The Homebrew maintainers who built a startup - Mike McQuaid and John Britton from Workbrew

31 October 2024

Interviewed by Scaling DevTools

Mike McQuaid and John Britton are cofounders of Workbrew - a tool that gives you the missing features for enterprises running Homebrew.

Show transcript

0:00 Today's guests are Mike McQuaid and John Britton.
0:02 They met while working together at GitHub.
0:04 Mike and John are the founders of Workbrew,
0:06 a startup built to provide the missing enterprise features
0:10 for managing Homebrew.
0:11 Mike is the project lead and longest serving
0:15 maintainer of Homebrew, applause,
0:17 the package manager for Mac
0:19 that almost every one of you will have used.
0:20 I used it today to install Docker.
0:22 We get the inside story on Homebrew,
0:25 including its London pub origins.
0:26 I would never have imagined that it began in a pub in London
0:30 with a random chat.
0:31 I mean, I guess the naming and the beer theme
0:34 might have led one slightly more in that direction.
0:38 We learn what it's been like
0:39 to work with the biggest company in the world, Apple.
0:42 He's based in the Shetland Islands of Scotland.
0:44 If you look at a map, Shetland is far away.
0:47 Essentially, that's the short version.
0:49 Someone in Apple being like,
0:50 oh, we have to get this thing here.
0:52 Like, can it not just go to an office
0:54 or a bigger city or whatever?
0:56 We can't.
0:57 Like, if you want this to happen,
0:58 this is what you need to show out, right?
1:01 We learn how Mike and John
1:02 think about open source sustainability.
1:04 You will find various parts of the internet
1:06 that don't like me very much for this viewpoint.
1:08 But in the end, to me,
1:09 the only way you are able to successfully scale that
1:12 is you say some of these features go away.
1:13 And how Mike and John are taking these lessons forward
1:16 to build Workbrew with a very unique approach.
1:19 Enjoy.
1:20 Probably everyone listening has used Brew, Homebrew,
1:25 to manage their packages on Mac,
1:28 unless they're, like, really hardcore on Linux or Windows.
1:34 Do you want to talk a bit about the actual, like, story of Homebrew?
1:40 Because this is a tool that we all use.
1:42 So I just want to say something about the hardcore Linux users.
1:45 Is that I would say that, like, both Mike and I
1:47 kind of came to be as hardcore Linux users
1:50 and ended up in this situation of working on a Mac
1:53 because of its similarity to Unix operating systems.
1:56 But there was something huge missing there.
1:57 And I think that the story of Homebrew's origin and whatnot
2:00 that Mike will tell
2:01 is definitely, like, going to encompass that as well.
2:05 So, I mean, I guess as John said,
2:07 I almost mentioned this earlier.
2:08 But, yeah, for me, I realized recently
2:10 I think I was a, you know, Linux user for,
2:13 I don't know, like, five or six years primarily.
2:17 And at less than a year after moving from Linux to macOS,
2:20 I ended up maintaining Homebrew.
2:21 So that's, you know, how I ended up over there.
2:24 I've, my, occasionally when Linux people give me trouble,
2:29 I point out I have, like, one commit in Linux kernel
2:33 from back in the day, like, 2000 and whatever year it was.
2:37 So, yeah, that's my nice little Linux fight back thing that I have.
2:43 Linux credibility.
2:44 Yeah, exactly.
2:45 Also, fun fact, Homebrew itself runs on Linux nowadays as well.
2:49 So we might talk about that later
2:51 because it's sometimes a little bit confusing
2:53 why you would want to do that.
2:54 But anyway, so, right, we're back in 2009.
2:58 I'm working in London for a stock called Mendeley.
3:01 There's a guy called Max Howell
3:03 also working in London for another stock called Last.fm.
3:05 He has irritations with the way package management
3:10 is currently working on Mac.
3:11 He's been playing around with Mac ports and think and coming by.
3:16 There was one other option that, like, leaves my mind right now.
3:19 And I think he was whining about it in the pub one day
3:22 and basically someone said to him,
3:23 hey, if you hate package management and you know all this stuff so much,
3:27 why don't you make your own package manager?
3:29 I think mainly intending him to just shut up.
3:31 But he went off and started work on Homebrew.
3:34 It's interesting.
3:36 You can kind of go back and look through the Git logs.
3:38 It's like one of the kind of few examples, I guess,
3:42 of a big open source project that was started with kind of re-be-driven development.
3:45 So the first project commit in Homebrew itself
3:49 is like a description of how Homebrew is going to work.
3:51 And Mac wrote all that down before he wrote the codes.
3:54 And I often think, like, that's a sign
3:57 when things are going to be optimized for usability and developer happiness
4:00 as you think from the outside in.
4:02 Think, like, how should people be interacting with this
4:04 rather than just, like, writing code and then being like,
4:07 well, they interact with it how I tell them to.
4:09 Anyway, so he'd been working on this for a little while.
4:12 He open sourced it.
4:13 It started to get picked up by a few more people.
4:16 I discovered it because I basically built a sort of hack on top of Mac ports
4:20 to make it work in the way I didn't know at the time.
4:24 But this Homebrew tool worked to use a lot more stuff from the system
4:28 so it'd be kind of faster to build various tools.
4:30 I discovered Homebrew.
4:31 I was like, this is cool, and sort of just started getting involved.
4:34 And, yeah, I think one other non-Macs person
4:40 had kind of been getting involved as a maintainer at that time.
4:43 And, yeah, like, I guess, fast forward 15 years,
4:45 we've got kind of gone from, you know, a few contributors,
4:49 a few maintainers to about 30 maintainers.
4:52 We've got about, I think, 15,000 or something contributors,
4:56 maybe more than that, and 20,000 packages
4:58 kind of in our, like, officially supported ecosystem
5:01 and then probably tens of thousands more
5:03 in the kind of wider ecosystem.
5:04 That's so cool.
5:06 I would never have imagined that it began in a pub in London
5:10 with a random chat.
5:12 I mean, I guess the naming and the beer theme
5:15 might have led one slightly more in that direction
5:18 if you ponder it sufficiently.
5:20 Yeah, that makes more sense now.
5:22 Wow.
5:23 So, Mike, could you just tell me
5:27 what you just told me about WorkOS?
5:28 Because I just told Mike that WorkOS is the sponsor,
5:31 and this is what Mike said.
5:33 Yeah, so WorkOS isn't paying me any money for this.
5:36 I pay WorkOS money for this,
5:38 but WorkOS is, like, one of the best developer tools
5:41 I've, like, ever used.
5:42 It's the documentation and the experience
5:45 with building with them is so, so good.
5:48 Like, I initially was almost like,
5:49 okay, this seems expensive,
5:51 but then, like, I built an integration
5:54 with them in about 20 minutes
5:56 that I had spent two hours,
5:58 sorry, two days,
5:59 banged my head off the wall
6:00 trying to build it directly with Okta.
6:02 I then have, like,
6:03 many, many SSO providers,
6:05 like, support instead of just one.
