those tools that you

5:19

use and you're like, where have you been all

5:21

my life? Tool? I didn't know I needed

5:23

you. And in fact,

5:26

I didn't even know this entire subset

5:28

of things, of shell history, was

5:30

so powerful until last summer when we

5:32

did a show about the Linux command

5:34

line. And I confessed on that show

5:37

that I had just learned control R

5:39

after multiple decades

5:42

of Linux use. I

5:44

had never known about control R. And after

5:46

that show came out, I had a few people

5:48

on Mathadon say, you have to

5:51

check out A2N. It's like control R

5:53

on steroids or control R on with

5:55

levels up. I don't know. I can't remember exactly

5:57

what they said. And I checked it out and indeed it was...

5:59

It was awesome, we put that on changelog news and

6:02

multiple people found it and thought it

6:04

was cool and then a few

6:06

months later you announced you're working on it full time

6:08

and I thought, we gotta talk to

6:10

Ellie. In fact, Adam, you also linked this up

6:12

to me privately. Like, should we talk about this?

6:14

Yeah. And I was like, we already

6:17

booked it. I was like, that's the best day ever, really.

6:19

Whenever I suggest something like that,

6:21

already done. Yeah, so we're

6:23

super excited to have you. Let's open up

6:25

with A2N, what it is, and

6:28

why you built it. Yeah,

6:30

sure, so you've kind of covered it quite well, but firstly,

6:33

A2N replaces control

6:35

R, provides like a nicer search

6:37

interface for your command history, but

6:40

it goes a little bit further than that. So

6:42

there's a few similar tools that make control R

6:44

nicer, like FZF, a lot of people use. But

6:47

the main sort of this killer feature, I guess,

6:49

is that it lets you sync the shell history

6:51

you have between N different

6:54

machines. I started this because

6:56

I had the problem myself kind of scratching

6:58

my own itch. I had this really long

7:00

command that I regularly used at work that

7:02

was suddenly useful at home as well. But

7:04

my work laptop was like the other side

7:06

of the room, closed, locked, whatever. And

7:09

I couldn't quite remember how to do it

7:11

exactly. So I had to go and like boot the

7:13

machine up, login, copy paste it across, and it was

7:15

really long winded. And I

7:17

actually didn't really want to make a new tool.

7:19

I kind of assumed someone would have solved this

7:21

problem already, but I couldn't find anything. So went

7:24

from there and I had a little bit of

7:26

time off work for some reasons. And I kind

7:28

of polished up a GitHub read me, put it

7:30

online and forgot about it for a little while.

7:32

But it turns out a lot of people had

7:35

the same problem too. So it grew from that.

7:37

I think I have that problem as well. I

7:40

do a lot of homelabbing, so not a lot of production work

7:42

really. And a lot of my machines

7:44

are ephemeral because they're VMs. So I will like be

7:46

embedded and do things and test out. And like that

7:49

command history just like stuck in that machine that just

7:51

now not even real.

7:53

It's just back to digital

7:55

dust on the disk again. So the command

7:57

history is there, not with me.

8:00

Now I'm also a user of

8:02

warp which does supplant some of that and

8:04

I thought well This is sort of like an interesting thing

8:07

like does warp use something like this behind the scenes How

8:10

does this work with you know that kind

8:12

of terminal because warp is doing some smarts

8:14

there and keeping some of that history I

8:16

don't know how it works, but it does work

8:19

to some degree but not always Perfect

8:22

long history. So I imagine that a twin is a little

8:24

bit better than that Yeah I mean from what I know

8:26

of warp they're doing a lot of different things shell

8:28

history there as far as I know it's not synchronized

8:31

and The kind of frustrating

8:33

thing right now is that a two and doesn't

8:35

work on warp They're like key binding stuffs not

8:37

customizable completely as far as I know But yeah,

8:39

there's a comparison to be made but the main

8:41

goal with a two and I guess is to

8:43

run in as many places As possible. I actually

8:45

heard from someone that it runs on their phone

8:47

And I only knew this because they were complaining about

8:49

compile times on phones being quite high. I was like,

8:52

sorry what? It runs on phones So

8:54

that's pretty cool. Are they using Terminus

8:57

what are they using? Did you know what app they

8:59

were using on the phone? So it's off in my

9:01

head I think it's like term oxal or something like

9:03

that. Yeah, the one I'm using is called. I

9:06

think it's called termus Yeah termus.

9:08

Yeah, T-e-r-m-i-u-s really cool. I

9:10

like it Well, I'm not

9:12

using warp because they do

9:14

not yet support tmux. No, they

9:16

get there. I will Continue

9:19

not to use it but excited

9:21

about that I've actually had

9:24

this problem inside of tmux even I think this is

9:26

fixable with a config where

9:28

I'd have multiple tmux sessions and

9:30

even they would not share their history between

9:32

the two sessions and At

9:35

some point I found a way of fixing that particular thing

9:37

to have those shared But it was kind

9:40

of a pain and a to and

9:42

will solve that I assume as well even

9:44

on the same machine between sessions Right. Yeah,

9:46

it does That's actually something a lot of users say

9:49

there there is for most shells some config you

9:51

can do to make sure That it is all the

9:53

sessions are shared, but it's not

9:55

usually the default and it even just does

9:57

that automatically So open

9:59

source You know brew

10:01

install or app get

10:03

or actually you have a knife You know

10:05

if you're cool with curling and piping into

10:07

bash or whatever Ellie will

10:09

take care of you and just figure out

10:11

what system you're on and what particular

10:15

Shell you're running whether it's bash, you know

10:17

the shell etc and get it all

10:19

installed. Of course, you can do it manually but

10:23

Sink, you know immediately I

10:25

hear sink and I think service right

10:27

and so this I guess leads

10:29

into the bigger conversation around The

10:31

sink service because that's kind of a core aspect

10:34

of what a twin does for you

10:36

in addition to just a better UI

10:38

around finding and navigating and executing your

10:41

history but Immediately,

10:43

it's like login, you know Because

10:46

the sink service is there and I wouldn't have

10:48

and did it with a little trepidation now Your

10:50

docs are nice and you say right up front

10:52

end-to-end encrypted. So, you know, everything's safe here,

10:55

but That's gonna be a

10:58

core thing of what you're you're doing and

11:00

you're doing for an open source project now You

11:02

want to talk more about the sink side of things? Yeah,

11:04

so you said you were a

11:07

bit hesitant, right? And yeah, trust is like

11:09

super important. So then to end encryption was

11:11

a definite requirement from the

11:13

very beginning It did make

11:15

a lot of development very frustrating because like

11:17

key sharing key management All of this stuff is

11:19

just not that easy. There's also

11:21

two sides to that I guess

11:23

because users don't want to give me their shell

11:26

history if it's not encrypted But I also don't want

11:28

I think right now we've got like 70 million

11:30

lines of history on the server and I don't

11:32

want that not encrypted because Like all of the

11:34

API keys all of the credentials all of the stuff

11:36

that you shouldn't paste into your shell But lots of

11:38

people do it's just gonna be sat there and that

11:41

would make me a huge target as well So

11:43

it's kind of best for everyone the idea with the

11:45

sink service as it's all encrypted Pretty

11:47

much just a dumb blob store it sinks

11:49

encrypted data that data can be shell history

11:51

in this case it is But it doesn't

11:53

really have to be it's just small little

11:55

bits of information There's a

11:58

very tiny amount of metadata that we think

12:00

as well just in order to make the

12:02

sync work, mostly just timestamps and host names.

12:05

The host names are hashed just as long as they're different.

12:07

But yeah, it's actually very simple. I'm giving a

12:10

really in-depth tech talk about it next month, which

12:12

I'm looking forward to. But yeah. I

12:15

guess sync is required to use

12:17

A2N? Sync isn't

12:19

required. So you don't have to sign in, you

12:22

don't have to register. You can just use it

12:24

locally if you want to as like

12:26

a better search interface. If the

12:28

fact that it's encrypted still isn't enough for you

12:30

to trust me, you can go ahead and host

12:32

your own. So there's a Docker image, there's some

12:35

documentation on how to get it running. There's a

12:37

Helm chart as well. I know there's quite a

12:39

lot of people that sell host. I normally get

12:41

like questions and stuff about it most days. So

12:43

I've tried to make sure there's options there for

12:45

whatever your risk tolerance is, I guess. What's

12:48

it all written in? Can you tell us how

12:50

it works and all that, Chas? Like not the

12:52

full tech talk, but like the quick tech talk.

12:54

The teaser. Yes. So it's all in Rust. Back

12:56

when I started, my Rust was like, I've done

12:58

it on and off for years. I never had

13:00

a job writing Rust. So it was only when

13:02

I had the energy for side projects. And at

13:04

the time I was working in Go and

13:07

the idea of like doing more Go after writing

13:09

Go all day wasn't really something I

13:11

wanted to do. And I think one of

13:13

the biggest selling points of Rust, which maybe isn't

13:16

included in all of the Go versus Rust blog

13:18

posts is that Rust is just really fun. There's

13:20

loads of different ways of doing things. When it

13:22

all works, you sort of feel quite satisfied about

13:24

it. And it's just like a sense of satisfaction.

