Oxide and Friends

Oxide and Friends Twitter Space: September 13th, 2021

Docker, Inc., an Early Epitaph
We’ve been holding a Twitter Space weekly on Mondays at 5p for about an hour. Even though it’s not (yet?) a feature of Twitter Spaces, we have been recording them all; here is the recording for our Twitter Space for September 13th, 2021.
In addition to Bryan Cantrill and Adam Leventhal, speakers on September 13th included Steve Tuck, Tom Lyon, Dan Cross, Josh Clulow, Ian, Nick Gerace, Aaron Goldman, Drew Vogel, and vint serp. (Did we miss your name and/or get it wrong? Drop a PR!)
Some of the topics we hit on, in the order that we hit them:
  • Topic: Scott Carey’s article How Docker broke in half
  • Andrej Karpathy’s tweet showing InfoWorld.com spamming ads
  • Carey talked to:
  • [@5:21](https://youtu.be/l9LTJdT0sZ8?t=321) Akira Kurosawa’s 1950 Rashomon ~90mins. Watch a 2min trailer
  • Box office bomb “The Hottie and the Nottie” movie. Other stinkers: Gigli, Gotti
  • [@9:31](https://youtu.be/l9LTJdT0sZ8?t=571) Jerry Kaplan’s 1996 book Startup: A Silicon Valley Adventure
  • Steve’s take on commercialization > Bryan: There’s no question that they hit on something very big. > We saw a container as an operational vessel, but we failed to see > a container as a development vessel.
  • [@14:36](https://youtu.be/l9LTJdT0sZ8?t=876) dotCloud (PaaS) struggles to find a buyer; ultimately open sources as last resort > All of a sudden a company that nobody had heard of, > was a company that everybody had heard of.
  • They took too much money.
  • [@17:40](https://youtu.be/l9LTJdT0sZ8?t=1060) Pitfalls in raising money and scaling sales by imitating big companies
    • HBO’s Silicon Valley
      • Clip ~1min with Jan the Man, Keith, and Doug (I’m shadowing Keith) > Everybody should be spending time arm in arm with customers understanding > how is this technology going to solve a problem > which they’ll want to pay to have a solution.
  • Tom: Was there actually a business anyways? Or was it just technology?
  • What if developers are attracted to those things they know cannot be monetized?
  • There was this belief that if a technology is this ubiquitous, it will be readily monetizable.
  • [@27:26](https://youtu.be/l9LTJdT0sZ8?t=1646) Docker Swarm and Kubernetes > Hykes: We didn’t work at Google, we didn’t go to Stanford, > we didn’t have a PhD in computer science.
  • Stinemates: (The Kubernetes team) had strong opinions about the need for a service level API and Docker technically had its own opinion about a single API from a simplicity standpoint. We couldn’t agree.
    • DockerCon 2015: No mentioning Kubernetes! 
      • Brendan Burns’ talk “The distributed system toolkit: Container patterns for modular distributed system design” was unfortunately made private by Docker sometime in the last two years. The internet archive only has this. Burns wrote a blog post about the topics from his talk.
    • rkt (“Rocket”), CoreOS
  • [@36:11](https://youtu.be/l9LTJdT0sZ8?t=2171) Docker coming to market
    • Enterprise teams wanted support
    • Initial support offerings were expensive and limited (no after hours, no weekends) > Bryan: I floated to Solomon in 2014: run container management as a service.
    • Rancher Labs, K3s (lightweight kubernetes)
    • People care about GitHub stars (for better or worse)
  • [@48:02](https://youtu.be/l9LTJdT0sZ8?t=2882) Monetizing open source technologies
    • Triton implementing the Docker API
    • The support relationships are the foothold to figure out the product.
  • [@54:36](https://youtu.be/l9LTJdT0sZ8?t=3276) Venture capital going into Docker
  • Product market fit
    • Acquisitions
  • [@1:04:42](https://youtu.be/l9LTJdT0sZ8?t=3882) Could the outcome have been materially different?
    • Who made money on Docker? Cloud companies? Developers?
    • VMware acquires Heptio
    • Who invented containers? 
      • BSD Jails, Plan9 namespaces?
    • Tyler Tringas’ post about how small teams can create value with little outside investment, as a result of the Peace Dividend of the SaaS Wars.
If we got something wrong or missed something, please file a PR! Our next Twitter space will likely be on Monday at 5p Pacific Time; stay tuned to our Twitter feeds for details. We’d love to have you join us, as we always love to hear from new speakers!

Creators & Guests

Host
Adam Leventhal
Host
Bryan Cantrill

What is Oxide and Friends?

Oxide hosts a weekly Discord show where we discuss a wide range of topics: computer history, startups, Oxide hardware bringup, and other topics du jour. These are the recordings in podcast form.
Join us live (usually Mondays at 5pm PT) https://discord.gg/gcQxNHAKCB
Subscribe to our calendar: https://sesh.fyi/api/calendar/v2/iMdFbuFRupMwuTiwvXswNU.ics

Speaker 1:

So had alright. So are we still waiting? What are we doing?

Speaker 2:

No. No. We'll do whatever you want. It's 5 o'clock now.

Speaker 1:

Let's see. It's 5 o'clock now. Okay. So alright. There there we go.

Speaker 1:

Somewhere. It's 5 o'clock somewhere. So alright. This, and definitely I mean, as always, folks wanna jump in, ask to speak, especially folks that have, we're gonna be talking about Docker, for those of us who are veterans of the container wars. And, had you I mean, you obviously read this article when it came out, Adam.

Speaker 2:

That's right. But, you know, I think you pointed me to it or, but it it kind of floating floated across my my eyes. But one of the things that was interesting is like this the same the same journalist had written some stories recently about Docker and its impact. And this one took a real negative turn obviously.

Speaker 1:

So were were the other stories I mean, the when you say negative turn, do you mean that you mean the headline how Docker broke in half? I mean, that seems like

Speaker 2:

Yeah. I mean, I I

Speaker 1:

Are you accusing that I did like, by the way, did you see someone tweeted out, like, they took a screen cast of them going to the InfoWorld? Did you did you see this? No. Oh, god. It was actually very funny.

Speaker 1:

It's a screen cast of them going to the InfoWorld site, and all of the pop ups just, like, exploding all over them. It's like a like, this is like the Internet circa 2021.

Speaker 3:

It's pretty funny.

Speaker 4:

Yeah.

Speaker 1:

Yeah. How Docker broke in half as the headline. It's definitely not it does it does not have better just law of headlines. There's no question in there. There's not like, is this how Docker broke in half?

Speaker 1:

It's like, here it is.

Speaker 2:

Yeah. And, yeah. I mean, I just some of the there are other ones we're talking about the innovation of Docker and and its significance. And then a little bit of a negative turn on its recent, licensing changes, which I hadn't heard about. And then I went to upgrade my Docker and it said, oh, by the way, big licensing changes just just before you click upgrade.

Speaker 1:

It it I guess it does have to tell. Maybe it has to be the licensing changes. It's just a shame that it, like, trolls you into a licensing. It's like, I'm getting into a licensing discussion with my upgrade software. Like, this kinda sucks.

Speaker 2:

Yeah. Yeah. Yeah. Previously, I had seen it pop up and say, hey, if you wanna downgrade, just so you know, that's a paid feature. And I thought, that's a funny paid feature.

Speaker 1:

That is a funny paid feature. No. You can go back. You can

Speaker 2:

definitely go back. It's just gonna cost you. Yeah. Right. Be a shame if something happened.

Speaker 1:

The so what was your take on reading the article?

Speaker 2:

I mean, that confirmed, I guess, my perceptions to be clear from the outside. And as Brian said earlier, like, we'd love to hear from folks with their perspectives, inside or outside or whatever. But, certainly, like, the the hype wave felt just enormous in, I guess, like, probably, like, 2012 or 2013 feeling like the the sort of this impossibly like, like, I I couldn't see how they could achieve the amount of hype that they had attained.

Speaker 1:

Yeah. 2014 is when it really hit fever pitch, I think.

Speaker 2:

Is that when they they started accumulating capital?

Speaker 3:

Like, it

Speaker 1:

was going out of style? Yes. Many rounds of capital over the the course of 2014. And the and I guess I think maybe it's worth saying also too at the kind of the top in terms of, like, the value. I mean okay.