6:08 So, yeah, like, for me,
6:09 WorkOS is one of the nicest developer experiences
6:11 I've encountered in the last, like,
6:13 five years, probably.
6:14 And it's still surprising
6:16 because a bunch of the developer team
6:17 are ex-GitHub
6:18 and therefore very good at their job.
6:20 Amazing.
6:21 And was it a case of, like,
6:24 build it and they will come?
6:26 I don't think so.
6:28 I mean, the funny thing in open source land,
6:30 back then we used to talk a lot more about this
6:33 than we do now when it's, you know,
6:35 there's a bit more of, like,
6:36 oh, clout and you should get some GitHub stuff
6:39 on your resume or CV or whatever.
6:41 Like, back then I remember people said
6:44 a lot of open source is just scratching your own itch, right?
6:47 Where essentially Max had a problem himself
6:50 that he wanted to solve primarily for himself
6:52 and then he was like,
6:53 okay, well, I'll put this on.
6:55 At the time, also quite new,
6:56 this thing called GitHub.
6:57 And I guess that was the other sort of interesting thing
7:00 and it's been interesting for me
7:01 kind of being on both sides of that
7:03 is, like,
7:04 Homebrew was also one of the first software projects,
7:06 certainly the first package managers
7:08 to kind of really lean in heavily
7:09 on the kind of GitHub way of doing things, right?
7:13 So Max, from the outset,
7:14 I guess he would say that he was being lazy
7:16 and I guess, you know,
7:17 maybe he was being lazy,
7:18 but he was also being smart.
7:20 The best engineers, in my opinion,
7:21 are often fairly lazy
7:23 because he had the attitude of, like,
7:25 hey, I don't want to maintain this all myself.
7:27 I want to have the community maintain this.
7:29 And if you want the community to maintain it,
7:31 you need to make getting involved,
7:33 merchant pull requests,
7:36 all this type of stuff,
7:37 like, as easy as possible.
7:38 So that was a big focus of his in the early days
7:41 and that was, I guess,
7:43 a big focus of mine increasingly
7:44 with Homebrew and GitHub and things like that
7:47 is that generally you want to take the behaviors
7:50 that you want people to do
7:51 in a developer tool
7:52 and make them as easy as possible.
7:54 And if the stuff you don't want people to do,
7:56 you want to make that either impossible
7:57 or harder for them to do
7:59 than the easy thing,
8:00 than the right thing.
8:01 That makes sense.
8:02 And what does that, like,
8:03 functionally kind of look like?
8:05 To me, in something like Homebrew,
8:07 it looks like,
8:09 I guess it depends on
8:10 what layer you look at, right?
8:11 So part of it is,
8:13 essentially,
8:14 we really try and optimize
8:15 the friction of, like,
8:17 onboarding, I guess,
8:18 both from a user level
8:20 and a contributor level
8:21 and a maintainer level.
8:22 So for a user,
8:23 Homebrew was, I think,
8:24 maybe one of the first projects,
8:26 maybe the first to the now infamous,
8:28 get a bash script
8:30 and pipe it into curl
8:32 that some people don't like very much.
8:34 And if you don't like that
8:36 then you might want to hear more
8:37 about Workbrew later.
8:38 We'll talk about that.
8:39 But I guess from that perspective,
8:42 a lot of these tools
8:44 that have installers,
8:45 having something where you can,
8:46 like,
8:47 get installed
8:48 a single command,
8:49 no dependencies,
8:50 everything gets done for you.
8:51 that was very heavily optimizing
8:54 Similarly for contributors,
8:56 the docs,
8:57 the automation,
8:58 the tooling,
8:59 everything like that,
9:00 such that the flow
9:02 as possible
9:03 for that person
9:04 getting started
9:05 with the project.
9:06 And when we see people
9:08 multiple times,
9:09 we're quite aggressive
9:11 about jumping on them
9:12 and really prioritizing
9:13 fixing that stuff
9:14 the project scales,
9:16 And we have still
9:18 very, very many more
9:19 than we have maintainers.
9:21 And then the last thing
9:22 is similarly doing that
9:23 for the maintainer approach
9:24 and they're making it easy
9:25 for us to add new maintainers
9:27 but also making it easy
9:28 to do their job,
9:29 trying to automate
9:30 the absolute living hell
9:32 out of everything
9:33 we possibly can,
9:34 trying to lean into,
9:35 robots,
9:36 automation,
9:37 CI all over the place,
9:40 linting all over the place
9:41 to try and essentially
9:42 just remove
9:43 the amount of manual work
9:44 that a human has to do
9:46 that can be done
9:47 by a robot
9:48 or a CI
9:49 or a GitHub action
9:51 or whatever it may be.
9:52 I mean,
9:53 what makes it possible
9:54 for such a relatively
9:56 small team
9:57 of volunteers
9:58 to run
9:59 a huge part
10:00 of the infrastructure
10:01 of Mac development.
10:03 unbelievable to me
10:04 how small but mighty
10:07 that group of people is
10:08 and they deserve
10:09 all the accolades
10:10 and all the support
10:12 for, you know,
10:13 as a developer myself,
10:14 I'm a user of Homebrew.
10:15 I use it every day.
10:16 I couldn't imagine
10:18 and it's just, like,
10:19 unbelievable that,
10:20 you know,
10:21 it's possible
10:22 to do something like that
10:24 you know,
10:25 small,
10:26 dedicated crew.
10:27 Yeah,
10:28 I mean,
10:29 it almost feels like,
10:30 I don't know if I ever
10:32 really thought about,
10:33 oh,
10:34 where did Brew come from?
10:35 It's, like,
10:36 like,
10:37 inherent part
10:38 of, like,
10:39 like,
10:40 developing with Macs.
10:41 such a,
10:42 did,
10:43 did, like,
10:44 Apple get involved
10:45 in it at all?
10:46 Not really.
10:48 Like,
10:49 they've helped us out
10:50 So we have
10:52 contacts at Apple
10:53 who we can kind of
10:54 escalate stuff to
10:55 and help us out there
10:57 and ask questions
10:58 I guess the most notable
11:00 case of Apple
11:01 helping us out
11:02 because obviously
11:03 they realized that
11:04 it was very much
11:05 in their self-interest
11:06 in that case to do it
11:07 was when there was
11:08 the Apple Silicon transition,
11:09 you know,
11:10 pretty much,
11:11 as soon as they
11:12 announced, like,
11:13 move to Apple Silicon
11:14 and there was these
11:15 and stuff like that,
11:16 they got in contact
11:17 with us straight away
11:18 and we're basically,
11:19 like,
11:20 of these kits
11:21 of your maintainers
11:22 and we also want to,
11:24 like,
11:25 get these kits
11:26 in data centers
11:27 so you can start
11:29 building,
11:30 building stuff in CI
11:31 and automating things
11:32 and packaging things
11:33 and integrate that
11:34 into your workflow
11:35 and you're not having to,
11:36 pay out the nose
11:38 for a lot of this stuff.