13:26

I don't think it's there with a lot of other languages. So

13:29

that was a big selling point for me. I've never

13:31

heard that one. I've heard a lot

13:33

of Rust enthusiasts and enthusiasm. I've

13:35

heard some comparisons, of course, hosting, not

13:37

hosting, but producing a Go time

13:39

podcast. Rust comes up a lot on

13:42

that and certain Gofers are excited and

13:44

other ones are dreading Rust or,

13:46

you know, combating it. And

13:49

I've heard Rust hard to learn

13:51

and sometimes complicated,

13:53

but worth it. I've never

13:55

heard fun. Never heard that. So

13:58

satisfying, I think is the keyword you use.

14:00

there. What's so satisfying about it? I

14:02

think especially in the early days, the

14:04

compiler shouts at you a lot. It's

14:07

actually very friendly. If you're new, don't be scared by

14:09

me saying shouting. It's super friendly. But there's a lot

14:11

of things you can do slightly wrong. There'll be a

14:13

lot of borrow checker errors, all these things.

14:16

When you first get to the point where you can

14:18

write a bunch of code and then there's not

14:20

really very many errors to fix, that's

14:23

a big point of satisfaction. There's

14:25

also quite a lot of functional

14:28

programming inspired functionality, so

14:31

maps and filters, all this stuff. When

14:33

you first get to the point where you're like, hang

14:35

on this big if else chain can

14:38

be written as some pattern matching and some

14:40

combinators and stuff, that also feels really nice.

14:42

I guess it makes you feel a little

14:44

bit smart, which people like. Now

14:47

you're speaking my language. I love good functional

14:49

tooling in a language or a standard library.

14:52

I think it's kind of the opposite to Go there.

14:54

Like Go is obviously intentionally very simple and Rust is

14:56

kind of the opposite, I guess, in a lot of

14:58

ways. Well said, well said. You wrote

15:00

it in Rust because this was a side project. This was

15:02

just something that you were doing to scratch an itch and

15:05

people liked it. If you look at

15:07

these stars on the GitHub star watch

15:09

thing embedded in the readme, it was

15:12

kind of slow and steady and then eventually

15:14

there were a few inflection points. How

15:16

did people find it? Did you get the word out on

15:18

your own? Did you slap up a website? Talk to us

15:20

about the boring kind of

15:23

branding and marketing side of open source. Yeah,

15:25

I think at first I just put it

15:27

on Twitter and I had a

15:29

decent number of retweets. I wouldn't say it

15:31

went viral or anything, but that got me

15:33

my first slow

15:35

momentum. The inflection points

15:37

you mentioned were mostly Hacker

15:39

News and Reddit. I tried posting on Hacker

15:42

News myself a few times and it just

15:44

never really picked up. It's always been when

15:46

other people post it. There's one inflection point

15:49

from last year in May where I was actually on holiday

15:51

and that morning I was meant to be going on this

15:53

big trip and I woke up and my phone was going

15:55

crazy and I was like, what's going on here? So I

15:57

was like in the car replying to comments and stuff. That

16:00

was really cool. I think a

16:02

lot of the marketing is down to just

16:05

having a read me that's very clear as to

16:07

what it is that makes a lot of sense.

16:10

Having a nice logo and stuff can

16:12

be very helpful. I tried to make

16:14

the install very, very straightforward. So the

16:16

curl to bash you mentioned earlier, literally

16:19

just calls your package manager for you

16:21

and add something to your ZSH or

16:23

whatever config file. So kind of

16:25

the friction for people giving it a try

16:27

is very low. So that helps a lot. I

16:29

think a lot of it's also just luck. It was

16:31

like the right thing at the right time that had, you

16:34

know, it was a need a lot of people had

16:36

and it turned out. One other thing that

16:38

really helps. I gave a talk at Fuzzdam that

16:40

wasn't really intended as marketing. It was more like

16:42

I wanted to give a talk and I had

16:44

this project. So it was a easy fit that

16:46

helped a lot. A lot of people shared it. A lot of people have

16:49

told me they saw my talk. So that was really helpful to. How

16:51

many people suggest this is a problem?

16:54

Like how many people really want their shell history

16:56

to just go with them? I

16:58

guess this is a pervasive issue with developers

17:01

at large. I think it's a

17:03

very almost bimodal problem because

17:05

I've heard some people just say

17:07

to me, like, why would I want that? I don't get

17:09

it. This is like not something

17:11

I need. I've had a lot of

17:13

people almost say what you said earlier

17:15

and like, wow, where's this been? I

17:18

guess there's been users sort of tell

17:20

me that they've been manually keeping their

17:22

bash history file over the

17:24

last 10 years and having something

17:26

automated. Really helps them out. And

17:29

I think there's a lot of people who never

17:31

really realized the value of their shell

17:33

history and having something long-term until they've

17:35

had tools that make it easier because

17:37

most people might use the default setup

17:39

where not very much history is maintained.

17:41

It's not shared across sessions, like that

17:43

kind of thing. So. I

17:47

think you don't need it until you know you need

17:49

it because you're like. Trying

17:51

to do some stuff like you did where you had a

17:53

command on your work laptop and it's like, I'm not going

17:55

to remember that. Maybe I'll throw it in docs somewhere. Maybe

17:57

that's the easy way to sort of move some things around.

18:00

Which I do that as well, but at

18:02

the same time like your your history in

18:04

some cases like muscle memory if

18:06

you could like easy search it and find

18:09

a ZFS command or The

18:12

there's a docker PS command

18:15

I run it like just has like a

18:17

nice formatting that doesn't do that like which

18:19

machines on this machine are running docker and

18:22

How many instances are there? It's just like a

18:24

prettier format that I've? carboculted

18:26

essentially around that it is just now part

18:28

of my history and then when it's not

18:31

I bring it back in but having this it's like

18:33

well I don't ever have to lose that history ever

18:35

again Plus with the nature of

18:37

the terminal the more you use it the

18:39

more you kind of Scrap

18:42

together you know longer

18:44

and longer commands over time that

18:46

do something very specific and

18:48

then not need that For a very long

18:50

time yeah I found and then six months

18:52

eight months two years later, and you're like I've

18:55

done this before and I can't remember exactly what

18:57

I did, but I know I Called

18:59

cut at some point you know or like some

19:02

Terminal command is in there And

19:04

if I can just type in cut and look at

19:06

all the I don't cut very often But if I

19:08

have the history you know I can remember a cut

19:10

I can look at the seven times I use cut

19:12

and one of them you'll see like there it is

19:14

right there And you just save yourself

19:16

so much time being able to just call you know Remember

19:19

that without having to remember it it

19:21

really is a time saver and

19:24

I got sold on you know setting your history

19:26

to like a Good

19:28

million or whatever is it infinity? I don't

19:30

know it's a many Ellie you probably know

19:33

how this works better than we do Is

19:35

there like a maximum history size histo size? I

19:37

think might be the yeah, it's his size I

19:39

always used to just spam a bunch of numbers

19:41

until I thought that was probably enough Pretty

19:45

much like how large is my disc you

19:47

know yeah Yeah, I had this one command

19:49

back in my you know more tried and

19:51

true Ruby days whenever I would blow

19:53

away gems It was

19:55

a for loop that is now on a

19:58

Mac that doesn't exist for me And

20:00

if I wanted to, like you said, recall

20:02

that command, like it's just gone. It's not

20:04

part of this new environment. Maybe

20:07

I'm running Jekyll and I want to, you

20:09

know, blow away the gems and do something

20:12

different. Now I don't have that for loop readily

20:14

available. I could probably Google it. Maybe

20:16

there's a gist out there where I like stuck it

20:18

in a gist on GitHub somewhere from back in the

20:20

day, which is how we used to sort of like

20:22

stow away some ideas and kind of share them too.

20:26

But, you know, now if I wanted to do that

20:28

for loop, I'd probably just be like, like

20:30

you, I don't know how I did it. You

20:32

know, forget it. I find as well that as

20:34

time goes on, the number of sort of original

20:37

commands you write goes down. Like even if

20:39

you were recalling something from three, four years ago,

20:41

like you've written it before and having that history

20:43

means that like, the longer you have

20:45

it for, the less time you spend typing and that sort

20:47

of compounds over time. For sure. It's

20:50

like, maybe there's a saying like command

20:52

line, I don't know what prompts, commands

20:55

are code too, you know, like.

20:57

For sure. I don't know, maybe

20:59

that translates here because like you kind of want to

21:01

keep it in a way and it's not quite code

21:03

you write and gets stored. It's

21:05

history. Well, the entire website

21:07

is dedicated to awesome command lines, right? Like

21:10

commandlinefu.com. I'm not sure

21:12

if it's still active, but I used to go there

21:14

and just read other people's command lines and be able

21:16

to learn what I can do. Ellie,

21:19

your statement about reusing

21:21

the same commands over time rings

21:23

true because one of the coolest things about A2N

21:26

is the stats subcommand. It

21:29

has stats built into it. And

21:31

so you can get a top list of

21:33

your commands and also list out total commands

21:35

and unique commands. And I have in

21:38

mine 15,000 total command history, which

21:41

I just installed it and then loaded

21:43

my history into it, 15,511. I

21:46

wonder if my hist size is 15,000 or something like

21:48

that. Or if that's just really how many

21:50

commands I've done in my days, at least

21:52

on this computer. And of those, 15,500, 4,800 are unique.

21:59

So what does that mean? Something like that.