Speaker 1:

We are interested in history. We are interested in failure. So we are, of course I am always interested in what happens to companies and why. I think that just to I I don't think we're necessarily, I think it's we're trying to reason about it. We're trying to learn from it.

Speaker 1:

How can we not make some of the how can one not make some of the mistakes that Docker made? Because I think that some of the mistakes that Docker made are preventable. And then some of them are just like, this is you know, it's hard to know. Yeah. Totally.

Speaker 1:

K.

Speaker 2:

I So what what what was your I mean, I know that I you had you had kind of been even more in the Docker world than I had. So a lot of this must have felt very familiar as you read through that article.

Speaker 1:

Well, so I thought the article was very sourced. So I thought that it was I was impressed by they talked to Ben. They talked to Solomon. They talked to Craig McClockey. They talked to a lot of I mean, there are there are a couple of folks that I would probably additionally add, but they they talked to the client mates.

Speaker 1:

They talked to a lot of people that were there. Now I think, you know, I I I know some people online definitely took some issues with the article. And I'm like, are you taking an issue with the article? Are you taking an issue with the people the article interviewed? Because the article is just kind of a vessel for those conversations, my read is.

Speaker 1:

Okay. Then a headline how Docker broke in half. But I don't think it's like that doesn't feel it doesn't feel controversial to me that Docker did not succeed the way its investors hoped it would have.

Speaker 2:

That feels incontrovertible. I mean, just based on the valuations and the investments made, like, that does not seem up for debate.

Speaker 1:

Right. And that's not to say that, like, you there there are lots of valuable things that that don't necessarily succeed the way people initially envisioned. So that's not necessarily a criticism. But I I so I thought the article was very good. I thought it it it it it's always you always wanna get the voices of the people that were on the ground, that were the people in the room, for better and for worse.

Speaker 1:

And I think you definitely have a Rashomon effect, for sure. What is the last time you watch Rashomon, by the way, Adam? You you

Speaker 2:

I think never,

Speaker 1:

unfortunately. Are are you serious? You've never worked for Shamans. Yeah. How has this not come between us?

Speaker 1:

Do you

Speaker 2:

I I I I I don't I don't know. You would've thought it would've been the next thing we watched after The Hottie and the Noddy, but it wasn't.

Speaker 1:

It do you know? So this is unfortunate that you've now created many more questions for folks. It is true. I wanna be clear. Let's just get it on the record now.

Speaker 1:

Let's just be out with it. Adam and I saw The Hottie and the Noddy in the theater being 2 of at most 20 people nationwide who did so.

Speaker 5:

That's right. And That's right.

Speaker 1:

A, The Hottie and the Naughty, a film you haven't heard of for a reason, say Paris Hilton Vanity Vehicle, that was true was a very bad movie, but not the worst movie that we've seen together in theater.

Speaker 2:

Are you gonna put that on Julie?

Speaker 1:

I am definitely gonna put that on Julie.

Speaker 3:

Now we're going

Speaker 2:

on it. It's totally different.

Speaker 1:

I felt like I've been around an open can of paint.

Speaker 2:

Are we forget?

Speaker 1:

Okay. I had a headache. I had a headache that took 2 days to clear after Geely. Geely was that bad.

Speaker 2:

Are you forgetting Gotti? You are forgetting Gotti.

Speaker 1:

Gotti was also yeah. Alright. So okay. We we we we need to to stop this off now. We're kind of this is where we need we have analytics on the Twitter space so we can watch everybody, like, run into their fucking lives.

Speaker 1:

And they're like, what why am I listening to these people say anything about anything? Like, a bunch of morons that saw Gotti, Gigli, and The Hottie and The Nottie all in the theater. Did I hear them correctly? Yes. I mean, there's a story there, but it doesn't make it any better.

Speaker 1:

It's it's terrible. And we took the intern to Gotti. That was bad. The I I felt like and that intern's a lot closer to my kids' age than my own age. I'm like, I'm abusing a child by taking him to Gotti.

Speaker 1:

I I

Speaker 3:

did see a couple 100 when you said people running for the exits, so maybe a good time to

Speaker 1:

Oh, oh, thanks. Oh, thanks, Steve. Thanks. Yeah. Exactly.

Speaker 1:

Just helping us read between the lines there. I appreciate that. Alright. So Steve But but but to bring it back,

Speaker 2:

the slow moving disaster that we were watching in real time in in Docker.

Speaker 1:

Yes. So well, and also so Rashomon is a I think we're actually talking about how you've not seen Rashomon. I think that's actually that's that's

Speaker 2:

that's that's that's that's

Speaker 1:

just I mean, not please, don't change we'll change the subject in a moment. But you you Rashomon is really worth watching. Steve, have you seen Rashomon? No. How do you endure Just just just listening

Speaker 2:

back to me.

Speaker 1:

So many doggy. So many I make a lot of Rashomon references. Do you guys just all, like, can collectively roll your eyes when I make a Rashomon reference? I mean Yeah.

Speaker 2:

I thought I thought it was a brain injury, honestly.

Speaker 1:

And it's worth watching. And the whole plot of Rashomon is you've got 4 different people who saw the same thing. And they are the way, Kearsawa tells the film they as they are describing what they saw all his films are set in medieval Japan. And as they're describing what they saw, they're kinda reenacting it on the film. And you realize that these 4 people believe they were at were at the same event, but they saw totally different things.

Speaker 1:

And it's it's absolutely that film has aged very well. It's a great film. But it's also, like, very on point for this because I feel that you've got, you know, all these people, McCloughkey and Ben Golub and Solomon Hykes and and, you know, us to a degree because we were kind of, you know, on in the Docker orbit. We're all in the room, and everyone's perception is gonna be slightly different. And there's it's very hard to get, like, you know, what is the one story.

Speaker 1:

You're, like, you're not gonna get to that. You're gonna get to a bunch of different perceptions. And in that regard, I have this article is great because it just it got people on the record about it.

Speaker 2:

Yeah. I mean, it felt like the last two books of our of our book club around around Next and then, around Windows NT of, of that effect of, like, lots of different perspectives here. And, I I think you made the comment, Brian, that you'd love to read the book on on Dockers rise and fall or splitting in half, and we'll see if they have a second act.

Speaker 1:

Oh my god. I would read the book in a heartbeat. I'm not to not to tack into your accusation of us as being in a book club, but I guess it's it is it's on point. I'm reading Jerry Kaplan's startup right now. Rereading actually because I read it when it initially came out in 1995.

Speaker 1:

Great book. But the part of what is so interesting is he has got very detailed notes from this company he started, Go. Not that Go. Different Go. And so I think I love to read a book on Docker.

Speaker 1:

I don't know if such a book is possible because I don't know if the notes really exist. You really need someone contemporaneously to be recording everything, I think, to write something definitive. But I would read it in a heartbeat. I mean, would you not?

Speaker 2:

Oh, yeah. Absolutely. And I was just thinking the I think maybe the only way that happens is is if there's some, document discovery or something like that. Yes. And your years of emails get pouring into the public.

Speaker 2:

But you're right. It's otherwise, it would have to

Speaker 1:

get Oh, I see. We so you're saying we need to sue the I I I like what you're doing. I I see what you're doing there. That's right. We're gonna use discovery to write a book that has a very small it's a it's a very small demographic book club, which is, you know, the name is right on the tip.

Speaker 1:

It's a very small demographic.

Speaker 2:

That's right.

Speaker 1:

That's right. But I would read it in a heartbeat. I think it'd be really interesting. We did so just to add for history for those and so, Steve and I worked together at Joynt, a bunch of a bunch of other folks over here. And Joynt, well, had been very early.

Speaker 1:

I've been a container pioneer, I think it's fair to say. And when Docker was open sourced in 2013, we were very excited to see all the energy around it. And so we became kind of we became a partner of Dockers. I dare say, Steve, we got kind of a unique look in the Docker because we were neither, like, a Docker subordinate nor were we a a belligerent. We were kind of in this, like, very weird, quasi neutral stance with respect to Docker, almost uniquely.

Speaker 1:

So I think we got a view that other people didn't get.