11:39 And even that process
11:40 was kind of an interesting
11:41 case of, like,
11:43 worlds colliding
11:44 there was conversations
11:46 with some of the Apple folks,
11:47 some of them knew already,
11:48 some of them would go,
11:49 so where's the homebrew office?
11:50 And it's like,
11:51 there is no homebrew office.
11:52 There is no homebrew office.
11:53 Okay, so what country
11:54 are your maintainers based in?
11:56 Like,
11:57 All of them?
11:58 Maybe, like,
11:59 20 countries, right?
12:00 And I remember, like,
12:02 one of the folks
12:04 who was very early
12:05 getting a lot of our stuff
12:06 sorted out
12:07 and he's a workbrew employee
12:09 nowadays as well
12:10 and he's based
12:12 in the Shetland Islands
12:13 of Scotland.
12:14 So, like,
12:15 if you look at a map,
12:16 Shetland is far away,
12:18 That's the short version.
12:20 And, yeah,
12:22 basically, like,
12:23 someone in Apple being like,
12:25 how, like,
12:26 we have to get this thing here.
12:28 can it not just go to an office
12:30 or a bigger city or whatever?
12:31 And it's like,
12:32 no, we can't.
12:33 This is, like,
12:34 if you want this to happen,
12:35 this is what you need to
12:36 sort out, right?
12:37 And it did all get sorted out
12:39 and it was all fine.
12:40 But it's,
12:41 it's definitely an interesting
12:42 cultural thing.
12:43 And I think Apple,
12:45 you know,
12:46 I'm a massive Apple fanboy.
12:47 I basically own
12:48 essentially almost every
12:50 thing they've ever made
12:51 with the exception
12:52 And I'll take one
12:53 if anyone wants to give
12:54 one to me for free.
12:55 But anyone's listening.
12:57 for my startup lifestyle
12:59 But, yeah,
13:01 you know,
13:02 But they definitely,
13:03 like,
13:04 don't quite get open source,
13:06 there's certain teams
13:09 and individuals
13:10 in the organization
13:11 who have that background
13:13 or relationships
13:14 or whatever.
13:15 organizationally
13:16 is, like,
13:17 they're not,
13:18 super tuned in
13:19 until almost,
13:20 how to partner
13:21 with a bunch
13:22 of volunteers
13:23 with no legal entities
13:24 scattered around the world.
13:25 Yeah,
13:26 it's actually really surprising
13:27 that they're not
13:28 giving bucket loads
13:29 of money
13:30 to you guys.
13:31 I don't know.
13:32 Well,
13:33 this is a lot
13:34 around open source
13:35 right now,
13:36 where it's like,
13:37 why?
13:40 I'm actually not
13:41 being too facetious
13:42 hopefully,
13:43 why should they?
13:44 They don't have to.
13:46 Humbro is operating fine
13:49 without getting buckets
13:51 from Apple.
13:52 And from both parties'
13:55 it would be hard
13:56 to justify going
13:57 from nothing
13:58 to buckets
13:59 without some sort
14:01 understanding
14:02 that relationship is,
14:04 of quid pro quo.
14:05 And,
14:06 you know,
14:07 that's when
14:08 you have
14:09 a volunteer
14:10 scattered around
14:11 and,
14:12 you know,
14:13 maybe the world,
14:14 I haven't checked
14:15 Close to the largest
14:16 company in the world.
14:17 in the world,
14:18 there's an imbalance
14:21 there.
14:22 also,
14:23 maybe it's pertinent
14:24 to maybe the way
14:25 right now,
14:26 where there's an awful
14:27 in and around
14:28 open source right now.
14:29 And I'm maybe
14:30 one of the old school
14:31 maybe even free
14:33 software people,
14:34 kind of the point
14:36 is they don't have
14:37 to give you anything.
14:38 I don't feel
14:39 like any company
14:40 who uses Homebrew
14:41 or relies on Homebrew
14:42 is obliged
14:44 to Homebrew.
14:45 a bunch of them do
14:46 and that's great
14:47 when they do
14:48 appreciate that
14:49 useful for us.
14:50 But,
14:51 if the license,
14:53 if we demanded
14:55 that every company
14:56 gave us money,
14:58 then in my mind,
14:59 we're not open
15:00 We're a different thing.
15:01 We're a proprietary
15:02 product with a free tier,
15:03 right?
15:04 My message would be like,
15:05 don't try to use
15:06 open source as a playbook.
15:07 You know,
15:08 build your product,
15:09 get people to use it.
15:10 Like,
15:11 the idea that,
15:12 open source is a marketing strategy
15:14 is something I just see
15:15 like way too often.
15:16 Yeah.
15:17 And I guess it depends
15:18 what you're optimizing for,
15:19 right?
15:20 If you're trying to
15:21 maximize reach,
15:23 then,
15:24 yeah,
15:25 make it open source,
15:26 put it under the most liberal
15:27 open source license possible,
15:28 the MIT license
15:29 let everyone in the world
15:31 But then converting
15:33 those people into paying
15:34 customers later
15:35 is not something
15:37 that you have built
15:38 your relationship
15:39 right?
15:40 And it's funny
15:41 because I guess
15:42 the open source
15:43 in some ways,
15:44 I think like open source
15:45 is a short-term
15:47 business model.
15:48 It's not a long-term
15:49 profit-making business model.
15:51 And it's one that
15:52 when these companies
15:53 get bigger
15:54 and zero interest rates
15:56 and investor pressure
15:57 and all these types
15:58 of things happen
15:59 and you get squeezed more,
16:00 then it's like,
16:01 you know,
16:02 to what,
16:03 are you able
16:05 to your,
16:06 like,
16:07 whoever it may be,
16:08 your community,
16:09 your customers,
16:10 your employees
16:11 hey,
16:12 all this open source
16:13 like,
16:14 turns out,
16:15 changed our minds.
16:16 Bad idea.
16:17 I don't know how to do that anymore.
16:18 yeah,
16:19 and I think John and I
16:20 share a similarly
16:22 bad taste in our mouth
16:24 from places
16:25 that have done that
16:26 where it's not even
16:27 doing the wrong thing.
16:28 It's just,
16:29 you cause a lot of aggro
16:31 that you don't
16:33 and that aggro
16:34 ends up feeling like
16:36 it is open source's fault
16:38 whereas I think
16:39 it's the exact opposite of that.
16:41 open source to be
16:42 something that it's not.
16:44 that makes sense
16:45 and yeah,
16:47 it's like,
16:48 it definitely is like
16:49 one of the biggest
16:50 conversations right now.
16:51 I think even
16:52 like the whole
16:53 WordPress stuff
16:54 is like
16:55 this kind of like
16:58 whether you should pay
16:59 things that you're not
17:00 obligated to
17:01 or whether you should
17:02 that you're not obligated to.
17:04 I mean,
17:05 to like looking at the origins
17:07 and looking at the licensing.