22:01

Something like repeating myself quite a bit. You

22:03

know? Like two out of three commands is not

22:05

unique. I think that's right. That's top of the head

22:07

mass. Here's the moment where I

22:09

come off later sounding like an idiot, Ellie. Like, well, this math is awful.

22:13

But the point stands. Lots of repeat

22:15

commands in there, you know? Oh,

22:17

it's fine. We've got a channel on the Discord where people

22:19

share that sort of a two in stats output.

22:22

And the number of times it's CD

22:24

or LS at the top is huge.

22:27

Yeah, I got CD at number one as well. What

22:29

else do you do? Mix. I

22:32

run mix a lot. So I'm an elixirist. So I run mix.

22:34

Sure. Which is usually just mix test.

22:36

I just run mix test over and over again. Git

22:39

ST, which is Git status. That's my shorthand

22:41

for Git status. IEX,

22:43

which is also elixir. That's

22:46

starting the elixir shell. And

22:48

then Smug. What's Smug? Smug

22:50

is a Go-based tool

22:52

for managing TMUx sessions. So

22:55

you can define different TMUx sessions using

22:58

YAML or some sort of little config file format.

23:01

And you can use Smug to manage them. So

23:03

I can say Smug start change log and that

23:05

will launch a specific TMUx session with three terminals.

23:08

And this one runs that command. This goes to

23:10

this directory. This starts that command. And

23:13

then I can just switch between them quickly.

23:15

Because I have a few things that I do,

23:17

but that's about it. And so I use

23:19

Smug to just start that thing. So spend this

23:21

one, start that one. And it always is just set up

23:23

the way you like it. So I run that command a lot in

23:26

order to get into whatever it is that I'm up to. So that's

23:28

a cool one. It's based off

23:30

of, there's an old project in

23:32

Ruby called TMUXinator, which does the same

23:34

thing. And I think TMUXinator

23:37

either got unmaintained

23:40

or I got sick of having

23:42

to deal with Ruby with regards

23:44

to it. And I wanted a universal binary. And

23:47

Go is very good at that. And so

23:49

there's one called Smug in Go, which

23:51

I replaced it. And I think

23:54

it even uses the TMUXinator config format or something. So

23:56

it's like an easy switch. So that's my

23:58

top five. And then LS

24:00

on it coming to number six for

24:03

sure. That's kind of a cool tool

24:05

because getting started half the

24:07

battle right? I guess it creates that momentum.

24:09

There's no minutia to do no ceremony to

24:11

do so it's like what was

24:13

it smug change log that was it? It's

24:16

not a start change log. Yeah. Yep. And you're off

24:18

to the races because like hey, this is what my

24:20

environment is whenever I do this development and starting

24:23

is pretty easy. It's already set up for me.

24:25

It's pretty cool. I like that. So

24:29

LS, CD kind of boring. And

24:32

FG as well. You mentioned the

24:34

command Z thing earlier. So that's on there

24:37

for me. Get status

24:39

and lvim as well. So

24:41

lvim is basically neovim but it's

24:43

like a distribution. I had a neovim

24:45

setup I maintained for like years and

24:48

years. And I got to the point where

24:50

I was tired of constantly updating plugins. So

24:52

luenovim just has them all from the get

24:54

go. I would share

24:56

mine but I haven't trusted you yet. I'm

25:01

actually stuck. I was like let me trust

25:03

Ellie in a VM. And

25:05

so I use warp issh in this

25:07

VM. And the thing I

25:09

get whenever I do H1 status

25:12

is maybe you can help

25:14

me with this. It says

25:17

error could not fetch history.

25:19

I did import my history. And then

25:21

if I do a different stats,

25:24

for example, it says error failed to find

25:26

a two in session in the environment check

25:28

that you have correctly set up your shell.

25:30

Now use your command. This is

25:33

a boom two. I think it's

25:35

2304 potentially. And I

25:37

just use your command. I'm like yeah, I trust your bash

25:39

command. I'll just whatever you want to do on my machine

25:41

because this is a VM. I don't care, right? And that's

25:43

all I did. And then I was like, well, maybe, maybe

25:46

I missed that up somehow. And so

25:49

I did the import process to import

25:51

my zsh history. And thus

25:53

far, I only get errors whenever I do

25:55

a twin stats or history because there's nothing

25:57

there yet. We usually see that

25:59

error later. when the sort of

26:01

shell plugin half isn't properly installed. So

26:04

maybe it installed to a different

26:06

config file to the shell you're using, which

26:08

can happen sometimes depending on the setup. So

26:11

a two end session basically just tracks

26:13

like the current shell session. So we

26:15

track history per, even though they're all merged,

26:18

we also track the history per session. And

26:20

if it can't find that variable, it's not

26:22

being set by the shell integration. Would

26:25

that be my like ZSH file then? My RC file?

26:27

Yeah, it would be your ZSHRC. It

26:29

should be like a two end in it. Yeah,

26:32

I don't think it's in there then. So

26:34

that's probably my issue. Well, actually it's eval.

26:36

Yeah, there's an addition at the very bottom.

26:38

It's eval A2N and it ZSH. Doesn't

26:41

need to be a live debug here, but that was my

26:43

hurdle here. I was trusting in a VM. Okay. I was

26:45

trying my best to

26:47

protect my machine from Ellie at all

26:49

costs. Well, you wouldn't have your

26:51

history in there either then. Well, I actually did it

26:53

in a VM that I know I don't need anymore,

26:55

but does have some ZFS history. So I was like,

26:58

well, it's probably going to be a ZFS fine

27:01

tune command. It's like an older thing that... Retainted

27:03

man. Yeah, it's got some history in it. So

27:05

I was like, okay, I can use this. And

27:08

the VM has a backup yesterday. So I'm cool

27:10

with like, if this VM dies, I'll just like

27:12

blow it away and restore from a backup. That's

27:15

how Proxmos works. You know, it's pretty easy. Okay.

27:17

So I've got this issue. Maybe I'll try to

27:19

my Mac and I haven't gone deeper, but don't

27:22

feel bad. I do trust you. I

27:25

do trust you. Wait, if you don't, it's cool. You can write

27:27

it on your home lab too. That's true. Well, I

27:29

was trying to. That'd be a cool thing. Adam has

27:31

set it up on your home lab and run it.

27:33

Yeah, I actually like that idea a lot like running

27:35

your own server. I think it's super cool that you

27:37

offer that. I, you know, I did like a

27:40

lot of the install process. I can at least share some

27:42

of my thoughts on experience here, which was your thoughtfulness

27:45

in the messaging, I think is super cool.

27:47

Like you have the ASCII art going on

27:49

and you've got these very polite prompts. Detected

27:52

Linux checking distro Ubuntu detected. It

27:54

gives them all this details,

27:57

you know, all this different stuff. Runs.

28:00

AppGet update for me, which is super kind of

28:02

you to give me the latest repositories available. Super

28:04

kind. Yeah, super kind of you. And at the

28:06

very end, it does all its thing and it's

28:08

like, hey, thank you. And this is how you

28:10

can like use it and contribute and become

28:13

a part of the community, which I think is super important

28:15

for any tool that wants to be

28:17

adopted. Like tell me how to adopt, right? And you've

28:19

done that. So very, very well done on

28:21

that part of it, even though I've had this ZSH

28:24

eval issue. I'll figure it out. Sorry

28:27

about that. But no, thank you. It's, obviously I

28:29

don't know how many people click on the links

28:31

at the bottom of the installer, but I hope

28:33

that it does help to some

28:35

degree. Yeah. And you did give a prompt

28:37

with the next step too, which is A2N register

28:39

or A2N login. So you've already

28:41

done that. It is kind

28:43

of hard. I would say if I didn't go there

28:45

and do that, it would be kind of hard to

28:47

understand what the next steps are. Like give me a

28:49

quick win. Like the

28:51

import, like, hey, you might want to import because

28:54

I didn't know that. That might be

28:56

something to add there. Like, hey, to get

28:58

started, just run this command to import.

29:00

And then you can run stats as an example,

29:02

because you probably have all the data and you

29:04

can run stats because you have all of it

29:06

imported. And you get a quick win. Like, Oh,

29:09

I get how this thing works. I get how this

29:11

thing will help me. That's a good suggestion. Thank you.

29:13

I'm actually thinking of doing like A2N setup, which just

29:15

is kind of like a little interactive wizard that just

29:18

does the whole lot for you. Better, even better. Yeah.

29:20

A2E, that could be fun. Yeah. Well,

29:23

I would definitely, I would agree on the setup process

29:25

and the documentation was all very handholdy and nice. And

29:27

even when I was feeling like, I don't know if

29:29

I want to be like right there, you're like, no,

29:31

it's A2N, I'm like, all right. So this person cares.

29:34

I also kind of figured that like, if I was a

29:36

user, I would be like, no way are you getting my

29:39

data. So I have to make sure it was like something

29:41

I would use if I was on the other side. What's

30:03

up friends? This episode is

30:05

brought to you by our

30:07

friends at Sanadia. Sanadia is

30:09

helping teams take NASA to

30:11

the next level via a

30:13

global multi-cloud, multi-geo and extensible

30:15

service fully managed by Sanadia.

30:17

They take care of all

30:19

the infrastructure, management, monitoring and

30:21

maintenance for you so you

30:24

can focus on building exceptional

30:26

distributed applications. I'm here with

30:28

VP of product and engineering Byron Ruth. So

30:30

Byron, in the Nat versus

30:33

Kafka conversation, I hear a

30:35

couple different things. One I hear out there, I

30:38

hate Kafka with a passion. That's quoted by the

30:40

way on Hacker News. I

30:42

hear Kafka is dead, long live Kafka.