Speaker 3:

Yeah. And I think we we were in a position, which I I think is not dissimilar from a lot of people in the ecosystem at the time in 2014, where, our customers, folks that were, deploying, running their own infrastructure in their on premises data centers, folks running in the public cloud, were seeking better support and services around this widespread adoption in their company of Docker. And they they were looking for how do we start more meaningfully stitch this together in terms of operational services, you know, general management kinds of things around log logging and repeatability around, you know, container images. And our perspective was we have this customer base that would be very eager to pay Docker for help in making it easier to run and operate containers at scale because the adoption's growing

Speaker 1:

quite a bit. And so, yeah, we

Speaker 3:

you know, with that perspective, it was, customer need that that that desire for a lot of companies to pay them money for kinda operate, operationalizing the Docker container usage.

Speaker 1:

Yeah. Well, I think we because we had seen the kind of the commercial application of this, I feel, in a way that they didn't. I know Steve do you remember you that you had that conference with in Vegas that you came back from? And you're like, you know, I I was actually I realized that I knew more about Docker than the person on stage. And and I'm like, Steve, of course you did because we've been, like, we've been we've we've been kinda waiting for this container space to really materialize.

Speaker 1:

And I'm not sure that was a a Docker Inc person or not. But, Steve, what did you see in terms of the way?

Speaker 3:

It was. But, you know, I think one thing just to back up a little bit is, you know, I I think it is important to decouple the kind of the financing and the venture expectations from where they were, where the project was, and, you know, their evolution trying to understand the market and and what the opportunities were to support that market. Because, I mean, it was pretty tremendous what transpired over the course of 20 13, 2014 just in in the technology and the adoption of the technology.

Speaker 1:

The adoption of Docker.

Speaker 3:

Yeah. The the adoption of Docker. I mean, it was it was remarkable how quickly it was adopted, that it persisted, that it still is in as much use today as it was then. I mean, that is rare error in some regards. And I think what this article and maybe maybe where some folks bristled at the article itself was the the focus on failure of the commercialization, of which I know we'll spend some time on today.

Speaker 3:

But, there were also a lot of things done right.

Speaker 1:

Oh, absolutely. From a technology perspective.

Speaker 3:

Technology and a project.

Speaker 1:

Well, I mean, there's no question that they hit on something very big in terms of a because we saw a container as an operational vessel, but honestly failed to really see a container as a development vessel. And and that's what they saw. They saw a container as a development vessel, which is extremely important. I do I mean, clearly, we're gonna talk about how Docker broke in half. So there's going to be yes.

Speaker 1:

We're we're gonna talk about the the the the challenges of Docker Inc. I do think that we gotta go back to dot cloud though to really talk about Yeah. Docker. And Steve, I was trying to remember if had they approached us they approached us to buy dot cloud.

Speaker 3:

Yeah. There was a a conversation about that, early on. You know, how real that that conversation was or could have been.

Speaker 1:

I mean, that conversation was, like, pretty real. Like, would you let me, like, take my baby? I mean, it was a very real conversation in that regard. I mean, it wasn't real from our perspective. We're like no.

Speaker 1:

No. It's like you should you, like, give it a home. Like, that's not but was that before they had open source Docker or afterwards? Do you remember?

Speaker 3:

I think it was before.

Speaker 1:

I think it was before too. Because they were based I mean, part of what's interesting to me about Docker is this kind of open source angle of it that they they were a PaaS that was really struggling. They had and I I I feel I've seen Solomon speak on this, but they basically got the company down to, like, 3 people. So they were a YC company founded in, like, 2008, 2009, something like that. They now were you know, it's in 2012, 2013.

Speaker 1:

The company is is not succeeding as a PAS. They are trying to sell it and are having a hard time finding a buyer because it's tough to find a buyer for some business like that. And then they open source it as kind of a Kind of a last resort. Kind of a last resort. Yeah.

Speaker 1:

Yep. And to a certain degree, I thought it was on there there's a degree which is like, boy, this is a real victory for open source and open source as a commercial vehicle because all of a sudden the company that nobody had heard of was a company that everybody had heard of. And, you know, that happened kinda overnight. I think they open sourced what? Like, February of 2013.

Speaker 1:

Yep. And that happens over the course of 2013. We start seeing more and more enthusiasm.

Speaker 3:

Yeah. By 2014, it was just going full steam.

Speaker 1:

2014, it is going bonkers. And this gets us to my first I feel like I do feel this is an important decision point. And not to dwell on it, but this is a a mistake that they made that is not a mistake that that every company has to make. They took too much money in 2014. And they you know, when you are raising when you are raising venture capital, there are venture expected outcomes from raising that VC.

Speaker 1:

They are in it for a 10 x plus return, actually. We did have one VC firm that were made name of us that passed on oxide because they said it was gonna gonna only be a 10 x return. It's like, okay. That's isn't that good? Okay.

Speaker 1:

That's not good enough. Alright. Fine. So but it but it's VC firms are looking for outsized returns, and they want an outsized return on that capital. And if you take on a lot of capital without knowing what you are going to sell, there's a real danger that you're not gonna spend it wisely.

Speaker 1:

And, I mean, ask me how I know. Well, and and you're gonna draw

Speaker 2:

the wrong inference from that signal to to confuse that those v two dollars with success, at at least at some level.

Speaker 3:

Yeah. Geez. With signal. You're gonna confuse that with signal and that you're on the right path. And I think one thing the article got right is it it got to a number of people that that acknowledged not spending enough time frontline with customers.

Speaker 3:

And that was what was clear as day. And that that's where, you know, coming back from that presentation, in Vegas and then seeing kind of some of the iterations on the business side in there. There was a clear disconnect between the end users that wanted to pay them for the product and, you know, how they were thinking about packaging and selling it.

Speaker 1:

Alright, Steve. I gotta put that bait in the water for you. The one of the ways that raising money makes you stupid is especially for technologists that don't understand really how the go to market motion works and what someone in sales does. If you are it is many technologists think that people that are in sales are just kind of like money magicians, and you just kinda like rub them on your company and all of a sudden, like, pipeline appears. And the if you wanna get the best magicians, where do you go, Steve?

Speaker 1:

You go to the biggest companies.

Speaker 3:

You you go to the biggest companies and the former VPs, SVPs, EVPs that have run go to market at those companies. And you bring them into your startup and watch the magic happen.

Speaker 1:

And hopefully, no one just listens to that bit of audio content without taking some broader context because this is what not to do. This is what not

Speaker 3:

to do. Unfortunately, the magic is, that person can then make 20 people appear who then can make 40 people appear. And those people, of course, are also gonna ask, like, where is the rest of my apparatus to ensure I'm properly fed and cared for? And

Speaker 1:

I see.

Speaker 2:

Like a like a pyramid.

Speaker 5:

Well, and

Speaker 1:

and this is where I mean, this is I mean, and, you know, god bless Silicon HBO's Silicon Valley for getting this unbelievably right. I mean, this is just I mean, Adam, you've Adam, are you in the demographic that can't watch HBO Silicon Valley because

Speaker 2:

it's too real? Look. I've told you this like, I I got it through through about a a season and a half, and then it was 2 weeks of them talking about exactly the acrimony I was dealing with at work, and it was just not funny. I gotta get back to it.

Speaker 1:

You I mean, okay. Forget Rashomon. Put Rashomon. Like, you know what? Don't you know what?

Speaker 1:

I don't care. Don't watch Rashomon. You've gotta watch HBO Silicon Valley. Alright. I mean, it is I mean, Steve, the the any Steve, what I know you've you've made it most of the way through HBO Silicon Valley.

Speaker 3:

Yeah. I'm through season 4.

Speaker 1:

Right. I mean Yeah. So you you you definitely got Bob from Northwest Regional and Keith, I'm shadowing Bob.

Speaker 3:

Oh, absolutely.

Speaker 2:

And I mean, I got Jan I got Jan the man.

Speaker 1:

Jan the man. And you've got the sales team. So the for those who haven't seen it, I mean, it's just sublime. It is so good. I wanna know who wrote every word of that episode, episode where this where action Jack Barker brings in the sales team.

Speaker 1:

And, of course, the sales team has nothing to sell. So they are all I I I love the line, Steve. I am paraphrasing, but where it's like Richard's like, wait a minute. Like, are these these folks are the best? Like and, like, no.

Speaker 1:

No. They're the best because the product sells itself. And Richard's like, that that that's right. That make me sense. But I feel like that is a very common failure mode that we have seen a lot.