17:09 for me,
17:10 it's like
17:11 we have this internal
17:12 that we talk about
17:13 which is
17:14 don't give what you'll
17:15 later have to take away
17:16 and when you think
17:18 about open source,
17:19 like what you're really doing
17:20 is if you come from
17:21 like the free software
17:22 it's like the freedoms.
17:23 The freedom to copy,
17:24 the freedom to change,
17:25 the freedom to distribute,
17:26 you know,
17:27 all of those
17:28 like core tenants.
17:30 if you're starting
17:32 where you're telling
17:33 what you're giving them,
17:35 then of course
17:36 the internet
17:37 is going to be mad at you
17:38 when you change that.
17:39 Whereas like
17:40 from our perspective
17:41 the idea
17:43 you know,
17:44 building a business
17:45 around open source
17:46 is that
17:47 this is an open source project.
17:49 It's
17:50 open to
17:51 for anybody to do
17:52 right?
17:53 that doesn't have to change.
17:55 You can come along
17:57 and build a business
17:58 related to it
17:59 try to solve more problems
18:01 the same problem space
18:02 and
18:03 not put yourself
18:05 where your entire
18:06 business model
18:07 is like
18:08 hinging on the fact
18:09 that you need to
18:10 change something
18:12 about the way
18:13 your open source project
18:16 Like I think
18:17 I'm not actually familiar
18:18 with the
18:19 situation with WordPress.
18:20 like read over it
18:21 like very briefly
18:22 but
18:23 there's definitely
18:24 this case of
18:25 open source projects
18:26 there's like
18:27 a valuable service
18:28 and the business model
18:29 provide support
18:32 or the business model
18:33 is provide a hosted version
18:35 and to turn around
18:36 we're the only people
18:38 that can provide
18:39 is a big challenge,
18:41 And that's what a lot
18:42 kind of come down to
18:43 hey,
18:44 we invested all of this effort
18:45 an open source project
18:46 and now we're going
18:47 to like try to
18:48 retain for ourselves
18:49 the ability
18:50 to be the only one
18:51 as a hosted service
18:53 and that just seems
18:55 very challenging
18:57 to operate
18:58 and to execute
18:59 and it
19:00 doesn't really jive
19:02 with the expectations
19:03 want out of open source
19:05 and so
19:06 I just think
19:07 a flawed logic.
19:08 I think it's a much
19:09 to say
19:10 there's a huge
19:11 infrastructure
19:12 people make
19:13 get a lot of value
19:14 as it is
19:15 and
19:16 if there are
19:17 why not go build
19:18 those bits
19:19 additional bits
19:20 that are not
19:21 that are
19:23 kind of complimentary
19:24 I think this brings us
19:25 really easily
19:26 into the concept
19:28 building this
19:29 it's the missing pieces
19:31 Mike uses this analogy
19:33 of like the Lego set
19:34 if you think about
19:36 and you think about
19:37 the code
19:39 and the available modules
19:40 the standard bricks
19:41 you can go and build
19:43 with the standard bricks
19:44 but when you buy
19:45 that set
19:46 that's the Millennium Falcon
19:47 all these pieces
19:49 that are super custom
19:50 to your exact
19:51 use case
19:52 and those are the bits
19:53 that we build
19:54 right
19:55 you've got Chewbacca's
19:57 yeah exactly
19:58 yeah
19:59 they're not
20:00 the Chewbacca is too
20:01 specific to like
20:02 your company
20:04 you know
20:05 that makes total sense
20:07 and I guess
20:08 it's also reassuring
20:09 in the sense of like
20:10 the open source part
20:12 you
20:13 you couldn't
20:14 even if you wanted
20:15 to do like
20:16 a massive rug pull
20:17 and be like
20:18 homebrew
20:20 no longer
20:21 like
20:22 or whatever
20:23 because you
20:24 you're working
20:25 that I guess
20:26 homebrew is not
20:27 part of workbrew
20:28 it's something
20:29 that you're heavily
20:30 affiliated to
20:31 but you're not
20:33 you don't own
20:35 you're like
20:36 building on top
20:37 as well right
20:39 Mike should talk
20:40 like in a little
20:41 I mean for me
20:42 this sort of
20:43 like
20:44 why
20:45 why now
20:46 because
20:47 various people
20:49 including John
20:50 for the last
20:52 like you know
20:53 10 years
20:54 10 years
20:55 hey you should
20:56 go and make
20:57 like a work
20:58 a homebrew company
20:59 and I've been like
21:00 I don't like that idea
21:02 and I think like
21:03 for various reasons
21:05 but one of those
21:06 reasons was
21:07 the fact that
21:08 homebrew was in
21:09 a place that
21:10 there was a risk
21:11 have happened
21:12 10 years ago
21:13 that like
21:14 if a big
21:15 injection of VC
21:16 happened
21:17 to go into
21:18 something around
21:19 you know
21:20 in the homebrew
21:21 that you could
21:22 quite easily
21:23 capture that project
21:24 was so nebulous
21:25 right
21:26 whereas now
21:27 governance structure
21:28 like I'm
21:29 leader of
21:30 as was mentioned
21:31 am literally
21:32 elected to that
21:33 position every year
21:34 like
21:35 it just so
21:36 no one
21:37 against me
21:38 and no one
21:39 has ever been
21:40 elected except me
21:41 had that position
21:42 but that's
21:43 not a guaranteed
21:44 and if I
21:45 were to
21:46 really annoy
21:47 everyone involved
21:48 community
21:49 and all the
21:50 I would not
21:52 get re-elected
21:53 like someone else
21:54 and they would
21:55 and they would
21:56 homebrew
21:57 right
21:58 so
21:59 that structure
22:00 means
22:01 as well as
22:02 boring licensing
22:03 details stuff
22:04 I wouldn't
22:05 essentially
22:06 is its
22:07 it's a
22:08 and it
22:09 that way
22:10 forever
22:11 workbrew
22:12 may or may
22:13 like involved
22:15 with homebrew
22:16 like past
22:17 present future
22:18 whatever
22:19 the idea
22:20 that workbrew
22:21 could somehow
22:22 homebrew
22:23 and take
22:25 it's not
22:26 right
22:27 it's set up
22:28 that that
22:29 a tenable
22:30 in some ways
22:31 that I actually
22:32 like the most
22:33 and I don't
22:34 work through
22:35 so I'm not
22:36 paid to say
22:37 the relationship
22:38 git and github
22:39 was like
22:40 I think that's
22:42 look at like
22:43 like a big
22:44 has with
22:45 that's the
22:46 internally we
22:47 kind of jive
22:48 because it's
22:49 like was github
22:50 the development
22:51 of a bunch
22:52 in git
22:53 sure like did
22:54 github help
22:55 of git
22:56 but back
22:58 when like
22:59 and bitbucket
23:00 and various
23:01 of relatively
23:03 git hosts
23:04 were around
23:05 like github
23:06 trying to
23:07 like eliminate
23:08 or fork git
23:09 or like find
23:10 a way to
23:11 like everyone
23:12 who uses
23:13 git on their
23:14 needs to
23:15 us money
23:16 that was
23:17 thing but
23:18 I also think
23:19 the growth
23:20 of github
23:21 hand in
23:22 back and
23:23 being a git
23:24 bunch
23:25 of I guess
23:26 some of
23:27 like I
23:28 a bunch
23:29 git that
23:30 be there
23:31 not been
23:32 for github
23:33 working full
23:34 features
23:35 right and
23:36 the same
23:37 homebrew
23:38 some ways
23:39 the features
23:40 substantial
23:41 we're still