30:44

And I hear Kafka is the default,

30:46

but I hate it. So

30:49

what's the deal with Nat versus Kafka? Yeah,

30:51

so Kafka is an interesting one. I've

30:53

personally followed Kafka for quite some time

30:55

ever since the LinkedIn days. And I

30:57

think what they've done in terms

31:00

of transitioning the landscape to event

31:02

streaming has been wonderful. I think

31:04

they definitely were the sort of

31:07

first market for persistent data streaming.

31:09

However, over time, as people have

31:11

adopted it, they were the first

31:14

to market, they provided a solution.

31:17

But you don't know what you don't know

31:19

in terms of you need this solution, you

31:21

need this capability. But inevitably, there's

31:23

also all this operational pain and

31:25

overhead that people have come to

31:28

associate with Kafka deployments. Based

31:31

on our experience and what users and

31:33

customers have come to us with, they

31:35

would say, we are spending a ton

31:37

of money on spend on a team

31:39

to maintain our Kafka clusters, or

31:42

managed services or something like that.

31:45

The paradigm of how they model

31:47

topics, and how you partition

31:49

topics, and how you scale them is

31:52

not really in line with what they fundamentally

31:54

want to do. And that's

31:56

where Nat can provide, as we

31:58

refer to it, subject based addressing, which

32:01

has a much more granular way

32:03

of addressing messages, sending messages, subscribing

32:06

to messages and things like that, which

32:08

is very different from what Kafka does.

32:10

And the second that we introduced

32:12

persistence with our Jetstream subsystem as

32:15

we refer to it a handful

32:17

of years ago, we literally had

32:19

a flood of people saying, can

32:21

I replace my Kafka deployments with

32:24

this NATs Jetstream alternative? And

32:26

we've been getting constant inbounds, constant customers

32:28

asking, hey, can you enlighten us with

32:31

what NATs can do? And oh, by

32:33

the way, here's all these other dependencies

32:35

like Redis and other things and some

32:38

of our services based things that we

32:40

could potentially migrate and evolve over time

32:42

by adopting NATs as a technology, as

32:45

a core technology to people's systems and

32:47

platforms. So this has been largely organic.

32:49

We never from day one, you know,

32:51

with our persistence layer Jetstream, the intention

32:54

was never to say we're going to

32:56

go after Kafka. But because of how

32:59

we layered the persistence on top

33:01

of this really nice PubSub, Coronets

33:03

Foundation, and then we promoted it

33:05

and we say, hey, now we

33:07

have the same, you know, same

33:10

semantics, same paradigm with these new

33:12

primitives that introduce persistence in terms

33:14

of streams and consumers. The floodgates

33:16

just opened and everyone was frankly

33:18

coming to us and wanting to

33:20

simplify their architecture, reduce costs, operational

33:22

costs, get all of these other

33:24

advantages and NATs has to offer that Kafka

33:26

does not whatsoever or any of the other

33:29

similar offerings out there. And you

33:31

get all these other advantages that NATs has to offer. So

33:33

there's someone out there listening to this right

33:35

now. They're the Kafka cluster admin. The

33:38

person in charge of this cluster going down

33:40

or not. They manage the team, they

33:42

feel the pain, all the things.

33:44

Give a prescription. What should they do? What

33:47

we always recommend is that you can

33:49

go to the NATs website, download the

33:51

server, look at the client and model

33:54

a stream. There's some guides on doing

33:56

that. We also have, so NAT provided

33:58

a basically a packet of resources. to

34:00

inform people because we get again so

34:03

many about requests about how do you

34:05

compare NAPs and Kafka and we're like

34:07

let's actually just put a thing together

34:09

that can inform people how to compare

34:11

and contrast them. So we have a

34:13

link on the website that we can

34:16

share and you can basically go get

34:18

those set of resources. This includes a

34:20

very like lengthy white paper from an

34:22

outside consultant that did performance benchmarks and

34:24

stuff like that and discuss basically the

34:27

different trade-offs that are made and

34:30

they also do a total cost

34:32

of ownership assessment between people who

34:35

are organizations running Kafka versus running

34:37

NAPs for comparable workloads. Well

34:39

there you go. You have a

34:42

prescription. Check for a link in

34:44

the show notes to those resources.

34:46

Yesterday's tech is not cutting it.

34:48

NAPs powered by the global multi-cloud

34:51

multi-geo and extensible service that is

34:53

fully managed by Sanadia. It's the

34:55

way of the future. Learn more

34:57

at sanadia.com/change log that's s-y-n-a-d-i-a dot

35:00

com slash change log. So

35:07

when you decided to quit your job to do

35:10

this what were

35:12

you thinking? What when who what when were

35:16

why? So I

35:18

started 2023 and this was just

35:20

like a random little side project right? It

35:22

was it was going okay. It was some

35:24

bit of fun and then um after my

35:26

talk at FOSDEM uh some of the other

35:28

speakers I was speaking to one of them sort of says

35:30

to me like you need to take this more like you

35:32

should take this more seriously think about monetizing

35:34

it etc etc and

35:37

I wasn't 100 convinced at the

35:39

time but it sort of sat in the back of my head

35:41

and throughout that year after a

35:43

lot of the reception I got I started

35:45

spending like an extra few hours per

35:48

like few days um just polishing improving

35:50

that kind of thing and

35:53

the user growth we got was huge so like

35:55

we had more growth in 2023 than we'd had

35:57

in the previous two years by a long shot.

36:00

and kind of got to the end of

36:02

the year and I'd always fancied having

36:04

a go at building my own company and

36:07

I had this project that was continuously growing.

36:09

It was sort of continuously demanding more time

36:11

as well. And I got to the point

36:13

where I was like either I give this a

36:15

go and try to make something of it or

36:18

I reduce how much time and energy I'm

36:20

putting into it because like it's not sustainable.

36:22

I can't like have a full-time

36:24

job and be like handling support

36:26

and prioritizing issues and all of this in a way that's

36:29

going to continue to grow. I think

36:31

there's a bit of a gap in sort

36:33

of the shell as it is. It

36:36

hasn't really changed a whole lot in a

36:38

long time. And the way you

36:40

said earlier like about how you know the

36:42

shells code to or something like that maybe I'm not quite

36:45

in you exactly but I think there's a

36:47

ton of developer tools focused around writing

36:49

software but people who spend a lot of

36:52

time in the terminal are almost underserved right

36:54

now. Yeah, I agree. The

36:56

command line code is code to I guess

36:58

I don't know your command line commands are

37:00

code to. Exactly. I feel that even

37:02

during the process like you recognize which dish drawer

37:05

I'm using like that's a cool code. I'm sure

37:07

is that in Rust as well like what

37:10

are you doing to confirm like which dish drawer do you

37:12

have like a massive statement or

37:14

something like that or a case statement. It's like

37:16

a chunky bash script. Yeah, okay. I mean that's

37:18

fun too. I mean code to. Yeah. And

37:21

I like I like bash scripts even too.

37:23

I mean I like them because they're useful

37:25

tooling that you can make for you. They're

37:27

very bespoke for what you need and I

37:29

think you know in my last I say

37:32

year and a half I've become more empowered

37:34

with chat GPT encoding tools because there's just

37:36

so much knowledge out there in a island

37:38

to generate or at least

37:41

guide me on bash scripting where I never

37:43

really I felt intimidated by it because

37:45

I was like I'm not that kind of hacker Adam

37:47

come on. You know you can't do this but

37:50

then I'm like I need to do it and so I found

37:52

out how to do it with my buddy code

37:54

Jenny I stuff and now

37:56

I have lots of Linux tooling that

37:59

I use myself. for various things

38:01

that are not really probably useful

38:03

to anybody else, but they're useful to me only,

38:06

and they could be highly specific. Whereas

38:08

before I would have never done it, but it's

38:10

telecode too, and I have no idea where to

38:12

keep it at, like it's in a repository, but

38:14

now it's the true version of it, how do

38:16

you deploy it, how do you easily

38:18

get into like an apt repository you

38:20

can install yourself, that whole world is weird.

38:22

Like if you write your own little tools,

38:25

how do you install your own little tools? I

38:28

would love somebody to solve that problem better, or at least

38:30

document how to better add a

38:32

repository to apt so that I can easily

38:35

install my own things from like a repository,

38:38

because my central repository of it essentially is

38:41

living in my bin folder, my own personal bin folder,

38:43

wherever I put my stuff at, that's the version, and

38:46

it changes without being versioned, because I'm

38:48

an idiot, it's just my own tooling.

38:51

That was better, that'd be cool. Definitely be very cool.