Speaker 1:

And to be clear, Steve and I lived at joint. Like, that is like, the reason that this is so visceral for us is because we watched it act that in front of us.

Speaker 3:

And And and again, it's a that that motion of bringing in a sales leader who then can recruit and bring in and develop and kind of build that apparatus is exactly what certain companies at certain points in time need. It's yeah. I think the failure mode that that you're referring to is companies that are still early, that really are still trying to figure out what their product is, who in the market it's the right fit for, what problems it's solving for. And, you know, in that learning phase where they just need to be spending a lot of time with the potential users of the product absorbing as much as they can, you know, technically steep so that they can have those those relevant conversations with those those end users. And if you start with that that top down apparatus, one, it can get you can get into an internal feedback loop that can create very bad long term problems.

Speaker 3:

And and, 2, you are burning an incredible amount of money while you're trying to figure out what the product is, who it's for, what it's what it does well. And, yeah, that we we we certainly live that, and and this is something I think that

Speaker 1:

You are now you're bringing you are you've also good news, but wait, there's more. You've also bought yourself a civil war because that sales org is going to rightfully now blame the product and maturity. Like, I will I would sell this, but, like, the product doesn't do any of this stuff. And now your engineering leadership is spent instead of, like, actually spending time, what you should be doing is everybody should be spending time arm in arm with customers understanding how is this technology gonna solve a problem for which you're gonna wanna pay to have the solution. And that's I I mean, Steve, I don't know.

Speaker 1:

I I mean, that was my my read is, like, their kind of, like, fundamental failing at the kind of the deepest possible level is that they felt their customers were the developers that were downloading it and not the people that were gonna pay the money to buy a product.

Speaker 3:

Or, you know, at least at least not having gone and spent the time to figure that out before rearchitecting pricing models and organizational models and and the rest.

Speaker 2:

And then back to money, that's what enabled them to do that in in such a kind of high scale fashion that separated the folks making the decisions from, the input from those,

Speaker 1:

great question, Tom. And I actually had this I there was a bit of, I would say, a boom in because I put NPM here as well in these open source technologies that from my perspective, and I think maybe I dare say from many people's perspective, just like when you say, Tom, it's like, this is a technology. I don't see any like, what someone walked me through something that someone's gonna pay for here. And I remember having a conversation with the VC in, like, 2014, kind of the middle of this 2014, 2015, the middle of this boom. And saying I'm like, actually, you know, just like just bear with me for a little bit of a thought experiment.

Speaker 1:

What if the thing you you guys are investing in things because of their downloads. You're looking at the downloads. The up into the right chart that you're investing in is downloads. It is not people consuming a product. It's people downloading this software.

Speaker 1:

What if developers are, and it's kinda speaking to myself here, attracted to those things that they know cannot be And he

Speaker 4:

And this and this is the big business conundrum now with open source. Right? You open source it to make it really popular with developers, when was the last time a developer wanted to pay for anything?

Speaker 1:

Well, that's it. Exactly. And and it's, like, deeper than that, Tom. Right? You know exactly.

Speaker 1:

What did it I mean, all of it, one, like okay. Look. We're we're we're cheap. I mean, I like to say we're thrifty, but we're cheap. But we also have all seen companies that end up being built on proprietary bits that are then end up not being able to scale because they're built on one of these proprietary bits.

Speaker 1:

And we don't wanna do that with the technologies we build. We do we want to build out on things where we've got that kind of investment protection. So we absolutely are seeking out things that are open source. And

Speaker 4:

Right. And then and then Docker would be nowhere if they hadn't open sourced it, so that's not a viable strategy either.

Speaker 1:

That's not a viable strategy either. No. I think you're exactly right. Docker we in fact, we know what Docker would have been without outsourcing it because that's what they were. They were dot cloud looking for someone looking to please take my baby on the in 2009.

Speaker 2:

But that distinguishing characteristic from the high downloads of, like, Elastic or Mongo that then turns into a real commercial entity versus something like Docker. Like, that may be hard, especially for investors to infer. Well, I think

Speaker 1:

that that's exactly it. Is that, like, it is it is hard. And I, you know, I tried to have a conversation with folks that were investing in Docker in 2014, and they were not really interested in having the conversation about, can you just walk me through how this gets to a product that we that someone buys? Because I'm I'm I think this is someone who would be a part of your your target demographic in terms of, like, a software developer. And, you know, there was not there was this kind of belief that, like, no.

Speaker 1:

When the technology is this ubiquitous, it will be readily monetizable. And But

Speaker 2:

I mean, don't you think Swarm was the I mean, obviously, Swarm was their swing at that. And I guess they just sort of I mean, I really missed the boat with that.

Speaker 4:

Well, Kubernetes happened.

Speaker 1:

Well, yeah. So this opens yeah. So this is and there are some bits. So of the there's some pieces that are missing in the story that I do think are important. The Docker convinced themselves that Kubernetes well, first of all, that Google was the enemy.

Speaker 1:

And that Google wanted to bring Docker to be part of Docker and move it in a direction that would allow them to replace Borg with Docker. And though and perhaps there are people here who can provide insight in those how those conversations went, but I gather, like, not well. And certainly, the the and I think Craig even said this in the article. McCluckie and and beta and crew would have been happy to have to not have done their own thing. And that is a major misstep, to have that expertise show up and not leverage it.

Speaker 2:

I gotta say I really enjoyed that bit of the article because it really spoke to my experience where they talked about, you know, the Docker folks feeling like the doc the Google folks were looking down on them because they didn't have PhDs, and the Google folks feeling like these open source clowns didn't necessarily know what they were doing. I'm paraphrasing here.

Speaker 1:

But you you're not paraphrasing by much. I thought that was a very reviewing I had that was a very reviewing one. That is a that's a that is a quote, I think, from Solomon.

Speaker 6:

It was, that that's pretty much on the nose. The only thing that I think was slightly different, and and and and I still think what you said is pretty much on the nail, is that Google didn't understand developer experience. I think, was the other side of the coin.

Speaker 1:

Yes. I think, Nick, you're exactly right. And, the the there was a real I mean, you know, there's there's an opportunity to take the best of both worlds. This is like when I thought we were gonna get, like, Oracle's business acumen with Sun's technology. There's this kind of, like, abstract possibility of taking the best of both worlds.

Speaker 1:

Like, can we take Google's operational know how and Dockers clear understanding of the developer and made them into a single technology. And, of course, like, no. That would always breaks up in the civil war. But the other thing is that I think is worth mentioning is that in, so in DockerCon 2015, you know, I don't know when that was in the year, but this is now Kubernetes is starting to get Kubernetes is started in 2014. It started to get momentum.

Speaker 1:

They got the 1st KubeCon, I think, was I wanna say in Steve, do you remember, is that 2014 or 2015? Remember 2014? And oh, god. I'm just remembering the person who ran the first coupon who's, anyway, but, anyway, the, we are the first coupon, and so Kubernetes is beginning to get some real momentum. And then, Steve, you obviously were at DockerCon in 2015 in San Francisco.

Speaker 1:

And you remember the the ban on Kubernetes. Do you remember this? Yeah. It

Speaker 3:

had it had to be removed from any slide, could not be mentioned. I mean, it was it was a hard, hard ban on any

Speaker 1:

So you could not say the word Kubernetes. And that is you know, that's this is like when you're, like, you're needing to put up, know, a wall in Berlin to keep people in. It's like, it's not working. Something's something else is not working. If you need to ban people using the word Kubernetes at your conference, like, you're on and, Nick, I don't know.

Speaker 1:

Did you work for Docker at the time, or did you it's obviously, you were in the ecosystem.

Speaker 6:

Oh, no. My perspective was I was in university. So I graduated, undergrad in 2019, and by that point, by that point, we were all using Docker, and we had heard of Kubernetes. Not a lot of folks knew how to use it. And and we we would keep getting mentions of DockerCon, Docker this, Docker that.

Speaker 6:

And the way that I've sort of read it from the outside was there was this migration happening, and it was unclear as to why and Swarm existed. Why not just use Swarm? So right as I think we were all about to look into that, Kubernetes seemed to have popped up out of nowhere. And I think the fact that Google had it had a lot to do with why we did not continue. Because I still remember there was this, regrettably a patent that that a professor was working on and it was going to use, Docker as sort of the, like, as the proof of concept.