23:42 company
23:43 but there's
23:44 still stuff
23:45 in homebrew
23:46 which was
23:47 and would
23:48 there
23:49 not for
23:50 that we
23:51 in the
23:52 right and
23:53 for all the
23:54 use however
23:55 yeah that's
23:56 actually really
23:57 cool and I
23:58 the kind of
23:59 comparison I
24:01 ever heard
24:02 anyone even
24:04 make the link
24:05 between git
24:06 and github in the
24:08 oh like this is
24:10 you know like
24:11 people do it
24:12 like I don't
24:14 people complain
24:15 about like this
24:16 project is too
24:17 tied to this one
24:18 um I you
24:20 about git
24:21 github um which
24:23 is really cool
24:24 as well
24:25 thing you hear
24:26 people periodically
24:27 on the orange
24:28 site is people
24:29 being like hey
24:30 people can think
24:31 github are the
24:32 same thing and
24:33 they're not and
24:34 isn't that bad
24:35 right like and
24:36 it's you know I
24:38 don't think github
24:39 cares very much
24:40 about um about
24:43 about not being
24:44 associated one-to-one
24:45 with git because
24:46 they feel like that's
24:47 not necessary
24:48 for them right
24:49 grown a brand
24:50 which is bigger
24:51 than git right
24:52 and that's you
24:54 know I think the
24:55 way that both
24:56 parties handled that
24:57 admirable really
24:58 other other ways to
24:59 look at that too
25:00 similarities between
25:02 git and github
25:03 and between what
25:04 workbrew and
25:05 you know git is
25:06 kind of the
25:07 behind the scenes
25:09 like technical
25:11 enabler of the
25:12 product but you
25:14 know the stuff
25:15 github successful
25:16 it wasn't just
25:17 obviously there
25:18 was like timing
25:19 with git and
25:20 decentralized version
25:21 branching and merging
25:22 all that stuff but
25:23 like you know
25:24 and the pull
25:25 big you know the
25:26 big drivers of the
25:28 and I think that
25:29 it's similar you
25:31 that it's like
25:32 there's this
25:33 underlying architecture
25:34 that brew is the
25:35 software onto a
25:36 machine there's a
25:38 to do more than
25:40 you know just the
25:42 libraries that
25:43 there's a huge
25:44 opportunity for
25:45 teams to
25:46 have shared
25:48 environments and
25:49 do things that
25:50 don't make sense
25:51 player model and
25:53 model and so
25:54 you know we
25:55 the opportunity
25:56 add a lot of
25:57 complementary
25:58 functionality and
25:59 you know build a
26:00 business on that
26:01 another kind of
26:02 company that I
26:03 really admire in
26:04 this space of
26:05 it's not quite
26:06 open core like
26:07 open core right
26:08 like it's not
26:09 they control git
26:10 thing it's this
26:12 complementary style
26:13 and I'm not
26:14 sure if you're
26:15 company called
26:16 tail scale but
26:18 amazing kind of
26:21 I want to call
26:22 them like a VPN
26:23 not quite a VPN
26:24 it's a peer-to-peer
26:27 tunneling like
26:28 overlay network
26:29 tool but it's
26:30 built off of this
26:31 called WireGuard
26:32 and you know
26:34 people don't really
26:36 make the correlation
26:37 and WireGuard
26:38 unless they're really
26:39 nerdy about it
26:40 underlying enabling
26:42 technology that
26:43 they build kind
26:44 of a multiplayer
26:45 peer-to-peer
26:46 administration layer
26:47 on top of and
26:48 they run you know
26:49 all these like web
26:50 do like stun
26:51 tunneling and
26:52 they do you know
26:53 NAT traversals and
26:54 complicated networking
26:55 stuff that you
26:57 but really there's a
27:00 huge value in them
27:01 and what's what's
27:03 it's like all of
27:04 functionality makes
27:05 WireGuard feel like
27:06 magic and that's
27:07 their product and
27:08 it's like WireGuard
27:10 enabling technology
27:11 and I think that
27:12 you know for us
27:13 similarly Brew is
27:14 enabling technology
27:15 that's really awesome
27:16 and we want to
27:17 magic to teams
27:18 right bring it to
27:19 can say you know
27:20 get your engineers
27:21 up and running as
27:22 fast as possible on
27:23 day one make sure
27:24 makes a change to
27:25 the developer
27:27 environment it works
27:28 you're not repeating
27:29 yourselves over and
27:30 over again when
27:31 of date you can
27:32 easily when there's
27:33 a vulnerability you
27:34 away you know about
27:35 it you can you
27:36 know remediate you
27:37 an audit history of
27:38 what's going on like
27:39 all of this stuff
27:40 is really really
27:41 valuable from a
27:42 business perspective
27:43 and it's not
27:44 something that you
27:45 know the open
27:46 source community is
27:47 like eager to build
27:48 like I don't see a
27:50 lot of open source
27:51 engineers in their
27:52 what does this
27:53 billion dollar
27:54 company need to
27:55 with my open source
27:56 project let me spend
27:57 my time working on
27:58 it for free right
27:59 so we can build some
28:00 of those bits and
28:01 and I mean that was
28:02 in some ways the
28:03 earliest motivation
28:04 for work brew
28:05 really was like
28:06 we sort of decided
28:07 on this we're going
28:08 to build a commercial
28:12 layer around
28:13 homebrew before we
28:15 nailed down the
28:16 exact ideas but the
28:17 exact ideas came
28:18 from people in the
28:20 homebrew community
28:21 including one grumpy
28:23 scotsman who may or
28:24 may not be a very I
28:27 don't know whatever
28:28 adjective you want to
28:29 like I said no to
28:32 a lot of these
28:33 things over the
28:34 would come from a
28:35 big company and say
28:36 my compliance
28:37 requirements mean I
28:38 need to ABCD and
28:40 and it would be a
28:41 non-trivial lift for
28:42 the homebrew folks
28:43 and I'd be like now
28:44 we're not going to
28:45 periodically I would
28:47 other maintainers and
28:48 be like I'm just
28:49 no one's super
28:50 passionate about doing
28:52 they're all like no
28:53 don't want to do
28:54 this how about
28:55 project receives a
28:57 large company to
28:58 complete your
28:59 questionnaire yeah
29:01 exactly classic
29:02 example right but I
29:04 mean I found this
29:05 actually like a
29:06 previous job so like
29:07 Linux guy and I
29:09 worked on KDE like
29:11 the Linux desktop
29:12 environment and and
29:14 one of the things I
29:16 about back then I
29:17 worked for a company
29:18 called KDAB who
29:19 consultancy company
29:20 around QT that the
29:22 tech that and KDE
29:24 was built in and I
29:25 because they were
29:26 like the the biggest
29:27 contributor to KDE
29:28 of any like company
29:29 right so I joined
29:30 and I was like I'm
29:31 this open source work
29:32 gonna be sorry for
29:33 money and it's gonna
29:34 whatever and then I
29:36 learned that like a
29:37 lot of the open
29:38 source work for money
29:39 work around this
29:41 stuff is the stuff
29:43 right because all of
29:45 the fun jobs that's
29:46 the stuff that they
29:47 problems getting
29:48 volunteers in their
29:49 spare time to make
29:50 sure that they do
29:51 this but like a
29:53 essentially when you
