
Ep. #16, It’s The Future
In this episode, Edith and Paul discuss Paul’s blog post titled ‘It’s The Future‘.
transcript
Paul Biggar: Not a single person pointed out that Katy Perry did not write the song "Call Me Maybe."
Edith Harbaugh: Do you think nobody wanted to admit that they know female pop stars?
The biggest insult of all is not to be insulted, it's to be ignored.
Paul: It was published because it was content marketing, and that's why we argued with the PR folks that we should publish this.
It doesn't age that well because Docker moves so quickly.
Edith: Like a ship.
Paul: Hi, I'm Paul Biggar, founder of CircleCI.
Edith: I'm Edith Harbaugh, CEO and co-founder at LaunchDarkly.
Paul: And you're listening to To Be Continuous, a podcast about continuous delivery and software development.
Edith: You can get in touch with us anytime at our Twitter handle, @continuouscast.
Paul: The show is brought to you by Heavybit. To learn more, visit heavybit.com. And while you're there, check out their library, home to great educational talks from other developer company founders and industry leaders.
The Story Behind "It's the Future"
Edith: Paul, so you had an article that literally, when you published it, I looked up and Oren, who sits next to me at Heavybit, and Max, who sits next to me, were literally laughing out loud and quoting different lines. This was an article called "The Future," and I have to say it was extremely funny, and it was one of the reasons why we started doing the podcast, 'cause I was like, "This guy is really funny. We should start having a conversation." So I'd love to hear more about why you wrote that article and the reaction to it.
Paul: Yeah, cool. I feel we should do many more episodes that involve a lot of compliments and flattery at the start of the episode. It's a pretty good way to start an episode.
Edith: Well, anytime you wear the LaunchDarkly t-shirt, you just always look good.
Paul: Yeah, the article was called "It's the Future," and it was basically a kind of a send-up of the Docker ecosystem at the time, and it doesn't age that well because Docker moves so quickly.
Edith: Like a ship.
Paul: Basically the premise of it was that there's a lot of complexity in the Docker ecosystem. And it was sort of a screenplay between two characters, one of whom was asking, "I'm trying to launch an app, how do I do this?" And the other, who was a complete Docker and container and "it's the future" fanboy.
And it went back and forth of, like, each time that they would ask something, there'd be a new rabbit hole to go down. And actually the image that went with it was a rabbit looking down a rabbit hole. And yeah, it was very, very well received.
It was funny, though, that when I first wrote it and I showed it to a couple of people at Circle, they didn't know that I had written it. They thought it was just, like, a thing I was sending around.
Edith: Oh, it was clever. So you were A/B testing it?
Paul: No, no, not deliberately. I just, you know, had been ambiguous with how I was phrasing it.
Edith: Wait, did you deliberately leave your name off so that they weren't tainted, or was it just accidental?
Paul: No, they just didn't look closely enough. I posted it in, like, cut and paste into a Slack pastebin sort of thing. And so my name was not... It was just, like, they thought it was like a pastebin meme that was going around or something like that.
But I had showed it to our PR people, and the PR people said, "Under no circumstances can you publish this." Because, you know, there's others in the ecosystem that were mentioned, potential partners and actual partners and that sort of thing. So they said, "You can't publish this," and then when everyone in the office was saying, like, "Oh my God, you have to publish this," so I sat down again with PR and negotiated its release.
Edith: Was this an external firm or your internal PR?
Paul: This was an internal PR. So there's our regular PR person, and there's someone who was providing sort of external PR services at the time, and then our social media person, and the three of them sitting on one side of the table while I'm trying to negotiate the terms of release.
Edith: Well, so they must have felt strongly, 'cause I'll say usually people don't want to disagree with a CEO or a founder.
Paul: Yeah, so I mean, definitely in the first draft there was...
Edith: Oh, is there a legendary first draft?
Paul: I have a first draft somewhere. But basically some names were changed to protect the guilty. To protect people who are our partners or who we wanted to be our partners or something like that.
Edith: Which in Silicon Valley is basically everybody.
Renaming the Partners
Paul: Right, exactly. So I mean, you can't really tell from the situation, but I think the people who got kind of the... There was a lot of "someone is dead." Right. In the article. It's like, "Something is dead, you need to use Docker now, or you need to use CoreOS, or you need to use one of the new shiny things."
Virtualization was one of the things that was dead, and maybe owning bare metal was dead or something like that. So, you know, companies that were involved in that were first named, and then we had to pick other companies who were not going to be our partners but who were vaguely in that space.