Speaker 6:

And then he ended up stepping back and saying, you know, I'm not sure if this run time is what I need. I might need something more scalable. I'll wait for this Kubernetes thing to shake out. And that that's probably the earliest I remember of it is that is that there were sort of this next step. What what now?

Speaker 6:

We can run our apps on top of our rel servers, but there isn't really anything expansive beyond that unless you use compose and other things, which are still very beautiful from a a a DX perspective. But I I I I think that is the vibe I got just being in the ecosystem for a long time from like a, you know, a a university student and then before that a bit in high school is that it was sort of this it it it was almost like it felt as though in the in the research groups, we were told that companies were switching to Kubernetes. And as a result, it felt as though, that conversation happened behind closed doors. So when I read the article, it kind of affirms the suspicions of the past that it was very strange that these competing solutions came up seemingly out of nowhere. Because it sounds like this all happened in the background.

Speaker 6:

I mean, I could be wrong. They this Yeah. I mean all anecdotes.

Speaker 1:

It it it definitely generate there was a lot it generated a lot of bad blood. And, I mean, certainly, having people at your conference, but then banning them in terms of what they can say and how they can say it. So Brendan Burns was going to give a talk at Con 2015, and they forbid him from from he couldn't give the talk, and he couldn't say the word. So he had to give a canned talk that was and he, so, Adam, if you go back to that talk, the video of which I think it would still be online, he is wearing a Kubernetes shirt, and he's got, like, a Kubernetes screensaver. And there's, like and they, like they've done every passive aggressive thing they can possibly do to get Kubernetes there.

Speaker 1:

Oh, wow. Yeah. Which is, like, man, this is not good. This is, like, this is not a good use of energy. I also think we need to talk about Rocket, which I mean, Steve, do you remember where I mean, I remember Rocket very viscerally, because we were so we joined.

Speaker 1:

I had a new CEO at that point. A CEO who's just coming on board. Right? I think, Steve, am I getting my the the the dates right here? This is, like, in December of 2014.

Speaker 3:

Yeah. Yeah. Yeah. He had been on for

Speaker 1:

a month. And I remember, like, we were in an estaff meeting, like, talking about, you know, the the you know, how we were partnering with Docker and some of the complexities of that. And then all of a sudden, they're right before DockerCon EU, now CoreOS, now Red Hat, has launched this totally rival container format. You're just like, oh my god. The container wars just started.

Speaker 1:

Am I remembering that accurately, Steve?

Speaker 3:

Yeah. Sorry. I had to pull over. Yes. It was it was that that was I mean, first, the just back to the Kubernetes and Swarm announcements.

Speaker 3:

Fact that that everyone that was part of the ecosystem needed to scrub their materials of Kubernetes then made everyone wonder, was this, you know, were these two sides really pitted against each other, which just, you know, raises a bunch more questions and gets people kind of assuming that there is, even more there. And then, you know, back to how much money was getting pumped into this container space, you know, CoreOS was right there trying to, launch their own their own, you know, almost fork and being able and then trying to wrap a bunch of of container orchestration services around that, which just further confused me.

Speaker 1:

And then, Steve, what when did you first, like, see Docker actually in the market with, like actually go to market and an attempted product? And and notably I mean, we're not gonna get out of here without talking about the price list.

Speaker 3:

Yeah. Well, I think, you know, again, it was around when they they raised that big round, and they built out a large sales and partner organization that, again, as we talked about, was, you know, kind of internally focused to a certain degree and trying to figure out how to meet business plans, which consumed a lot of calories and did not focus as much energy on folks that had really started adopting Docker at scale amongst the developer organizations in these enterprises. In those enterprises, you had the management teams looking for some way to put some guardrails on it because there was fear about, you know, kind of uncontrolled sprawl and other security implications. There's consistency implications. Is it starting to sit underneath or around a bunch of our software pipeline?

Speaker 3:

And while it wasn't a quantifiable go build this product for me, you had a ton of enterprises that were saying they were willing to pay a lot of money to get some help and support around that DIY effort because it was just do it yourself for all these enterprises all the way to the other side of things, which is a platform, a product that you would buy that would manage all that for you, which became Swarm. There was this period of time over 12 months where these enterprises were begging for something in the middle. You know, start giving me some blueprints, of ways in which I can better, you know, operate and log and manage and, and and deploy and secure my Docker environments. And with those cries for can we go spend money, with you on that, what came back in in terms of kind of one of the big first product pricing announcements, was was a bit assaulting. So what you're Go on.

Speaker 1:

What I

Speaker 5:

think you're referring to,

Speaker 3:

Brian, might be, they they and and this kinda speaks also to, I think, a bit of the struggle where, inside Docker, they were wrestling with how far downstack they could go, should go, and how much they should be able to charge for, things that we even sit below Docker. And, they constructed a pricing model that for these enterprises that wanted to have the the largely support, but we're looking for more products to come post support. Something that if you have a server that you bought for, you know, $20,000 and then someone puts VMware on it, they they generally would spend on the order of about $3,000 a year back then for ESX. I I don't think it's much changed. But, the pricing model that came out said, then if you run kind of Docker management engine and the support services around that, you would then spend based on the number of VMs, the number of, of of how much infrastructure was on that was on that server, but on the order of another 3 or $4,000, per server, per year.

Speaker 3:

So you've effectively doubled the cost that one is spending already for the hypervisor and control plane on prem. And the that that wasn't the worst thing. That was, like, almost almost tolerable. And and and then But

Speaker 7:

but what have I

Speaker 3:

what have I done?

Speaker 1:

That's a problem.

Speaker 2:

Well Oh,

Speaker 1:

you're one of those, aren't you? One of those down voters.

Speaker 7:

Well, I just,

Speaker 1:

I mean, what if

Speaker 7:

I just don't pay, though? Like

Speaker 6:

Hey. He's helping pad the download stat. Okay? That apparently, that makes VC money, so he's fine.

Speaker 3:

And well, so by the way, that was 5 by 10 support. So that was Monday through Friday.

Speaker 1:

Can you imagine? Can you imagine the person?

Speaker 3:

That ended that ended that ended that ended at 6 PM Pacific. And then one would logically ask the question of, well, what if I wanted 7 by 24 and I I I wanted you know, we're really gonna be investing, betting on this for enterprise platform. And that went up from, like, 3,000 to on the order of 12,000 per machine. And, you know, I I do distinctly recall a conversation at their offices where we were just saying, look, we've got a bunch of enterprise customers with wallets open that want to pay. You know, they're they're they're believing that this is a path for a very important technology layer in their infrastructure.

Speaker 3:

And if and if it helps their developers go faster, and and and this can this additional support services and productization can help, you know, them have a more secure environment that is easier to operate, then they're gonna be leaning in. And by the way, they're they're willing to pay a premium on top of what they are already paying for their infrastructure. It it can't be 5.

Speaker 1:

Right. It can't be some 20 x. You you can't be more expensive than the computer. You you just can't be. Like, you've got the price point was just gone.

Speaker 1:

And I'd actually forgotten that that price point only included, like, Monday through Friday support. And you imagine having a a support issue on a Saturday, having spent, you know, $3 a year per server for again, for what? For what? And this is the other problem, Steve. So, like, it wasn't really clear what you were what were you buying?

Speaker 3:

Well, I mean, the hard part was that in some ways, back to Tom's original question is, was there a business there? Who knows? Because, you know, even though it was not on the customer side, they didn't have only a select set of requirements that they wanted. They were willing to pay for the the architectural guidance and support and and, you know, troubleshooting and troubleshooting and kinda augmenting their teams. And that is, you know, it was a pretty good sum times hundreds of companies.

Speaker 3:

And if that had not been prohibitive and then, you know, other question I have is whether or not, you know, flipping from an open source, easy to download technology to a proprietary platform in Swarm. I think there were definitely some opportunities there to allow one to walk their way into a container registration product, by not making it a hard cut over to a

Speaker 1:

Well, I also feel and, I mean, I floated this to Solomon when we first spoke in April 2014 that the I mean, the I felt that the best way to monetize it was to run container management as a service and where Docker Inc would buy the infrastructure. And it's like we and then go to I mean, clearly, that was, you know, it was pocket hard book to a degree, but it was clear from the dot cloud experience that they were never going to do that. That they've been so wounded from dot cloud that them running a service was, like, off the table. I don't know if, Steve, you got that same same inference. But every time I kind of tried to encourage them to explore operating a service, it was just like, no.