29:54 have a spec and
29:55 you need to
29:56 hell up something to
29:57 make sure that like
29:59 you behave in the
30:00 same way as some
30:01 software on windows or
30:03 a bigger open
30:04 whatever and you
30:05 want to just like
30:06 nail that down such
30:07 that you completely
30:08 of the specification
30:09 right that's that's
30:11 not so fun right
30:12 like but it is
30:13 really valuable and
30:14 important to big
30:15 companies and that
30:16 they will pay for
30:17 people to do right
30:18 and in some ways I'm
30:19 also inspired by that
30:21 where it's previously
30:23 with those
30:24 organizations it
30:25 wasn't that I was
30:26 saying like we
30:27 don't that your
30:29 right it's your
30:31 they did have
30:32 problems and they
30:33 them but it's like
30:34 we can't solve that
30:36 we're not the right
30:38 people we're not
30:39 qualified well enough
30:40 and in some cases
30:41 when even when
30:43 people would come
30:44 pull requests to
30:45 this functionality
30:46 functionality it was
30:47 like well now no
30:48 one's using it so
30:50 it bit rots and
30:51 then a year later
30:52 someone tries to use
30:53 it again it
30:54 because homebrew
30:55 doesn't use it
30:56 right and 99.99%
30:58 of homebrew users
30:59 whereas now we
31:00 have a world with
31:01 workbrew with this
31:02 it's like we have
31:03 customers using this
31:04 stuff every single
31:05 day right so if
31:07 project has an
31:08 issue there we
31:09 fix it within
31:10 minutes or hours
31:11 right instead of
31:12 it being like a
31:13 year of this thing
31:14 just being broken
31:15 right and so
31:17 yeah I think I
31:18 different way of
31:19 doing this stuff
31:20 this stuff and I
31:21 think it's where
31:23 people are best
31:24 aligned right we
31:26 in workbrew are
31:27 best aligned to
31:28 solve these problems
31:29 for customers and
31:30 we do it with some
31:32 some proprietary
31:33 software and we do
31:34 some stuff for free
31:35 and some stuff for
31:36 money right and
31:37 whereas workbrew is
31:38 sorry homebrew is
31:40 best aligned to do
31:41 what suits the
31:42 community to kind
31:43 of coalesce behind
31:45 the stuff that works
31:46 for 99% of people
31:47 right and if you're
31:48 in some very
31:49 niche use case or
31:50 0.1% or whatever
31:51 then it's you know
31:52 it's probably not
31:53 good for you as
31:54 workbrew might be
31:55 this is also the
31:56 configuration or
31:57 simple defaults or
31:59 solving the case for
32:00 the majority of
32:01 people you know
32:02 there are a lot of
32:03 you know I'm not a
32:04 I'm not a maintainer
32:05 but I'm a contributor
32:06 to homebrew and I'm a
32:07 heavy user and there
32:08 are some things in
32:09 work the way I want
32:10 them to but kind of
32:12 the consensus is like
32:13 this is the way it is
32:14 because we need to
32:15 make life manageable
32:16 for the maintainers
32:17 if we changed it the
32:19 support load or the
32:20 burden of maintain
32:21 maintenance would be
32:22 much higher so we've
32:23 optimized for you know
32:24 the homebrew project
32:25 has optimized some
32:26 things for maintainers
32:28 rather than for users
32:29 in the sense that like
32:30 this always works it
32:31 never has a problem
32:32 we're never having to
32:33 support it and so
32:34 kind of the the idea
32:36 of the project the
32:38 open source project is
32:39 kind of the one true
32:40 way to use the thing
32:42 that applies to the
32:44 most people and is the
32:45 most reliable and the
32:46 most robust and the
32:47 easiest to maintain and
32:48 know every time a new
32:49 mac os version gets
32:50 released homebrew is
32:51 waiting right like that's
32:53 accomplishment for the
32:55 project but when it
32:56 comes to you know big
32:57 companies something that
32:59 we learned at our time
33:00 at github was that you
33:02 know a small company of
33:03 100 employees telling a
33:04 company of 100,000
33:05 engineers how to build
33:06 software is kind of
33:08 laughable the idea that
33:10 the 100 person company
33:11 knows how 100,000
33:12 person companies should be
33:13 running their software
33:14 teams and that you know
33:15 controls aren't
33:16 necessary right like
33:18 these these kind of
33:19 like very I don't know
33:22 very basic ideas right
33:23 and the reality is that
33:25 those really big
33:26 companies or you know
33:28 the companies that have
33:29 requirements they get a
33:31 lot of value out of the
33:32 things to have checkboxes
33:34 and have options and to
33:35 be able to say this is
33:37 we make this tool you
33:39 know fit our working
33:40 style rather than having
33:41 to conform to what's
33:42 there and so another
33:44 took away from our time
33:45 make github and what
33:47 we kind of emphasize
33:48 that you know the idea
33:51 of configurability is
33:53 kind of a debt that you
33:54 have to maintain and a
33:56 burden and you know the
33:58 open source project is
33:59 not really willing to do
34:00 that in in some cases
34:01 but for us that's
34:03 willing to sell we're
34:05 to maintain to make this
34:07 thing work for you know
34:08 various sizes of
34:09 companies with different
34:10 controls and security
34:11 and compliance requirements
34:13 and all that kind of
34:14 stuff and make that our
34:16 value add to make it fit
34:18 your needs and so I
34:19 think it's just like a
34:20 great opportunity to work
34:21 with them work on an
34:23 open source project but
34:24 model and it's not
34:26 really open core it's
34:27 not really like an
34:28 open source company it's
34:29 more of like a
34:30 complementary business to
34:31 an existing open source
34:32 company or an existing
34:34 open source project I
34:36 comes back in some ways
34:37 to the title of the
34:38 podcast right this is
34:40 scaling dev tools right
34:42 and in some ways there
34:43 is the the open source
34:45 way and there is the
34:47 the business way and I
34:48 think where it gets
34:50 messy or people get
34:51 confused is when they try
34:53 and combine the two so
34:55 about sustainability and
34:57 open source nowadays right
34:58 and I have been you know
35:00 moderately critical on the
35:04 internet and more loudly
35:05 critical in person at
35:07 that about the idea that
35:09 essentially the way you
35:10 fix open source
35:11 sustainability is just
35:12 throw money at every
35:13 that fixes all the
35:14 problems right it doesn't
35:15 it doesn't I've seen it
35:17 in hubbrew and I've seen
35:18 it in many other
35:19 projects and many other
35:20 maintainers who are in my
35:21 inbox like trying to
35:23 done right the way in my
35:26 experience the way you
35:27 sustainability is by
35:28 figuring out what
35:30 sustainability looks like
35:31 open source project right
35:32 and some of that comes
35:34 from personal boundaries of
35:35 the maintainers with each
35:37 other with the community
35:38 personal boundaries with