And it's funny, because one of the founders of another Docker company had emailed me afterwards. So I wrote that post called "It's the Future," and then I wrote a follow-up post called "It Really Is the Future," where I said, you know, "The last post was just a joke. We actually think this is how technology is going."
And I mentioned, like, you know, twelve companies that were in the Docker space, whereas before I'd only mentioned like six of them. And someone wrote to me and said, "Thanks so much for mentioning me. We felt kind of left out in your previous article that you didn't make fun of us."
Edith: The biggest insult of all is not to be insulted, it's to be ignored.
Paul: Right, yeah, yeah. So we ignored some companies deliberately who actually wouldn't care. Big companies are at least much bigger than us, and so being ignored would not be a massive insult to them.
And replaced it with, I think Vagrant was one of the victims, and Heroku was one of the major ones that was like, "Heroku's dead, you need to do something else," and of course the punchline is, well, if you haven't read the article, read the article. But the punchline was at the end. It's going back to, you know, Heroku is the simple way to deploy all this instead of having basically a distributed systems degree to be able to launch a web app.
Edith: Well, that's why I think so. Oren and Max, who sit next to me, had both come out of Heroku. Of course, yeah. They thought it was absolutely hilarious.
Paul: Well, I sent the article to some people I know at Heroku, and a couple of minutes later the Heroku Twitter account retweeted this.
Edith: Was that your intended distribution strategy, or was...
Paul: Yeah, the way I distributed the article was I emailed everyone I knew in the industry and said, "You might enjoy this."
Edith: So basically you're saying you're a social media master.
Paul: Social media master, and then there was the Hacker News, and release it at the right time on Hacker News. I think you're supposed to release it at, like, ten a.m. on a Thursday or something like that. But there was, like, twenty minutes after I published it on Hacker News where it had, like, one or two upvotes. And then all of a sudden, I guess, you know, after people had read the thing, 'cause it was a long thing, all of a sudden it went quite viral.
Edith: Supernova.
Where the Idea Came From
Yeah. So how did you come up with the idea to write it?
Paul: It's hard to know how, like, you know, something spun into your head, but there's... There's how I thought I wrote it, and then based on the feedback that I got, I realized that there was another vector.
Like, it just popped into my head one day. I said in my head something stupid about, like, you know, how Docker was annoying or something like that. And then it just, you know, evolved into a, you know, he-said, she-said sort of situation.
I read afterwards, someone tweeted at me to say, "Oh, that reads a lot like something I wrote on the Joel on Software forums, like, ten years ago." And I went back and read it, and I had read this like five times before.
And so it was... This will probably ring a lot of bells about the hammer factory abstract protocol method or whatever. A guy goes into a hardware store to buy a hammer. And the shopkeeper says, "Oh, we don't sell hammers anymore, we only sell hammer factories." Like, "All right, I'll have one of those."
And it's like, "Well, actually we don't sell hammer factories, now we only sell hammer factory factories." It's like, "Oh well, okay, give me one of those." And it's like, "Well, actually people want it more than hammers. So, you know, we have a tool factory factory." And it goes on.
Edith: It's kinda like the Austin Powers joke where, you know, he says he has a factory that makes tiny factories.
Paul: Right, right, exactly. It was factories of factories of factories, and it went on for a long time and it was very deep. And it's almost exactly... I mean, it's not the same joke, but it's the same layout and it's the same setup.
I think that was probably a large influence. I didn't realize it until he mentioned it to me.
Edith: Well, I mean, you know, it's a funny pattern, which I think people have been doing. You know, there's always a straight man. And the question I've been trying to figure out for a while is, am I the straight man or are you?
Paul: Right. I don't know if you've seen "Who's on First."
Edith: Yeah.
Paul: Yeah, exactly. So someone else said, you know, "Oh, this is basically Who's on First." And I went to look at "Who's on First." I'd never seen it before.
Edith: You'd never seen it?
Paul: I didn't grow up in America.
Edith: Oh, I thought you grew up in New York.
Paul: I lived there for a couple years as a teenager, but, you know, not really. So yeah, I watched "Who's on First," and "Who's on First" is, like, one of the funniest things I've ever seen. It's amazing.
Edith: Yeah, and people only know the first part, but it just gets funnier the deeper you go into it.
Paul: Right, it's like seven minutes long or something.
Edith: Yeah, well, because the jokes just pile up on each other. So I think that's what I liked about the article, is that you name-checked so many things, they were right there.
Paul: Right. I was talking to some large company, which I'll remain nameless, who wanted to be our partner afterwards. And they want to be our partner entirely because of the article. Like, the article had been sent around in an internal list, and people thought it was hilarious.