Speaker 1:

That's a hard no. They're never gonna do that again. Yeah.

Speaker 3:

They were not. And and to some degree, you can understand in that time frame and ever after 13, 2014 being averse to trying to build public cloud services that we're gonna compete with. You know, I

Speaker 1:

think that's the I can understand it.

Speaker 7:

It's like then if not the public version, then why why not also replace the VMware bit? Like, why do I have to buy VMware and all this other stuff for $20,000? Like Yes. Like that's why not roll those 2 layers together and instead of VMware getting $3,000, you get 3 $1,000. But, you know, your your thing has containers.

Speaker 7:

Like,

Speaker 1:

Right. So that would have been another approach. They would probably have had to I think they probably would have had to acquire that realistically, but they could have done it. So you're saying Josh It's not

Speaker 7:

like Linux does. Like, Linux boots on hardware like ESX does. You know what I mean? Right. You know?

Speaker 3:

But even before going downstack, I think there also was a ton of opportunity to think horizontally about what are the different, again, operational, security, performance, like the the the standard operational tasks that that one is responsible for in running infrastructure. And you had to go to multiple projects or multiple companies to stitch together that overall Yeah.

Speaker 1:

I kinda feel like Rancher kinda did this much better in a go to market. Yeah.

Speaker 2:

I agree. They really did.

Speaker 1:

I mean, I kinda feel like if if Rancher has been dealt Docker's hand, I feel they would have played it a lot better.

Speaker 6:

I I might be able to speak to this to a degree. Full disclosure, I work at, Rancher Labs now, Sousa. We Oh, wow. Yeah. We Okay.

Speaker 1:

Well, then you can definitely tell me authoritatively.

Speaker 6:

Well, no. I couldn't because I I got I I knew they were getting acquired when I joined, but at the same time, I've heard bits and pieces of the story. We do have a hosted Rancher solution, where people can, you know, sorta use Rancher to manage downstream clusters. But we've sort of and and this is partially external, partially internal view, like, what what I thought before and what I think now. We've sort of ensured that we act

Speaker 1:

what's the best way

Speaker 6:

to put this? We what what makes something like e 2 and e ESX successful is the sort of, in my eyes, a a plug in my model. Bring your own ISO, bring your own AMI, execute your own statically linked binaries, which totally won't break somehow. And I I I I think our vision, especially at Rancher 2.5, has been you can provision a cluster from any major provider whether that's on VMs or on, hosted services, or you can import any so long as there is something resembling a cube config. And and I'm trying to not make this an advertisement.

Speaker 6:

I'm I'm trying to make it relevant to the conversation, so forgive me for the backdrop. But I think to sort of tie this back in is that I think they avoided this by essentially saying we're going Kubernetes native. And we don't know if the landscape is going to look like managed clusters or distributions. So Rancher sort sort of has multiple multiple verticals. Right?

Speaker 6:

There's k three s and RKE, which are like the the distros that run on the Linux systems themselves. And then there's the Rancher management product, which kind of like is its own vertical, but they both can work in tandem. So so I guess the answer to that is they did a pretty good job of moving in multiple directions at once. Rancher and k three s, I think have a similar amount of stars on GitHub, which is why I, by the way, requested to speak in the first place because your comment about downloads, people definitely do care about stars for better or for worse.

Speaker 1:

Oh, well, I mean, people care about stars. I yes. Oh, god. Trust me. I I, again, I was incarcerated by the CNCF on the TOC having to endure projects telling us how many stars they had.

Speaker 1:

Like, this isn't gameable.

Speaker 2:

Yeah. Yeah. But but but stars aren't there's not, like, a magic alchemical formula to turn stars into dollars.

Speaker 5:

No. But, I mean, Cocoa did have a pretty amazing opportunity ahead of them, based on the user base and based on the the, you know, ability to put marketing in front of the end users. And in terms of, like, the landscape at the time that they received funding, some of their contemporaries in that open source enterprise space like, Elasticsearch have managed to make it work. Npm NPM did get ended up getting acquired by by Microsoft, but, you know, that was around the same time. And the other big one that hasn't really been mentioned so far is HashiCorp.

Speaker 5:

Yep. Where where like Vagrant was at at the the thing that people use prior to Docker pretty much for this like, local dev environment, consistency consistency problem. And they found ways to monetize that into, you know, a suite of services, which, you know, they seem to be a fairly successful private company. So, the other space where they could have made money, was around that Docker Hub opportunity where, like, certain type is in that same sort of space selling their Nexus product. And they're incredibly successful company as well.

Speaker 5:

So like there's definitely a case to be made looking at the opportunity ahead of them in 2014 that maybe the valuation was was sensible. It's just, like, an execution and focus problem, which I think the article did kinda get into.

Speaker 1:

Yeah. I think you're exactly right. And I'm glad I mean, I'm glad you mentioned HashiCorp for sure. I think that the I mean, what I and actually, Nick, my point about Rancher and Hashi too is that to me, the difference is more fundamental. I just felt like with both Rancher and Hashi, they were actually, like, interested in I I

Speaker 7:

Just I'm still not clear on what, like, for the 3 to to $20,000 that you pay to get that enterprise product at the time. What what did it do?

Speaker 1:

Well, this is the challenge. I mean, they that they had already decided

Speaker 7:

you didn't get in the free box. Right. And and, like, you didn't get you still have to buy VMware, and you still have to buy much other stuff. Like, it's not like it did bare metal provisioning management stuff. It was just, like, this pretty thin layer at the very top of the middle that was very expensive.

Speaker 3:

Well, to be clear, not many people made it.

Speaker 1:

Right. But they they were but

Speaker 3:

I think what they were intending I mean, what what folks want to pay for at that point was support. And then they wanted to see product features start to show up in supported versions that they did not care whether they were I mean, prefer that they would remain open source. And what that dollar figure is they would have paid still seem to be, you know, in the realm of VMware plus, but not, you know, 4 x then.

Speaker 2:

Right.

Speaker 7:

It it feels like the UX stuff that they did was successful in part because they had a good enough understanding of what people on their laptops actually wanted to do, And also because it's hard to compete with free. Like, there's a very real sense in which things like TCP, IP, and Ethernet, by being effectively free, ended up in lots of things, and are still with us today. And, like, the open source parts of Docker, still with us today in very wide use, very, like, meteoric adoption. But it's not clear that anyone would have paid a bunch of money upfront for that part. Like

Speaker 1:

Well and I think that they were really struggling with how to monetize. I mean, I had a very awkward conversation with Nick Steinmates who's who's quoted in the piece. So we were a bit peculiar. I think unique, Josh. Right?

Speaker 1:

In terms of we were implementing the Docker API. I don't know that anyone else is doing what we were doing.

Speaker 7:

Yeah. When they they were definitely very confused when we didn't wanna give them any give them any money, like, any royalties for the because we shipped none of their software.

Speaker 1:

Right. So we'll yeah. Then Nick had this conversation with me. He's like, hey. Listen.

Speaker 1:

Like, I'm not saying it's happening now, but, like, at some point, we're gonna wanna get a Docker support contract in place. And I'm like Okay.

Speaker 7:

For for what?

Speaker 1:

For what? For or it's like, great. I'm like, I'm all for it. But what what are you even talking about here?

Speaker 7:

Well, for the for the Docker enterprise stuff that you

Speaker 1:

Yeah. The Docker enterprise. I'm like, well, we're actually reimplementing your API. So, I mean

Speaker 7:

It's not even Linux.

Speaker 1:

Right. It's a it's not even in Go at the time. Right? It's like it's not even it's not it it's nothing you would recognize.

Speaker 7:

And we're

Speaker 1:

which I think just because I think the other challenge that you have is that when you pour a bunch of money on a company that does not have product market fit, and then you it the the second you have that kind of first big miss, and now you, like, cash the team, the like, alright. The sales team, like, get rid of that sales team. Get the next sales team in here. You you start having this kind of, like, cascading organ failure where now you're getting, like, more and more desperate and weirder and weirder and weirder. And I feel like what would have been a viable because, Steve, one thing I I'm, like, dying to know is a bunch of the things you were talking about would have been kind of a service and support model, which I think would have been a totally viable model for them.