35:39 maintainers in terms of
35:40 like what what are we
35:41 doing here what are we
35:42 trying to build right are
35:43 we gonna do the Jamie
35:45 Zawinski rule of like
35:46 every all software grows
35:48 client right or are we
35:49 gonna say no this is the
35:51 line beyond which we
35:52 don't we don't go we're
35:53 not interested that's out
35:54 of scope use this other
35:56 project whatever it may be
35:57 and that's how you have to
35:59 do this because in you
36:00 know 10 years ago
36:01 homebrew had probably like
36:02 half the number of
36:03 maintainers maybe a third
36:05 uh but we had like maybe
36:07 one like one tenth if not
36:10 much less than that of the
36:11 users right and the way
36:13 you scale doing that as
36:14 John kind of mentioned
36:15 and alluded to as well is
36:16 you do that by filing off
36:18 the rough edges right and
36:19 if you have a feature that
36:20 blows up for users 10% of
36:22 the time right there might
36:23 be a bunch of used power
36:24 users who love that
36:25 feature and it makes their
36:26 life really happy and the
36:28 workflows are great but like
36:29 if 10% of users who bump
36:31 into that even they have
36:32 like 0.1% of those
36:33 users go and complain on
36:36 our discussions or like
36:38 in the past when we had
36:39 IRC channels or in our
36:40 issue track or whatever
36:41 that gets very very hard
36:43 for 30 people to manage
36:44 right and the only way in
36:46 the end like I've been
36:48 you know some you will
36:49 find various parts of the
36:50 internet that don't like
36:51 viewpoint but in the end
36:53 to me the only way you
36:54 are able to successfully
36:55 scale that is you say
36:56 some of these features go
36:57 away right if you want to
36:59 build a plugin or support
37:00 your own way of doing
37:01 things over here that's
37:02 fine but in the open
37:04 source project we do not
37:05 have the infrastructure to
37:07 do this stuff like we
37:08 can't do it and if you
37:10 there's that doesn't
37:12 solve the problem either
37:13 because you end up in
37:14 this I should probably
37:16 name for it but like
37:17 this sort of open source
37:18 sustainability money-wise
37:20 gulf of like too much
37:22 money for stickers not
37:23 enough money to pay
37:24 anyone right where like
37:26 if you don't have a
37:27 quite even one person's
37:29 full-time salary coming
37:31 it reliably but you've
37:32 got way too much like
37:34 what do you do with
37:35 that and how do you use
37:36 that to somehow magically
37:37 scale the project so
37:38 that everyone is able to
37:39 do more with less or the
37:42 same amount right but I
37:43 think on the business
37:44 side that's when it gets
37:45 can go and build a
37:46 business where if you
37:48 most businesses well no
37:51 most businesses certainly
37:52 are business you start
37:53 from the outset with a
37:54 model of looking at
37:56 being like well
37:57 ultimately we are going
37:58 to have to build
37:59 something and we're
38:00 people money for a
38:02 product and exchange for
38:03 that product like they
38:04 are going to get a great
38:06 user experience that they
38:07 don't otherwise get
38:08 right and to me like
38:09 that's another way that
38:10 you build sustainability
38:11 both for open source and
38:13 for dev tools in general
38:14 right like github as a
38:17 you know if you look at
38:18 the offering on github
38:19 nowadays and how many
38:20 billions of features
38:21 they have and ci runners
38:23 thing you couldn't do
38:24 that with a bunch of
38:25 volunteers like you can
38:26 do that with 30
38:27 volunteers working
38:28 right like maybe their
38:31 2007 or 8 or whatever
38:34 it was you could have
38:35 done that you maybe
38:36 that indefinitely with
38:37 like a small number of
38:39 people just doing things
38:41 but you wouldn't scale
38:43 that to you know like
38:45 hundreds of millions of
38:46 users for sure like
38:47 that's all of that stuff
38:49 requires many many
38:50 people being having a
38:52 level of dedication that
38:53 is not really
38:55 compatible with them
38:56 having a full-time job
38:58 yeah it may it makes
39:00 uh it makes sense what
39:01 you're saying there
39:02 about like um you know
39:05 how you unless you're
39:06 paying people the full
39:07 salary like money is not
39:09 really going to help
39:10 like do that and that
39:11 you do need a company
39:13 to go and like give
39:15 those like those extra
39:17 features that maybe not
39:18 so many people need that
39:20 cause a lot of problems
39:21 you know you're you're
39:22 willing to do all the
39:23 boring fiddly things
39:25 that big companies need
39:28 and that is not fitting
39:30 for open source it makes
39:32 so much sense this is
39:33 very cool i honestly like
39:35 i hadn't really seen
39:37 this sort of model or
39:38 like really thought about
39:39 this maybe it's probably
39:40 the right way i'd seen
39:41 it i guess but not
39:43 thought about it um it's
39:45 very cool and it seems
39:46 like you're not gonna have
39:48 to worry about really any
39:50 of the kind of issues
39:51 that a lot of open
39:52 source companies have
39:53 with like monetizing
39:55 uh the project um it's
39:59 very cool and we're kind
40:02 but i just wanted to ask
40:03 like who who are you
40:05 like kind of selling to
40:07 it's like it departments
40:08 or how yeah so there
40:10 there are really three
40:12 you know types of people
40:13 that are interested in
40:14 in work brew um
40:15 obviously first and
40:16 foremost is the
40:17 developer um homebrew is
40:19 incredibly popular with
40:20 developers on max and
40:22 at the minimum there are
40:25 developers out there that
40:26 work at companies where
40:27 they have security and
40:28 compliance requirements
40:29 that prevent them from
40:30 using homebrew you know
40:32 imagine trying to do your
40:33 work without access to
40:34 homebrew how much less
40:35 productive you would be
40:36 um so really one of the
40:39 core you know values that
40:40 we provide is just the
40:41 ability for engineers that
40:43 these security compliance
40:44 requirements to have
40:45 access to the tools that
40:46 they already know and
40:47 love and use um in
40:49 addition for engineers we
40:51 aim to you know make the
40:53 process better help them
40:55 get up and running faster
40:56 um and really make it so
40:58 that it's a treat your
41:02 local development
41:03 developer environment like
41:04 its infrastructure and the
41:06 idea being that you know i
41:08 have worked in vms i've
41:10 containers i've worked in
41:11 cloud idees and they all
41:13 have a trade-off in that
41:15 you get consistency speed
41:16 of startup all those
41:17 types of things in
41:18 exchange for worse
41:20 ergonomics i love working
41:23 on my local machine with
41:24 my local editor on my
41:25 local file system with my
41:26 local network with all the
41:27 things that are ergonomic