But when we went to talk to them, and I was chatting to one of the senior guys there, and he was saying that there was so much technical detail in it that it actually looked like we knew what we were talking about. And it wasn't just a joke. And that was kind of the point.
It was supposed to be... There is a lot of technical detail in here. And I think certainly at the time it was very accurate for what the ecosystem was like.
Edith: Yeah, I mean, part of why it was funny was it wasn't just a word salad of plain, like, anybody can look at Crunchbase and be like, "Here's 17 names," and combine them.
Paul: So a number of people sent me their versions of that. So they wrote a version for something else. And so one guy wrote it for front-end technologies. And it was super clear that he knew what he was talking about, and the joke really, really held up and it worked well.
Edith: Yeah, well, we have a similar... It's a cartoon, but... So there's somebody who's trying to ask somebody else how it worked, and they're like, "Well, you know, we have a Firebase backend that then commits..." And then the punchline is, "So you just tell me to fuck myself."
Paul: Right, right. I know the one.
Edith: Like, 'cause once our designer was asking how we collected emails on our website and we're like, "Well..."
Paul: So someone else sent me an article that was... I mean, it wasn't as bad as a collection of random words, but it didn't make as much sense for the joke. So I think that one didn't get published.
Content Marketing on Purpose
Edith: Yeah, so did you set out for it to be, you know, a widely circulated piece that would get you partnerships, or is it just scratching an itch, or a combination?
Paul: I mean, it was very intentionally, you know, a content marketing piece. So I mean, it wasn't like I sat down to write, you know, content marketing. But it was a, you know, "This is funny. We should publish this in our blog."
And obviously the point of the blog is that people come to the site and that they get retargeted and all this sort of thing, and that they eventually buy our product. Even, you know, through direct mechanisms of, you know, clicking over to the site.
And I think at the bottom of the piece we had a call to action. "Do you want to use Docker and continuously deliver that shit?" Which was one of the quotes from the piece.
And so yeah, I mean, it was published because it was content marketing, and that's why we argued with the PR folks that, you know, we should publish this.
Edith: So how many drafts did you go through? I mean, so some of the jokes are really funny. Did you have people...
Paul: Well, so by the time I published it, nothing was funny. Like, I can read it now, I can't see a single joke in there.
Edith: Oh, what is... There's the one about, like, BSD, and you're like, "It's San Francisco."
Paul: Right, BDSM. Right, yeah. So I recognize that they are funny in a purely academic sense. But I can't laugh at anything. Because I've read that thing like a hundred times.
Nothing is funny after you've read it a hundred times.
But the first time and the second time, the third time I read it, I could sort of see it, that it was funny. It went through maybe ten or twelve drafts. So it was a couple of drafts at the start, and then there was the draft to remove the partners' names and be rewritten with other people's names in its place.
And then there's a couple more drafts. There's a draft to make it ambiguous as to the gender of both participants.
Edith: Oh, that's very clever. I didn't realize that until now.
Paul: Yeah, so there is one place where the... What's the opposite of the straight man in the...
Edith: The comedian? The yuckster?
Paul: No, the straw man or whatever.
Edith: Oh, the straw man.
Paul: That's probably not the name, but the person who the joke is on.
Edith: That's the straight man.
Paul: Oh, that's... Okay, right.
Edith: So the straight man is the one who's just like, "Oh, who's on first?"
Paul: Okay, okay. I can see that. So that person said, "Oh, I'm more of a distributed system guy now." And that was the only kind of gendered reference in the whole thing.
And obviously there's a big thing at the moment of, you know, trying to make sure that everything is not just, you know, put out from the male perspective or, you know, the assumption that everyone is that everyone in the tech industry is male. So we deliberately removed all kind of gendered references and rewrote some of the jokes or things to make it ambiguous as to who is what or whether there was even any genders involved at all.
Edith: That's fascinating. So, like, the reason why I asked is I saw the producers of Airplane speak, and they say that they had tested every single gag in that movie on studio audiences. So I pictured you, like, putting out draft after draft and, like, sharpening each line.
Paul: So there was a little bit of sharpening, but for the most part it went out as it was.
Edith: So it was more like dulling as you're taking out the pointed...
Paul: No, no, so actually when we changed the companies that were involved, it got funnier. So we had the opportunity to kind of refine the joke and to make it a bit more appropriate for the companies that got mentioned.