Speaker 1:

But do you think that their investors would have puked on service and support because of margin issues? Absolutely. Right?

Speaker 2:

Yeah. I I mean, this is the I think this

Speaker 3:

is the fundamental issue with raising that much money is to, you know, you're in Adam's earlier comments. Like, the expectations are very high for, you know, short to medium turn wins. And unless you've coupled that large amount of money raised with a agreed upon long term investment strategy and return strategy, which, you know, that it is usually doomed to failure because that model what I think what that model would have done is it would have been that's the that's the foothold to figure out the product that is going to help scale the business to the size in, you know, that might have been envisioned when the the hype cycle was happening and the money was pouring in in that spectrum of this total open source DIY world all the way to the other end of what the what the products are either horizontally or vertically that, would be repeatable and wouldn't be a a services and labor business model.

Speaker 1:

Yeah. And just to give kind of some numbers behind, Steve, what you're talking about in terms of, like, the where we were in the hype cycle. So they raised a $10,000,000 a led that's led by benchmark. It's Peter Fendt, benchmark, the great investor. That's in 2011.

Speaker 1:

So that's Doc Cloud. And so they open sourced in 2013. Hail Mary. Company is dead. They they come off the operating table.

Speaker 1:

In 2013, they open sourced. They raised a $15,000,000 b in 20 in January of 20 14, a $40,000,000 c in September of 2014, and a $95,000,000 d in April of 2015.

Speaker 7:

Sooner or later, it's gonna add up to real money.

Speaker 1:

That is a colossal amount of money. And, Steve, you are, like, I feel like they were always moving offices. Is that my imagination?

Speaker 3:

Well, when you're when you're hire when you're hiring that not always. You they they I think only moved once or twice. But within, you know, 3 years, that's a lot.

Speaker 1:

That's a lot. And they acquired I I I don't know if he's here, but they they acquired so Borja Borjas at, at at TUDEM. So they acquired TUDEM in 2015, and, Borja's got a very good let's just say the Borja has Borja's got a very good lawyer recommendation for anyone whose company is being acquired. I think I'm gonna leave it at that. But Borja, the the but I remember it would be great to get his perspective because I think that he so coming in, I the they're acquired in October of 2015, kind of right as that, you know, that with the raising again, they're gonna raise another $18,000,000 in November 2015.

Speaker 1:

I mean, it's just the numbers are just off the chart. So I guess on that kind of capital raised, you can't go back to a services and support model.

Speaker 2:

Yeah. Whereas I think Hashi was willing to get those, like, singles and doubles, like, to kind of take take the wins where they could. They had a variety of different products, all of different phases, and didn't have that, that kind of built in hubris around requiring a home run because that's what they had promised to their investors.

Speaker 8:

Right. Well, my question is why were the investors expecting something different? I mean, right now, AT and T's business isn't based around the e. L. F.

Speaker 8:

Format and the make bio.

Speaker 1:

Well, we so okay. Great And the I we had the same, like, kinda question for investors to invest in NPM. Same thing. It's like, you know, I'm sorry. Where's the the the

Speaker 7:

the $1,000,000,000 company? NPM was worse because it was basically, like, curl duct tapes to tar.

Speaker 1:

Right.

Speaker 7:

And maybe it ran some things. I mean, the registry the registry was the the expensive part to run, and that was the only thing people wanted to use really.

Speaker 4:

So They they call that couple of shell scripts as a service.

Speaker 1:

That that's right. So, I I would like to say, Josh, that your description of NPM actually applies almost as well to Docker. I mean, it is basically curl plus. I mean, it's basically the same thing. Right?

Speaker 1:

The, I think, Erin, to answer your question so here's my hypothesis for whatever it's worth. Because I was trying to reason you know, I can't put the down okay. Downloads to all open source.

Speaker 7:

Yeah. Downloads are the only part that it costs you to provide money. Right. The

Speaker 1:

so my theory on this is that 1, you know, VC a venture funded business, one of the things you're very concerned about is that you don't have product market fit. In fact, that you that may be the thing you're most concerned about. And there have been companies, I mean, Facebook, Google, Amazon, that where the economics of the business didn't look that great, but they had this very clear fit with the market. Like, people liked what they were making. And I think people kind of over index on that and thought, okay.

Speaker 1:

So I'm looking for something where everybody likes it and we've got no idea how to monetize it. It's like, well, no. Because, so I think that that Erin, I think that that might be their answers. Like, yeah. But I didn't know how Facebook was gonna monetize itself either.

Speaker 1:

It's like, yeah. But that now seems, in in hindsight, kind of obvious.

Speaker 2:

Also also mix in a a bit of, like, VC FOMO. Right? Like, there's there's there's a hot, ostensible space here. At least a lot of people are excited about it. And, do you wanna be on the outside looking in?

Speaker 1:

So do you wanna explain FOMO on that trip? I think

Speaker 9:

the part of the reason they didn't, think too much about, building down the stack actually relates to how they thought they were gonna make money in the end, at least how their VCs did. The initial announcements about Docker were all about the pain that developers felt. And I think what they thought was other operating systems beside Linux would start implementing primitives like, process groups, CRMs, cgroups, and, all the other sort of lower level bits that make Docker possible. And then Microsoft or any of these companies that are have a newfound zeal for developer experience would have acquired them for ridiculous sums.

Speaker 1:

I am I try You actually

Speaker 9:

am curious with all the operating system aficionados on this call. Why don't other operating systems want these primitives? They seem so useful and so

Speaker 7:

In the case of Microsoft. Right? Like, I mean, I am not a big consumer of, like, Windows, but certainly, they have an extremely long history of being excellent for developers. Like, the the reason I mean, you talk to people that exist solely inside the Microsoft ecosystem and are unaware effectively of a world outside their own borders. Like, the documentation is good.

Speaker 7:

The tool the tooling, if you never have to leave, it's good. Like the developer experience has been good for decades as long as you are only developing for Windows. Right? And it's pretty different to everything else, but it's it's good. They've always MSDN has always been fantastic.

Speaker 7:

The licensing is, you know, atrocious, obviously, and a bunch of other stuff is a problem. But, like, the actual part where you write dot net software has been pretty good. And I I don't think they feel like they need it anymore.

Speaker 8:

NPM. By buying NPM, they had a lever with which to guarantee that developing for the web would be a good experience on Windows. Well, NPM was sold

Speaker 1:

on Right.

Speaker 9:

They also implemented primitives to support v 8, if I'm not mistaken. They have, synchronization primitives that they implemented specifically to support Node Better.

Speaker 1:

Yeah. And so I think that and, Drew, I'm gonna go ahead and match Microsoft because I one of the questions that the article does not get to, I don't think, is the phantom $1,000,000,000 acquisition by the the the kind of the the the rumor at the time was that Docker had a $1,000,000,000 offer from Microsoft that they rejected. And I mean, I don't know. Steve, I'm sure you heard this one. I mean Oh, it's true.

Speaker 1:

And I don't know, Steve. What do you think about that? Having or Adam, you we've all been through acquisitions. I don't know. What's your

Speaker 2:

take? I mean, yeah. I heard the same things at the time. And I guess, on one hand, it didn't make sense. But now seeing the acquisitions that Microsoft has made in the last, you know, 5 years since then, I don't don't know.

Speaker 2:

Maybe.

Speaker 1:

Well, so my question is, like, I think they may have been suffer given their what I saw in the marketplace with the way they engage the market, I really question whether they were suffering from happy years. Like like, in this $1,000,000,000 acquisition, like, did you have an LOI? Were you in due diligence? Like, where were you exactly in this $1,000,000,000 acquisition? Or was this just, like, some talk from some BD execs?

Speaker 1:

Or maybe even from, you know, from from execs from Scott Guthrie or execs who would have been, you know, kind of chartered the purchase. But it was not actually an offer.

Speaker 2:

Yeah. I agree with that. And also on the I mean, easy to have half years on the back of what was probably some very easy rounds to raise.

Speaker 1:

Yes. Yes. And I mean, certainly, if they had that offer, they should have taken it.

Speaker 4:

Well, they they raised 300,000,000 or something. Right?

Speaker 1:

Yes.

Speaker 4:

So the valuation must have been up around a 1,000,000,000 anyway.