38:54

Adam, do you remember RVM? Oh yeah,

38:56

of course. Ellie, are you aware

38:58

of RVM, Ruby, version manager? Yeah, there's a

39:00

few similar for other languages as well. Yeah,

39:03

it was kind of the first one that

39:05

ended up being like NVM, no version manager,

39:07

et cetera. And RVM was

39:10

famous, but also a little bit infamous

39:12

because the entire thing was written in Bash,

39:15

and it was a lot of code, and

39:18

it was very complex once you got into there, and

39:21

RVM had some, I mean, a lot of users,

39:23

a lot of bugs, a lot of

39:25

issues, and then as a user

39:27

you dive into it, and you're a Ruby person or

39:29

something, and all you're seeing is like, I

39:31

can't remember the guy's name who wrote it,

39:34

but he knew Bash very, very, very well,

39:37

to the degree where you're like,

39:39

oh, this is borderline, too much

39:41

Bash. Anyways, at

39:44

one point he decided that he

39:46

was going to make a package manager, I

39:48

guess for the listeners who aren't aware, RVM, Ruby

39:51

version manager. Michael Puppies is his name. Michael

39:53

Puppies. We talked to him back in 2013, Jared.

39:56

That was a long time ago. I feel like

39:58

there was another fellow that was, that's not Michael Puppies. I

40:00

remember Michael me call Wayne. Yes.

40:02

Yeah Wayne Seguin. Maybe he's yes

40:05

Wayne segwin or segwin segwin Yeah, did they work

40:07

on it together? I think he was the originator.

40:09

Okay, and Michael poppies took it over and had

40:11

to maintain it Okay, it's all coming back. I

40:13

think Michael took over a lot of the issues

40:15

really though And he yeah, there was like a

40:18

burnout stage there too with with Michael Wayne was

40:20

super into bash Maybe he still is I haven't

40:22

kept up with him Yeah, and

40:24

I remember at one point he talked about

40:26

riding a bash package manager For

40:28

X where X is whatever it is you wanted

40:31

to manage It was one of these like I'm

40:33

going to engineer the general thing that solves all

40:35

problems kind of situations that we get into as

40:37

engineers and the reason I think

40:39

of it is because his deal is like you should just be

40:42

able to Like package manage your bash

40:44

scripts just like you can other things which

40:46

is kind of what you're asking for Adam

40:48

Yeah, it's like just give me a way of like

40:51

Packaging up and deploying even if only to

40:53

myself Which a lot of

40:55

us should do with our dot files synced across machines,

40:58

you know, it's like you

41:00

write your little bash scripts and your functions and

41:02

stuff into your dot files and use git to

41:04

basically Synchronize those across machines, which

41:07

is a poor man

41:09

solution, but it works, you know to a

41:11

certain degree better than you know going the

41:13

route of like doing

41:16

it distro or package manager

41:18

specific like at for example, you

41:20

know, cuz if I'm gonna do the same if

41:23

I need the same bin options in

41:25

on my Mac then You

41:28

know, maybe there's you know

41:30

dependencies required or whatever. I don't know like

41:32

usually you'll end up with some sort of

41:34

platform specific Yeah conditions in your scripts which

41:37

I know I've had in my days where like check is

41:39

it Mac or is it Linux? All of those ones I've

41:41

been writing are they translate from Mac

41:44

to Linux? No problem Yeah, just as

41:46

so long as the package there like 70 or ZFS

41:48

or whatever like those things As

41:51

long as it's present and it runs and there's no way

41:53

to do in the bash. Anyway, yeah, we're in the weeds

41:56

I was just thinking I was just reminiscing. Yeah a little

41:58

bit about that as something that And that doesn't

42:00

exist that I know of. Maybe people know ways

42:02

that you can deploy out your

42:05

little scripts to yourself and

42:07

to others in a way that's

42:09

a lot like NPM install or

42:11

cargo install, where they have these

42:13

language-specific things. This would be

42:15

more generic. I think it would be cool.

42:17

Also, even for internal tooling, like I've been

42:19

in jobs before where some of the onboarding

42:21

is like, please copy paste all of this

42:23

from the wiki into your .file. Totally. It's

42:25

not great, right? Yeah. Maybe

42:28

a new frontier for A2N at some

42:30

point. I actually just made it sync

42:32

aliases earlier today. Really? So there's some

42:35

scopes for that too. Okay. So

42:37

now we're starting to see what maybe a potential future for

42:40

this looks like. Maybe. I use

42:42

something from Thoughtbot for some of these things.

42:44

I think it's RC Up or RC Something.

42:47

There's lots of different... People have taken different whacks

42:49

at this, I think, over the years. But

42:52

yeah, syncing aliases. Yeah,

42:54

that's kind of the inspiration there. I had a

42:56

lot of feedback from users that it's the

42:58

first thing they install on their machine. And

43:00

then from there, they recall other commands for

43:02

setting up the rest. And

43:04

it's like, well, if you didn't have to recall all

43:07

the commands and you could just install it on your

43:09

machine and then your setup's right there, then that would

43:11

be the next step. Bam.

43:14

Yeah. So this could be your .file

43:16

syncer and your environment setup syncer without

43:18

a GitHub repo or something.

43:20

All you need is A2N and you're

43:22

already logged in. Exactly. I

43:24

love it. I love it. We sidetracked

43:27

a little bit. Is it trivial to run your own

43:29

server in a home lab or on-prem? Is

43:31

it pretty easy? It should be. And

43:34

if you have any problems, let me know because I want to make it

43:36

easier. It's just

43:38

Postgres, at least Postgres 14

43:40

and run a binary.

43:43

And that's it. I

43:45

guess I could probably Dockerize that. I would

43:47

prefer Docker in that scenario. We

43:49

have a Docker image too. There's a Docker image, there's

43:51

an example Docker Compose, and there's a Helm chart. Oh,

43:54

that's too easy there. I think there's

43:56

some Kubernetes community distro

43:59

thing. have the home labs that has it

44:01

included too. I can't remember what it's called now. I'll

44:04

test it out after this conversation. I'll

44:06

boil it a bit. I'll give you some feedback.

44:08

Oh, I see it. Yeah. Thanks. Under your

44:10

docs, self hosting, server setup, usage, Docker, Kubernetes.

44:13

So very cool. Here's a setup that

44:15

I think has worked in the past. It's like open source,

44:18

library or framework or system

44:20

with a hosted business

44:23

attached to it where

44:25

the end user is

44:28

more mainstream, more normal, less

44:30

nerdy. But when you have

44:33

open source tool service

44:36

as the business model with

44:38

your core audiences, super

44:40

nerds, like Adam, who's

44:42

like gets excited because he heard Docker

44:44

or somebody even nerdier

44:47

who hears Helm Chart and is like, let's go. You

44:49

know, which I'm sure some of our listeners are like,

44:51

okay, the Helm Chart, cool. It

44:53

seems like that's harder for the business side, isn't it? Where it's like

44:56

the people who are your core demographic

44:58

of potential purchasers, because they're all

45:00

command line users, they also

45:02

are super down with self hosting. I

45:04

think it's split. I think a lot of the

45:06

sort of early users are very down for

45:08

self hosting, very down for customizing things.

45:11

But I think there's tons of people

45:13

that don't actually know how nice the

45:15

shell can be and how things can be

45:17

better. And if there was a

45:19

very easy way for them to have a nice

45:21

setup that felt modern and was good, and they

45:24

didn't have to know what sort

45:26

of scripts and plugins they need to install and

45:28

which things they need to add and all of

45:30

this, which most people will just stop caring as

45:32

soon as they start seeing this huge list of

45:34

stuff. It was nice with no effort,

45:36

I think we could have more people using the terminal

45:38

much more. I agree with you. And

45:41

I also want to recognize that

45:43

of amongst the nerdy ones like

45:45

Adam, there's also me and I would just I have

45:47

no interest at all in self hosting, even though I

45:50

totally could do it. I just don't want

45:52

to, I would happily sign up for the service and

45:54

pay the money on a recurring

45:56

basis to have you handle that for me. So even

45:59

amongst you know, There's a lot of home

46:01

labbers, a lot of self-hosters amongst

46:03

us nerds, and then it's also the ones that are

46:05

just like, yeah, I'd rather not.

46:07

I'd rather give Ellie my money. Maybe I'll hold

46:09

you to that recurring payment thing. Hey, I'm into

46:11

it. I'm interested for sure. Especially

46:14

if you are just gonna like automatically

46:16

sync all of my things. And

46:20

it's cool, it's very cool. My

46:22

reasoning is not to pay. We know that,

46:24

we know that you're just tinkerer. We even

46:26

talked about that with Obsidian. Nick and I

46:28

talked about that when we were at that

46:31

conference. We were talking about Obsidian. Are you

46:33

Obsidian? I love Obsidian.

46:35

It's my favorite thing. And you use their

46:37

sync service? I do. Were

46:39

you early enough to get the half off discount?

46:41

Nick got, Nick Nesey got this like, I

46:44

don't know, early adopter super cool person

46:46

discount that's like forever. Cool person discount. Because

46:49

Nick is cool. I don't think that strikes

46:51

true for Nick, but come on. We'll see.

46:55

Well anyways, I paid 10 bucks a month. You paid

46:57

10 bucks a month for a sync? Yeah. Yeah.

47:00

I don't not wanna pay that number, but

47:02

I think that's not exactly what

47:04

I think the value is that Obsidian gives me,

47:07

is syncing. It seems

47:09

to be pretty trivial to do.

47:11

Like why not just give somebody the option to

47:13

run their own server? I

47:15

don't know. But that's why I think about this.

47:17

Like not so much not paying you, but more like just

47:20

even data protection or just for learning. How

47:23

does this actually run? What does it take to, because

47:25

I don't get to do SRE stuff as a living.