Some of them had been a little bit of a reach before because they were, you know, in my head I was thinking of the people who were our partners at the time. And it had been a bit of a stretch to make the specific joke I did. And then when we had to remove them, I could search around a little bit and find the ideal company that could be poked in just the right way.
Edith: So the article was a lot of planning. It was a content marketing piece.
Paul: Well, I mean, it was a thing I wrote in about an hour that we then planned, how do we market this to get the most effective outreach, and that kind of thing.
How People Reacted
Edith: So what was the reaction? What was the thing that surprised you most?
Paul: I guess the thing that surprised me most was that a lot of people's reactions were, "Oh yeah, fuck that Docker shit." So most people who read it perceived that we meant it and that Docker was complete crap.
Edith: Interesting, because I think it was clear that it was very tongue-in-cheek.
Paul: In fact, when we launched it, we had a tongue-in-cheek disclaimer at the top, which we've since removed because it wasn't really necessary. But there was a lot of people complaining when we wrote the "It Really Is the Future." Nobody hated "It's the Future." A lot of people hated "It Really Is the Future."
Edith: Interesting.
Paul: So the idea that you're going out and taking the piss out of this new company, this new technology, yeah, people love that. But you go back and say, "Actually, you know, we really do think it's good," people hate that.
Edith: Oh no.
Paul: Yeah, so the haters really hated on the second one.
Edith: But the second one I thought was very innocuous. The second one was just kind of a rah-rah Docker-is-good piece. There wasn't a dialogue. It wasn't funny.
Paul: It wasn't that funny, no, but there was a discussion of why do people hate new stuff.
Edith: Oh, I like that line.
Paul: Yeah, the kind of people who would look at the Sistine Chapel and say...
Edith: "Why do we need to paint the ceiling?"
Paul: Yeah, like, "I have a fresco in my house already," or, you know, something along those lines. And, you know, obviously no one likes being dismissed as mere curmudgeons.
Edith: It's funny, so I wrote a kind of controversial article called "Kill the Staging Server." And I actually have a follow-up planned, which is "The Staging Server Is Dead."
Paul: Long live the staging server?
Edith: Yeah.
But I already got, like... So I got a lot of really good feedback, but the one tweet that made me a little bit sad was this person who just had a link, is like, "Call me when you start doing serious software development." And they're like, "Okay, you don't have any rebuttal or rebuke or, like, feedback." You're just like... And by the way, shouldn't software development be fun?
Paul: Right. Paul Graham has this article that he wrote years ago, like, you know, the six levels of how to disagree with somebody, and, you know, the main one is, like, refute the main point of the argument, and the simplest and least effective or least convincing one is, you know, dismiss it out of hand. Or, you know, insult the person who says it for, you know, who they are, or something along those lines.
So it's very easy to ignore those. Like, I don't even notice that hate anymore.
Mistakes Nobody Caught
The feedback that I found really interesting was the people who are mentioned in the articles who were kind of making jokes at each other about the article. And especially, so there's a couple of things I got wrong, but that people perceive them as part of the joke.
So one of the things I talked about is this guy, Diego, who in the article he worked for CoreOS and he wrote etcd. In fact, he did neither of those things. I had met him when he was, like, interning at CoreOS or something like that, but he didn't actually take a job there.
So Alex Polvi, you know, tweeted at him, it's like, "Oh, I hear you work for us now, I expect you to be, you know, here on Monday morning." But the thing is, I didn't know that. And so people thought it was just all part of the joke.
And so, like, people corrected, like, every single minor detail that they thought was part of the joke. They're like, "Oh, I love how you made, like, this joke here about how Diego works at CoreOS." And there's various things like that. But not a single person pointed out that Katy Perry did not write the song "Call Me Maybe."
Edith: Oh, wow.
Paul: Yeah, so that was, like, an intentional...
Edith: That's Carly.
Paul: Exactly, yeah.
Edith: Yeah, I'll own that one.
Paul: I think I got that wrong in the first draft, but after that I left it in deliberately because, like...
Edith: Do you think nobody wanted to admit that, like, they know female pop stars?
Paul: I don't think anyone knew. I think that people, like, "Call Me Maybe, Katy Perry, that sounds right."
Edith: No, no, no.
Paul: Like, a hundred thousand views, no one pointed that out.
Edith: Maybe they either didn't know or were embarrassed to admit they knew.
Paul: Maybe, maybe, yeah. But every little tiny bit of something else that was ever so slightly off...
Edith: What else?
Paul: There was another one, but I don't remember what it was. But it was kind of along those lines as well.