Speaker 1:

The valuation that was, I think, north of a 1,000,000,000, which Pretty close. And I you know, this is the this is the China's you know? Again, Silicon Valley, this is season 2 episode it's season 2 episode what? Episode 1, Sand Hill Truffle, and then then Runaway Evaluation, where the, where Adam, did you at least get to the season 2? Or did you, like Yes.

Speaker 2:

Yes. Halfway through season 2.

Speaker 1:

Okay. So where Monica counsels Richard to not take to take the less money, I think that that is, like, that is such a profound episode where, in particular, Richard is talking to the entrepreneur who took too much money and was forced out of his own company. And, man, that thing just it it it is absolutely nails. I don't know. Is that Dick Costello?

Speaker 1:

Who's writing that? It's like someone has clearly lived that who is writing that script.

Speaker 2:

Yeah. It's so spot on, but I think also, at least for me understandable Well, this gives us, you know, 18 months, 24 months, saying, well, this gives us, you know, 18 months, 24 months, 36 months of runway. I mean, I agree with you, Brian, but it's also tough to to walk away from.

Speaker 1:

It is tough to walk away from. I also love I mean, among the many lies in Silicon Valley, we haven't touched our previous round. I always love that one. Oh, we haven't touched our previous round. It's like, okay.

Speaker 1:

Can I see audited financials out of curiosity? Just like no. I mean, I totally believe you. You haven't touched your previous round, but you probably have. Because And and it may be

Speaker 2:

also I mean, you know, think of them as, like, AMC or GameStop. Maybe they're they're selling well or or, you know, they're selling pieces of the company while the pieces of the company are worth a ton.

Speaker 1:

Yeah. If if you could call AMC and GameStop.

Speaker 2:

The

Speaker 1:

The things I did not have on my 2020, 2021 bingo card. Exactly. So I so could I mean, I think one question is, of course, could the outcome have been materially different? I do feel like they didn't mention Heptio in the, in the article, McCluckie's company, that he sold the VMware for what, Steve, 500,000,000, something like that? And Yeah.

Speaker 1:

I'm I'm sorry. I am certain. You know, I I so for reference with Craig had told me when they were starting Kubernetes that he was gonna start a company and sell $500,000,000 for VMware. I remember thinking, like, dude, it's harder than that. And then that's exactly what he did.

Speaker 1:

I'm like, like, I don't know. It's not hard. I don't know. Wow. Okay.

Speaker 1:

I guess. What? But I all I mean, I guess one question I would have is how lasting what they built is versus was that just built for an exit? Certainly, my read. Yeah.

Speaker 1:

And looking at who who

Speaker 2:

did make money on Docker in clearly, the pub or it seems like the public clouds are making money on Docker.

Speaker 1:

Well, I okay. So

Speaker 3:

Well, I'm glad you mentioned Heptio because, again, like, that is a great example. It's a good marker for what people were willing well, at least what was valued of what people were willing to pay for, really, support services. I mean, that was largely consulting services wrapped up in, like, a little bit It

Speaker 1:

was yeah. It was. But

Speaker 5:

I think, Apple support services. Made a bunch of money off the phone because people keep the RAM upgrade, consistently for the past 5 years to try and be able to cram in a few more containers.

Speaker 2:

Got it. You're you're saying that the RAM vendors should have acquired Docker in that way. Right.

Speaker 1:

It would've been great. That makes sense. Analysis and acquisition of Docker. Oh, wait. Oh, that does make sense.

Speaker 1:

Of course. SK Hynix is realizing it. This is like well, I'll tell you who I think made money on Docker. I think the user made money on the did the developer. Ultimately, we had just like we've got, you know, VC funded taxi rides with Uber and Lyft, I think you had VC funded technology that developers are using that made a had made a does and and is making a big difference in the way people software.

Speaker 1:

And, yeah, no vendor monetized that. That went into customers' pockets and maybe it should have. You know? Maybe it was like a it's a donation to those who actually use technology. That's kind of why I view as like the actual money maker on on this stuff.

Speaker 2:

So Solomon So this is

Speaker 8:

the new one cent from AT and T that has to go to developing new technologies?

Speaker 1:

That exactly. Right.

Speaker 2:

So Solomon Solomon is the Robin Hood taken from the VC he's giving to all of us. And he's playing he's playing the ultimate trick.

Speaker 1:

Yes. Apparently, although those VCs apparently do not include Peter Fenton because they recapped the company and I think Benchmark is gonna get theirs. I mean, I'm I'm very impressed with how they recapped the I mean, not impressed. The recap is terrible. But

Speaker 8:

Although by that logic, a lot of other startups have become much more productive because they have these tools. Like, their portfolio over a career is probably better because they invested in Docker even if they lost all the money.

Speaker 1:

Oh, there you go. Yeah. That's it. There there's your glass half full. Yeah.

Speaker 1:

But I do think I mean, I think actually, Erin, this is an important point because especially as I'm

Speaker 8:

No. Sometimes you just have to grow peanuts for 1 year to restore the soil.

Speaker 1:

There you go. Peanuts. That's right. I I'm reading so many of these these, you know, books and memoirs and so on from software from the nineties, and we need to get out of this phase. But the just being reminded about how difficult everything was on when in this all proprietary world.

Speaker 1:

And I'm sure there are plenty of startups that that were able to move faster and better because of Docker. So that's where I don't know. I think that's that's, you know, the the the, it was the friends we made along the way.

Speaker 2:

So so that's the Docker legacy. Yeah.

Speaker 1:

It is the Docker legacy. And what's important I think I think Docker's got nothing to be ashamed of, Docker Inc. I do wish that I because I told Solomon so directly in April 2014, I wish they'd have the self control to take less money because I which is very hard. Hard to hard to ask, but you do wish that a company could have that kind of

Speaker 4:

Well, I'm I'm disappointed that we haven't developed into who who invented containers.

Speaker 1:

I I mean, I think the isn't the Internet, like, tops 20 or whatever? I think, like, made

Speaker 4:

it'll touch a flame fast when you start.

Speaker 1:

Sorry, Dan. Were you weighing in on on who invented containers?

Speaker 10:

Yeah. I mean, Plan 9 namespaces. Like, so much of what goes into Docker and and, you know, containers and c c groups and all that stuff was directly inspired by plan 9 and and what they did with namespaces in that system.

Speaker 1:

There we go. That and and that is probably a good note on which to end. I think we're, again, trying to keep this to, I I don't know, about an hour. Adam, do you have any or does anyone else have any any closing thoughts? Or there's some people who actually just Well,

Speaker 2:

we had a couple of folks who just got in. So any anything any last thoughts you need to insert?

Speaker 1:

Just wanted to there's an interesting concept about VCs, sort of funding development for other companies.

Speaker 11:

For more reading about that, there's a concept called the peace dividend of the SaaS wars. There's a blog post floating around that's there's a couple

Speaker 1:

of them from sort of medium sized VCs that are interesting to read just on that subject of the big copy

Speaker 11:

big competition between all the clouds encouraging all this open source that then later on is sort

Speaker 1:

of having a ripple effect through the rest of

Speaker 11:

the economy. So if you're interested in that

Speaker 1:

sort of thing, that might be something to read up on. That sounds

Speaker 2:

floats above.

Speaker 1:

Yeah. That sounds great. And, of course, a reference to the peace dividend after the end of the of both World War 2 and the Cold War. So definitely an interesting analogy. Yeah.

Speaker 1:

That's thank you for the the point of that. That sounds great. Yeah. It's a good read. I'll I'll try to DM it to you or

Speaker 11:

something once I find it again.

Speaker 1:

That sounds great. Well, again, we're recording all these. We've got, we've got show notes, and, you know, thank you to the those folks, especially Carl. I'm sure you're around, but Carl's done a lot of show note work, which has been terrific.

Speaker 8:

Thank you to Docker, the company that's ch rooted the future.

Speaker 1:

Exactly. That's right. Showed it brought Chirrut to the masses. Thank you, Docker. Well, thank you, Docker, for for for, all the technology, and I think it made a lot of impact on a lot of folks.

Speaker 1:

Definitely, at least from our perspective, our ringside seat, gave us things not to do. I don't think if we ever have an oxide conference where we're banning words, I think will someone please remind us that we are the animals are walking upright? Don't think we'll do that. Alright. Hey.

Speaker 1:

Thanks, everyone. Thanks for joining.

Speaker 2:

Yeah. Thanks, everyone.