47:27

I'm a podcaster. You know, my SRE

47:30

for a living is myself in my

47:32

home lab and tinkering. Yeah. And

47:35

so that's why I do it to learn. Less about keeping my

47:37

own dollars. Yeah, when I

47:39

have been SRE for a living, my

47:41

home lab got neglected quite a lot.

47:43

Yeah. What's

47:46

your home lab then? It's pretty simple.

47:48

It's just like a little Yvonne Tusser

47:51

of a bunch of ZFS stuff running

47:53

on it. What a bunch of Docker

47:55

containers for my home media setup and

47:57

like a Prometheus and Grafana dashboard.

47:59

sets up and stuff. Doing a little bit

48:02

of home automation at the moment just because maybe

48:04

future me will want to know the temperature

48:06

in my kitchen 10 years ago. What's

48:39

up friends, this episode is brought

48:41

to you by Image Proxy. Image

48:43

Proxy is open source and optimizes

48:46

images for the web on the

48:48

fly. They use the world's fastest

48:50

image processing library under the hood,

48:52

LibVeeps. It makes websites and apps

48:55

blazing fast while saving storage and

48:57

fast costs. And I'm joined by

48:59

two guests today, Sergei, Aleksandrovich, author,

49:01

founder, and CTO of Image Proxy

49:04

and Patrick Byrne, VP of engineering

49:06

at Dribbble, where they use Image

49:08

Proxy to power the massive amounts

49:10

of images from all of dribbble.com.

49:12

Sergei, tell me about Image Proxy.

49:14

Why should teams use this? Everyone

49:16

needs to process their images. You

49:19

can just take an image and

49:21

just send it to the browser

49:23

because usually it's a megabyte of

49:25

data. And if you have lots

49:27

of images like Dribbble does, you

49:29

need to compress and you need

49:31

to optimize your images and you

49:33

need them in the best quality

49:35

you can provide. That's where Image

49:38

Proxy shines. Very cool. So Patrick, tell

49:40

me how Dribbble is using Image Proxy.

49:42

Being a design portfolio site, we deal

49:44

really heavy in all sorts of images

49:46

in a lot of different sizes, levels

49:49

of complexity. And when we serve those

49:51

images to the users, those really have

49:53

to match exactly what the designer intended.

49:55

And the visitors need to receive those

49:58

images in an appropriate file size. dimensions

50:00

depending on whatever their internet connections beat

50:02

is or their device size is. And

50:04

that was just a constant struggle really

50:06

to really thread that needle throughout the

50:08

course of the platform using a handful

50:10

of different tools in maintaining that right

50:12

balance of high degree of fidelity, high

50:15

degree of quality without sacrificing the

50:18

visitor's experience. And when we

50:20

were exploring using Intracy we were able to

50:22

spin up using the open-source version of the

50:24

product, a small ECS cluster, to just throw

50:26

a few of those really tough cases that

50:28

went through our support backlog, looking at

50:30

some of the cases people were reporting

50:32

and almost to a T we aced

50:35

every one of those cases. Wow, so

50:37

it seems like image proxy was really

50:39

impressive to you. Yeah Adam, I just

50:41

have nothing but positive things to say

50:43

about using image proxy. The documentation is

50:45

crystal clear. Out of the box it

50:47

does everything you need it to. Tweaking

50:49

it to meet your specific needs is

50:52

incredibly easy. It's wicked fast. It deploys

50:54

real simple and our compute costs

50:56

have gone down over the open-source tools that

50:58

we've used in the past. Even

51:00

including the image proxy pro license

51:02

it still costs us less to

51:05

use and gives our visitors a

51:07

better experience. So as an

51:09

engineer I like using it and as an

51:11

entering manager I like what it does to

51:13

our bottom line. So image proxy is open

51:16

source and you can use it today but

51:18

there's also a pro version with advanced features.

51:20

Check out the pro version or the open

51:22

source version at imageproxy.net. The one thing I

51:25

love so much about this is that a

51:27

matter which you choose the open source route

51:29

or the advanced features and the pro

51:32

version you use the same way to

51:34

deploy it. A Docker image one

51:36

that is from Docker Hub that

51:38

everybody can use open source or

51:40

one that's licensed differently for those

51:42

advanced features that's the same delivery

51:44

mechanism via Docker Hub. It's so

51:46

cool. Once again imageproxy.net. Check out

51:48

the demo while you're there and

51:50

check out the advanced features for

51:52

the pro version. Once again imageproxy.net.

52:00

So when you decided to quit the job, let's go back

52:02

there. Yes, I like this go back there. I was about

52:04

to go back there. Yeah, good. We,

52:06

we, we moved far. I think

52:08

they were boring me, Adam. Yeah, I am good. You're

52:10

on track or on the same page. I also was

52:12

happy you brought us back to the, to

52:14

the home lab, but well,

52:16

you kind of gave a little bit of

52:19

your thought and your reasoning, but maybe just

52:21

expand more, talk about money, talk about life,

52:24

you know, like, tell us the inside scoop

52:26

on what you're thinking about it. Yeah,

52:29

sure. I mean, money wise, I'm quite

52:31

lucky. I've had a good career. I've

52:33

got plenty of savings. My outgoings are

52:35

pretty low. Like it's just me. I've

52:37

got no kids, mortgage, whatever. So it's

52:39

a bit of rent and money

52:42

for my motorbikes pretty much it. But

52:46

yeah, so the outgoings are low. I had enough for

52:48

like at least a year. So I'm good there. Um,

52:51

I just figured that, you

52:53

know, have you ever had something that you

52:55

have to work on and like, if you don't,

52:57

it's going to bug you forever. Like it's one

52:59

of those things I figured I had to give

53:01

it a go. Otherwise I'd always wonder like, what

53:03

if I had tried? What if I had done this?

53:06

And like at the end of the day, it's maybe

53:08

a small risk, but I don't think doing

53:10

this will make me any less employable if it

53:12

doesn't work out. So it's something

53:14

I really wanted to at least

53:17

give a shot. You set parameters

53:19

around what success looks like, what

53:21

failure looks like, when you're going to call

53:23

it a day, if there is

53:25

one, or do you have a business model,

53:28

like how much did you put into it?

53:30

You're like, I just want to work

53:32

on this. I'm going to give it a shot. Yeah, I've got,

53:35

I was kind of trying to get

53:37

a nice balance between planning enough, but

53:39

also not making it too rigid because obviously

53:42

when you go from fitting something into your mornings and

53:44

doing something full-time, like a lot can change. So

53:47

the way I saw it was like the

53:49

first six months of the year wouldn't be

53:51

focusing too much on getting revenue. They'd be

53:53

focusing on making sure I'm building the right

53:55

thing, making sure I'm building something people want,

53:57

making sure that like any. sort

54:00

of frustrations people had with souls. And then the

54:02

latter six months of the year, I'd be thinking

54:04

more about like making sure I can pay rent. But

54:07

that's, that's the main plan. I'd kind of like to

54:09

start addressing some more team use cases later in the

54:11

year as well, but the first, again, the

54:13

first beginning is just focusing on making sure

54:15

like individuals are very, very happy with their

54:17

shell. And how long

54:19

have you been doing it? So full-time,

54:21

like six weeks. Otherwise

54:24

it's like three years now. The

54:26

first majority of that was very sporadic though.

54:28

Yeah. So six weeks full-time.

54:31

Are you having a blast or what? Yeah, it's been great.

54:34

It's cool because I'm, I don't know, I've had

54:36

loads of ideas I've wanted to explore, but I've

54:38

never had the time and the energy. So being

54:40

able to tie up so many loose ends from

54:42

last year, like as an example, the first

54:45

version of the sync that Atuin had

54:47

was like mostly good

54:49

for most use cases and it

54:52

lasted way better than I thought. I think

54:54

there was a comment from the first version

54:56

that was like, this is really naive, improve

54:58

it, and I never

55:00

got around to fixing it. So that was like

55:02

a prototype sync V2 that was sitting around for

55:04

all of last year. And in

55:07

the last, the first few weeks of full-time,

55:09

I managed to finish it and we

55:11

rolled it out a couple of weeks ago and working

55:14

so much better. And it's nice to have been able

55:16

to do that. I think that's like very important for

55:18

laying the foundations for everything else. What

55:20

are you doing to ensure you're building the right

55:22

thing? Like how do you, is it

55:25

just Discord? Is it emailing? Is it interviewing?

55:27

Like what do you do to ensure this? It's

55:31

a big mix. So Discord's been

55:33

great. That's been growing constantly. People give

55:35

me some really nice feedback there. I think

55:38

I can't stick to just Discord. So there's also a

55:40

forum now, which is mostly for support.

55:42

And that's been really good too. I've

55:44

been doing user interviews too. So I have like little video

55:46

calls with people and talk about like

55:48

what they like, what they don't like, what other

55:51

problems they might have, what they'd be willing to pay

55:53

for, what they might not want to pay for, that kind of thing.

55:56

That's been going pretty well too. Did you see

55:58

mountain bikes when you said to... Yeah,

56:00

I'm kind of obsessed. It's like my not computer thing

56:02

to do. Where do you live

56:04

at again? I live just outside of London. A

56:07

lot of good trails there, right? Kind

56:09

of. I mean, the UK is a bit mixed. It's always

56:11

raining? It rains a lot. We can ride in

56:13

the mud though. We can't ride in the mud around here. I

56:15

mean, it's... It sucks, but I like

56:18

it. Yeah, I mean, most cycles in the rain, they're

56:20

still fun, but it's very

56:22

different and it's like more dangerous. Yeah,

56:26

okay. We'll sidetrack that. I

56:28

wanted to check that because

56:30

I thought I heard you say motorbikes or mountain bikes. I

56:32

wasn't sure which one. So for other people who are in

56:34

the motorbikes, which I'm not, I'm asking this for them. What

56:37

motorbike do you have? Is that awesome?