41:29 and the way that i like
41:30 them and anytime i
41:31 introduce a docker or vm
41:33 or cloud id or anything
41:34 like that into my
41:35 workflow it drains me it
41:37 makes me less productive
41:38 in addition all of those
41:40 options cost more they're
41:42 either you're paying for
41:43 vms you're using more
41:45 battery life using more
41:46 processor you're doing
41:47 whatever you have these
41:48 really amazing apple
41:50 silicon devices you're
41:51 dollars pop for and
41:53 then you're shelling out
41:54 into a cloud environment
41:55 like what what are we
41:56 even doing it doesn't
41:57 make any sense so that's
41:59 the story for the
42:00 developers it's like have
42:01 access and have really
42:02 great ergonomics the
42:04 next big category of
42:05 people is the it managers
42:06 so these are the folks
42:07 who in your company
42:08 they're responsible for
42:09 issuing you a laptop and
42:11 making sure that you're
42:12 up and running and for
42:13 engineers oftentimes
42:14 these people you don't
42:15 need to interact with
42:16 you can self-serve you
42:17 can fix your own
42:18 it department at github
42:19 was amazing the people
42:21 on that team you know
42:22 were there to support
42:23 you when you needed
42:24 them they helped you
42:25 situation they helped
42:26 you with getting access
42:27 to things but at the
42:29 had the situation where
42:30 we had a lot of talented
42:31 engineers who knew what
42:32 they were doing so it
42:33 where they had a lot of
42:34 trust and there was
42:36 make sure things that
42:37 were necessary got done
42:39 and so as it managers
42:40 we want to give them a
42:41 way to lighten their
42:42 workload and be able to
42:44 get brew into the hands
42:46 of their team in a
42:47 reliable repeatable
42:49 really easy way and then
42:51 once it's out there know
42:52 what's going on have
42:53 visibility into how are
42:54 people using this thing
42:55 do i need to be worried
42:57 that they're installing
42:58 some software that could
42:59 have a supply chain
43:00 attack or things like
43:01 that and then the last
43:03 people that were you
43:05 know trying to serve are
43:06 whether it's a CISO
43:08 or you know kind of
43:09 somebody who's on the
43:10 AppSec side but they're
43:12 they're interested in
43:13 is secure we're building
43:15 a bunch of functionality
43:17 around making sure that
43:19 compliance standards so
43:21 you know there's SOC 2
43:22 there's ISO 27001
43:24 there's you know all the
43:26 different kind of credit
43:28 card compliance like
43:29 there's there's all
43:30 different industries have
43:31 their compliance standards
43:32 and inevitably they
43:34 people are using
43:35 whether it means you
43:37 know the people on
43:38 their devices are using
43:39 admin accounts rather
43:40 than the standard users
43:41 and what what's entailed
43:42 interacts with that or if
43:44 it's about what pieces of
43:46 integrated into their
43:47 workflows so you know
43:49 we're serving those
43:50 people by saying hey
43:51 level overview of
43:52 happening you can get
43:54 alerts about when there
43:55 are vulnerabilities you
43:56 vulnerabilities you can
43:57 set policies and
43:58 guidelines to make sure
43:59 that the usage of these
44:01 tools are contained one
44:03 of the really powerful
44:04 features of homebrew is
44:06 that it's not tied to
44:08 one library of packages
44:10 they're the official
44:11 packages in homebrew core
44:12 and homebrew cask but
44:14 you can actually add a
44:15 third-party library called
44:16 a tap and install any
44:18 software homebrew you can
44:19 create your own taps you
44:20 can use taps by third-party
44:21 vendors and that
44:23 represents for security
44:25 folks a really really big
44:27 issue they're like oh my
44:28 god literally anything
44:29 could be installed in the
44:30 machines I have no idea
44:31 about it and so you know
44:33 for them we're making it
44:34 so you can see are
44:35 people using third-party
44:36 libraries third-party
44:37 taps which which ones are
44:38 they using do you have a
44:40 control policy that you
44:41 need to implement so
44:42 stuff around there that's
44:44 really valuable and so
44:45 you know those are the
44:46 people we're trying to
44:47 serve and I think that
44:48 kind of the last thing
44:49 I'll say about this is
44:50 kind of the business
44:51 model and like how
44:52 this so we have pretty
44:55 much two two stories
44:58 version and one there's
44:59 a paid version so
45:01 workbrew is available
45:02 for free for unlimited
45:03 users unlimited devices
45:05 it's the best way to
45:06 install brew on a fleet
45:08 whether you have 10 or
45:09 10,000 devices you can
45:11 go to workbrew you can
45:12 install brew on every
45:13 device have it be
45:14 consistent get it ready
45:15 minutes and have full
45:17 visibility into everything
45:18 there this is kind of
45:20 like table stakes for
45:21 any company that's doing
45:22 any kind of sock to
45:23 compliance or any kind
45:25 of you know good
45:26 security practices and we
45:27 made that totally free
45:28 then beyond that if
45:31 security controls or
45:33 if you want to add
45:34 remote management or
45:35 if you want to have
45:36 any kind of interaction
45:37 around automation and
45:39 default packages and
45:40 kind of environment
45:41 setup then we have a
45:42 paid plan that you can
45:43 get additional features
45:44 from but it's really
45:46 targeted those those
45:47 three use cases
45:48 very cool very cool
45:50 kind of reminds me of
45:51 like sneak in a sense
45:52 of like the kind of
45:53 model that if you've
45:55 thought about it that
45:56 way but like I can
45:57 totally imagine like
45:58 I've I've had to use
46:00 sneak at work because
46:03 we went through
46:04 sock to compliance and
46:05 it was like just like
46:08 my manager loved it
46:10 because you know you
46:11 can just everything is
46:13 like you know compliant
46:17 bits in with vanter and
46:18 stuff like that and
46:20 then for me it was
46:21 like generally good
46:22 because you know a lot
46:25 of the time you just
46:26 press a button upgrade
46:28 and stuff rather than
46:29 like solving it
46:30 yourself but um yeah
46:32 that's very cool it's
46:34 uh very very exciting
46:36 um yeah and thank you
46:38 so much for sharing the
46:40 journey and um yeah
46:42 and thanks uh thanks
46:43 mike for helping us
46:44 download all the stuff
46:47 we need onto our max
46:48 so um yeah thank you
46:51 john and mike um and
46:53 thanks everyone for
46:54 listening uh so workbrew.com
46:56 if people want to
46:57 learn more check out
46:58 workbrew.com and you
46:59 know thank you so much
47:00 podcast we're really
47:01 grateful to be here and
47:02 have this conversation
47:03 you know look forward to
47:04 coming back at some
47:05 point in the future and
47:06 giving an update amazing
47:07 yep thanks jack and
47:09 listening