56:40

I have a Husqvarna 701 Supermoto. Okay,

56:43

Husqvarna I'm aware of because they make

56:45

very good riding lawnmowers around these parts.

56:49

And chainsaws, apparently. Yes, chainsaws. And they're all just

56:51

three numbers, so you don't know what you're getting. It could be a chainsaw,

56:53

could be a sewing machine, could be a motorbike.

56:55

Did they make sewing machines too? They

56:58

do. Found that out recently. That's hilarious. Yeah. You

57:01

never know until the box shows up and you're

57:03

like, that's too small for a motorbike. It looks

57:05

like a sewing machine. Exactly. What

57:07

do you think the possibility is with A2N?

57:09

Like, what do you think you might be

57:11

going towards? I know this is

57:13

early days, six weeks full time at least.

57:16

But what do you think the... What's the big dream? I

57:18

think the big dream is to kind of hit all

57:20

the use cases, right? Like

57:23

I want power users to love using

57:25

it because it's customizable. It lets them get

57:27

what they want out of it. I would love

57:29

it if people who are less comfortable in

57:32

the command line and kind of freak out and

57:34

are a bit scared feel so much more

57:36

confident using it. And I would love for teams and

57:38

organizations to be using it just to accelerate their productivity,

57:42

to make it more comfortable. You mentioned, I

57:44

guess, warp to some point. to

58:00

some degree, a lot of

58:02

folks use terminal app on Mac or

58:04

pick your terminal flavor on Linux. I'm not even sure.

58:07

I don't use Linux desktop, so I can't say for

58:09

sure. It's still not the

58:11

year of Linux desktop, so there you

58:13

go. Maybe next year. Maybe next

58:15

year. It's always next year. It's

58:17

always next year. How does

58:20

H-WIN fit into the world

58:22

of warps, for example, where they want to supercharge

58:24

or be the terminal of the future and they're

58:26

trying to, there's

58:28

even our friends at, not

58:30

bubble tea, but what's their name,

58:33

Jared? Charm. Charm, yes. Y'all

58:36

are all wanting to improve the

58:39

command line. One

58:42

organization, project, dream at

58:44

a time, and I feel like there's a version

58:47

of you all that come together and just coalesce

58:49

into a great symphony or something like that. How

58:51

does it work with the warps and

58:53

the non-warps and the things that sort of add

58:56

things? Is it challenging to work around those things?

58:58

Like you mentioned, it doesn't work right. Describe

59:01

that. Yeah, I mean, I guess

59:03

it's different challenges. Warps are in

59:05

a nice place where they control

59:07

the sandbox. Everyone's using their terminal. They've got

59:09

the whole thing from the bottom up. And

59:12

the downside is there that they have to convince

59:14

someone to download and use a new terminal, which

59:17

I don't know anyone there, but I imagine they struggle with that. Whereas

59:20

the challenges that we have are

59:22

a little bit different in that

59:24

the friction to get started with

59:26

the two ends is quite low. It's just

59:28

a plugin to install into what you already have. But

59:31

the flip side there is that the

59:34

number of things it can conflict with

59:36

and the number of different setups people

59:38

have is like huge. And I

59:40

didn't realize when I started this, the number of very

59:43

weird installs some people would have,

59:45

like some strange terminal on

59:47

Windows that when you're SSH'd into a different

59:49

machine and have three specific plugins, it doesn't

59:51

work. And it's like, well, great. I never

59:54

really envisioned someone would do that. But let's

59:57

see how we can sort that out. Yeah, I'm

59:59

not sure that answered the question. Well, I think

1:00:01

just mapping around the, I guess the land of terminal,

1:00:03

which is like the Wild Wild West in a lot

1:00:05

of ways. There's no right

1:00:08

way or wrong way. It's just the way. And

1:00:10

there is no way to determine

1:00:12

this random SSH session into on

1:00:15

Windows that has these three plugins. You just can't predict that, right?

1:00:19

Exactly. So I think there's like somewhere in

1:00:22

the middle where everything converges and that's where

1:00:24

the answer is. But we'll see. So

1:00:27

you don't need much money. So I'm guessing you haven't raised

1:00:29

any money or have decided

1:00:32

to take on investment or do you

1:00:34

have a posse? Do

1:00:37

you have folks behind you rooting for

1:00:39

you? You have fans. I know that

1:00:42

because they hit us up on

1:00:44

the socials. And they're like, hey, everyone's awesome. The

1:00:47

support has been amazing, actually. I didn't really

1:00:49

realize that many people cared that much. So

1:00:52

that's been lovely, especially on Mastodon. Lots

1:00:54

of people are like, what on Mastodon? Why? Where?

1:00:56

And like every time I post about it

1:00:58

there, I get loads of nice feedback and support. So that's

1:01:01

lovely. Thank you to anyone listening to this. I

1:01:03

think funding wise, no, I don't have any

1:01:05

funding. Sort of consider a little bit of

1:01:08

angel money at the moment, maybe. But again,

1:01:10

I don't need loads of

1:01:12

money right now. The most expensive thing

1:01:14

is my time. And I'm good with

1:01:16

that for now. I guess I'm not

1:01:19

ruling it out for the future, but it's not a

1:01:21

thing in the present. I

1:01:23

don't think like if someone gave me

1:01:25

$2 million tomorrow, pretty much the only

1:01:27

thing I could do with it is hire people. And

1:01:29

I don't think that's the right thing right now. I

1:01:31

think it would be the wrong incentives at this point. Sure.

1:01:34

Is there anybody else working on it with you? Not

1:01:37

full time. Not like this. There's a

1:01:39

lot of open source contributors who are

1:01:41

amazing and I really appreciate them all.

1:01:43

It kind of varies. There's some regular

1:01:45

contributors and some of them have their

1:01:48

special sort of interest areas of what

1:01:50

they like to work on. And

1:01:52

there's a lot of drive-thly contributors too. So

1:01:54

someone will have something that annoys

1:01:56

them and then they'll come and fix it and then

1:01:58

vanish. on a sticker so

1:02:00

I mean I know it's not going to show up for the audio

1:02:02

but I got one of these and every

1:02:05

um... A turtle with a wizard hat on? It's

1:02:07

got a wizard hat on and it's sparkly. So

1:02:09

everyone can contribute. And the shell is kind of

1:02:11

a gem or something? Exactly. It's like

1:02:13

a magic shell. So

1:02:16

anyone that wants one I'll ship them out.

1:02:18

Nice. Yeah. I like that mascot. What's

1:02:22

the name? Et-et-et-et-et-et-et? It doesn't have

1:02:24

a name. I should

1:02:26

probably name it Et-et-et-et-et. Oh, like what does

1:02:28

Et-et-et-et-et mean? Why would you pick that name?

1:02:30

Oh, um, my favorite book

1:02:32

series is the sort of Terry Pratchett

1:02:34

Discworld universe. I don't know if you're familiar

1:02:36

with it. The

1:02:38

premise, it kind of comes from a lot of... The

1:02:41

premise is that there's like a giant turtle in

1:02:43

space and flying swimming around.

1:02:46

He has four elephants on his back and then the

1:02:48

world is like a flat disc on the back of

1:02:50

the elephant. It's kind of like a wacky fantasy series.

1:02:52

Okay. Oh, yes. And

1:02:55

the name of that turtle is the Great Ativan.

1:02:58

It kind of came from there and it's been

1:03:00

like adjusted slightly. But the concept comes through

1:03:02

like a lot of old religion and mythology of

1:03:04

like the world turtle. Gotcha.

1:03:07

Interesting. The world

1:03:09

turtle, the magical shell. Yeah,

1:03:12

I've heard some people say, I'm not sure what the

1:03:14

context was, but like we're all just on

1:03:16

a world on a turtle's back flying through space

1:03:18

or something like that. I've heard that at least

1:03:21

somewhere in a geek

1:03:23

land of some sort. I'm not sure

1:03:25

which. There's a big intersection between computer

1:03:27

nerds and Terry Pratchett fans. That makes

1:03:30

sense. Okay. Is there a tie-in to

1:03:32

turtles all the way down from that or no? No,

1:03:34

but it is fitting. I mean, there might

1:03:36

be, but not that I'm aware of. I

1:03:38

mean, the synapses connected in my brain, but

1:03:41

I don't know. I don't know the history of the

1:03:43

phrase turtles all the way down well enough. But

1:03:45

I just figured, you know, nerds and turtles, there

1:03:47

you go. Nerds and turtles. Well, does

1:03:50

the mascot, the icon, does it have

1:03:52

a name? No, it does not. I've

1:03:54

been debating. I suggest the name. Okay,

1:03:57

for it. Shelley. Oh, actually.

1:04:00

That's cute. Yeah, maybe it's

1:04:02

Shelly. Shelly. That's a pun, and

1:04:04

it's good. That's a good pun. Yeah,

1:04:07

I would dig it That's almost as good as

1:04:09

my TNT tie-in from the day. So good. Yeah

1:04:13

But not good so bad. It was good. I

1:04:15

like Shelly. That's actually good. Good. It's good good That's

1:04:18

not bad. Good. That's good. Good. So Ellie

1:04:21

which rhymes with Shelly. That's true, by the way in case

1:04:23

you didn't notice How can we

1:04:25

help how can change all the community our listeners? You

1:04:28

know fellow nerds terminal junkies Adam

1:04:32

and I like I'm rooting for you. I think

1:04:34

it's really cool I hope you

1:04:36

make it even cooler and that

1:04:38

you can do this, you know as long as you want

1:04:40

to so how can we help? How can get involved? What

1:04:43

would help you? I mean even just being here

1:04:45

and sharing it is huge So I think if

1:04:47

anyone who wants to help out could check it

1:04:50

out. Give it a try Let me know what

1:04:52

you think all feedbacks good even if it's not

1:04:54

obviously positive Sharing it with

1:04:56

people you think might be interested is also great.

1:04:58

So this is spread pretty much entirely through word-of-mouth

1:05:01

So continuing that would be fantastic, but

1:05:03

no just more of the same fantastic. Thank you

1:05:05

so much. Adam any final thoughts or

1:05:09

Lines of questioning not necessarily. I

1:05:11

think I think

1:05:13

the trust factor is Like

1:05:16

you got registered like in the whole

1:05:18

happy path of getting started is all

1:05:20

Registering and I think there's

1:05:22

an allergy Initially like even

1:05:24

with Jared on when we first talked to Zach

1:05:27

Lloyd with warp we talked on twice down the

1:05:29

podcast It was like we

1:05:31

don't want to have to register to use the terminal and

1:05:33

I get it I understand all this your project you get from

1:05:35

doing it and I get that we can trust you understand all

1:05:37

that I think just generally hackers are

1:05:39

like I would rather not so

1:05:42

I'd like a more squeaky or I

1:05:45

guess a more smooth path to the

1:05:47

non-register version to get a

1:05:49

win early on and Let

1:05:51

the the trust be earned over

1:05:54

the usage of the tool like that would be

1:05:56

feedback from me because that's That

1:05:59

would be my hurdle really And I know this

1:06:01

isn't feedback time, but that's my one

1:06:03

thought after talking to you and know what the

1:06:05

tool is and what it can do is like

1:06:07

that's I think probably a hurdle you're having to

1:06:09

deal with. And the server is

1:06:11

awesome, but I don't think somebody's going to stamp a server

1:06:13

just to like bifurcate the option to trust you. I think

1:06:15

it's just going to be like let's just figure out how

1:06:17

to get to a win with

1:06:20

it without having to do the whole let it

1:06:22

– let me put it in your cloud kind

1:06:24

of thing or whatever. However that looks. Well, what's

1:06:27

the third option? So if they're not going

1:06:29

to run their own server and you don't want them to register,

1:06:31

what could Ellie do? Well, you don't have to, right?

1:06:33

You could just sync it locally. Isn't it local if

1:06:36

you don't register? Yeah, you can use it offline. It

1:06:38

won't sync, but it's still useful. But that's the core

1:06:40

feature that it offers out. Well, you

1:06:42

get the usage of the tool. I think that's what's cool

1:06:44

for me is like, well, okay, now that I see this

1:06:46

tool is cool and how I can use it in this

1:06:48

one single environment, I know I want it

1:06:50

over there too and over here too. But in that way,

1:06:52

I actually want to opt in to sync later on. Give

1:06:55

me the tool and its usefulness and let

1:06:57

sync be a superpower that I

1:07:00

sought after after I trust you

1:07:02

because I don't need to have syncing

1:07:04

to enjoy stats immediately

1:07:06

as an example. That was my hurdle here in

1:07:09

the call was I just couldn't share that with you

1:07:11

all because I had my issues. I can

1:07:13

fix that and that's not your problem, but I

1:07:15

think ultimately you may have the

1:07:17

loss of somebody's attention in

1:07:20

that moment if they just – I can't figure it out. It's

1:07:22

not that big of a deal to me. Let me move on.

1:07:24

I think over time, there'll be a challenge. Yeah,

1:07:26

maybe you started with the stats and

1:07:28

the history, the actual control

1:07:31

R replacement as part of this setup

1:07:33

that you're working on. And at

1:07:35

the end of the setup was now

1:07:37

the sync option which requires either self-host

1:07:39

or login right now. Register.

1:07:42

I can see that because I think I

1:07:44

would potentially download it and use it to

1:07:47

get to the stats and just to replace

1:07:49

control R. If that was obvious

1:07:51

and good enough for me to be like, oh, I

1:07:53

love to control – I love control R on steroids

1:07:55

or whatever way you are able to

1:07:57

pitch that because I realize the steroids

1:07:59

are good. have also bad effects

1:08:02

on the body. So maybe not the best. Call

1:08:04

it on strawberries. Strawberries are great for you. They

1:08:06

give you energy. Oh, on strawberry. Yeah, antioxidants. Amazing

1:08:08

for you. Great for your skin. All right, so

1:08:10

control R on strawberries. Plus

1:08:12

the history, the stats might

1:08:15

get me to install it. And then by the way,

1:08:17

sync. Oh, wow. Yeah.

1:08:21

So I can see what you're saying there, Adam. Yeah,

1:08:23

I mean. But I think as you build out the

1:08:25

like your developer environment, sync everywhere,

1:08:28

a lot of people are going to come

1:08:30

for that exact reason. And they're going to go straight for it. For

1:08:32

sure. Because I think we

1:08:34

understand that a sync requires some sort of

1:08:37

quote unquote cloud computer. You know, like somewhere

1:08:39

that's not our computer. And so either registering

1:08:41

or hosting your own server makes

1:08:44

complete sense. Definitely. Thank you. No, I'll bear that in

1:08:46

mind. Whereas with warp, it was like it's a terminal.

1:08:49

I don't want to have to sign up and sign

1:08:51

in to use a terminal. But like it's a sync

1:08:53

service. This is, it has become the sync service. So

1:08:56

I get the reason. But

1:08:58

I agree. For sure. Thank you. I think I'll do

1:09:00

something like that for the like sort of

1:09:03

hand-holding setup to me, maybe. I would

1:09:05

wear. But I dig it. I'm excited for

1:09:07

you. I think Shelly is a hit. Turtle

1:09:09

in the back, you know, world's on a turtle's back, flying

1:09:12

through space. How do we get some of

1:09:14

those stickers? Yeah, let's get some stickers up

1:09:17

there. I see Sammy or dressed after this,

1:09:19

so I'll send you some. Fantastic. All

1:09:21

right. All right. This has been fun.

1:09:23

Thanks, Helly. Yeah, it's been great. Thank you so much.

1:09:26

Thank you. Thank

1:09:29

you. It took me a minute,

1:09:31

but I finally put my finger

1:09:33

on what the name Attuen was

1:09:35

reminding me of throughout this conversation,

1:09:37

Q Grand Moff Tarkin. I

1:09:40

grew tired of asking this, so it will be

1:09:43

the last time. Where is

1:09:45

the rebel face? They're

1:09:53

all down to me. Do you hear the resemblance?

1:09:56

Just me. Somehow, everything reminds me of

1:09:59

Star Wars. Okay, so

1:10:01

I've been using Atuin for a while now

1:10:03

and I don't think I'm ever going back

1:10:05

to the old built-in Control-R style. It's just

1:10:07

too good, and even though I primarily use

1:10:10

a single machine, I love the idea of

1:10:12

eventually using this to bootstrap my future shell

1:10:14

environments. What about you? Are

1:10:16

you a long time Atuin user who is excited? We

1:10:18

finally had Ellie on the Pod? Or is this your

1:10:21

first time hearing about it? Let us know in the

1:10:23

comments, we'd love to hear from you, there's a link

1:10:25

in your show notes. Or hop

1:10:27

in our community slack, there's great

1:10:29

conversations going on in there and

1:10:32

it's totally free to get yourself

1:10:34

an invite at changelog.com/community. Thanks

1:10:37

once again to our partners at fly.io, to

1:10:39

our friends at Sentry, who will give you

1:10:42

$100 off their team plan if you use

1:10:44

code changelog when you sign up, and to

1:10:46

Breakmaster Cylinder for Beat Freakin' For Us all

1:10:48

these years. Coming up next, homebrew

1:10:51

maintainer Mike McQuaid is back on

1:10:53

changelog and friends. We talk about

1:10:55

social media of course, the Apple

1:10:57

Vision Pro of course, Elon Musk

1:10:59

of course, homebrew and

1:11:01

workbrew, Mike's new business that's

1:11:03

building the missing features and

1:11:05

providing support for companies using homebrew.

1:11:08

Plus a few surprises along the way, I

1:11:10

think you'll dig it, it was a real

1:11:12

fun one. Alright that's all for

1:11:14

now, thanks for listening. If you dig this show,

1:11:16

share it with your friends, and if you really

1:11:19

dig it, hook us up with a 5 star

1:11:21

review. And if you really

1:11:23

really dig it, consider supporting our work with

1:11:25

a changelog plus plus membership. I hear it's

1:11:27

better, that's what people tell me at least.

1:11:29

It is better, it's been better for years.

1:11:31

Okay I'm done for real now, we'll talk

1:11:33

to you again, I'm free. and

1:12:14

a a