So I did know that he didn't work there, and I didn't know that he didn't write etcd. I think he did design Raft. I'm pretty sure that he designed Raft, which is the protocol in etcd. I don't remember the other thing.
Docker Since Then
Edith: Well, so you said you already felt that the article was a little dated. Like, so how do you feel like things have moved on since then?
Paul: Well, so the whole Docker thing moves really, really fast. And so there's probably a couple of, like, companies or technologies that were mentioned in there that have changed their name or been acquired or gone out of business or, you know, something along those lines.
There was also, I remember when I wrote it there was a bit of a war going on between CoreOS and Docker, which, you know, they've since buried the hatchet and become friends. But there's probably a couple of technologies there that have changed.
CoreOS released the Tectonic stack sometime after that, I'm pretty sure. And so that, you know, would probably get a little name-dropped a little bit, or, you know, differently, or something like that.
And I imagine that, you know, maybe Docker has had several releases of Swarm at this point, that might have gotten a bigger thing, but, like, the things that were big at the time were the things I mentioned. And those are not necessarily the things that are big now, just because of how early Docker is and how quickly that technology is moving.
Edith: Yeah, so, you know, this is a friendly podcast, but what do you think of Docker now?
Paul: I mean, I think Docker is awesome. It solves a real problem that people needed solved. And, you know, the complexity that results from that is real. It's getting managed and it's getting managed better over time, but it's real because you're writing distributed systems.
Distributed systems are hard. No silver bullet technology is gonna make distributed systems easy. It can just make them more manageable.
And Docker, certainly very useful as part of making distributed systems more manageable, but, you know, there's still a distributed system. Abstractions leak, and abstractions will continue to leak, and you still need to understand a whole set of new technologies.
Chasing the Dragon
Edith: Wow. So are you gonna write a new "The Future" in a year or so, or do you think...
Paul: I tweeted about this recently. I think that's the best thing I've ever written. And yeah, I think I probably can't do it justice again.
And I remember looking at the Google Analytics for this article. And so the follow-up article was bigger than the biggest thing I'd written before. And then the first article was like five times bigger than that or ten times bigger than that.
Edith: Wow.
Paul: Like, first day, a hundred thousand views. Yeah. So, like, you can't really follow that up. And I spent a couple of days thinking about, like, "Okay, what am I gonna write next? What am I gonna follow, you know, this wave with?" But, like, I quickly realized, you know, you can't be chasing that dragon.
The likes and social media and content marketing dragon is like, I mean, it's kinda like the problem that people have with Facebook, you know, or maybe with Instagram, but when you're constantly looking for, like, people to like your thing and you're looking for affirmation from social mechanisms instead of just what you're working on, everything's negative from that point.
Edith: Yeah, that's interesting, 'cause you're not that active on Twitter.
Paul: Right, right, exactly. And part of that is just accepting that I'm not really that good at Twitter. I mean, I occasionally say some stuff that's interesting or funny, but I'm very happy to not be chasing likes or affirmations or retweets or whatever the thing is.
Edith: Yeah, they're very seductive. I do content marketing too. I write articles, and I wrote first one about fundraising. You know, I compared fundraising to an enterprise sale.
You know, you have your champion, you have a process. I got a decent amount of traffic.
I thought I would get a lot more, but then I wrote something which I considered kind of geeky and technical, which was why people were doing continuous delivery. Much more popular than the fundraising article. And then my last one, which was on staging servers, which I thought was intensely narrow of an audience, it got the most traffic of any of them. And it actually got comments.
It got shares, it got comments, and I was like, "Whoa." I think I just proved that something I thought only really I cared about actually had this chord.
Paul: It tends to always be that way. There's something that you don't expect to be as big as it is. I remember someone telling me that... What is the name of that company? There was a Google Docs competitor that got acquired by Yahoo at some point.
Edith: Tumblr?
Paul: No, no, a Google Docs competitor. Anyway. The biggest piece of content that they ever wrote was something about Ajax. And it was written, like, in the early days of Ajax and all the JavaScript-y stuff. But, like, it wasn't a company that made JavaScript tools, or it wasn't a developer tools company, but that was, like, the biggest thing that they had ever written, and the thing that led to the most inbound traffic.
If you're judging things by, you know, the business metrics that you're trying to generate, you have to be looking at leads or sign-ups or, you know, money in the bank at the end of the day, rather than, you know, whether someone retweets your stuff or whether there's a like there, and you have to be looking at the business metrics, not the social metrics, and not the things that make you feel good.
Edith: Well, I mean, they go hand in hand.
Paul: Often they go hand in hand, and I think often they don't.
Edith: This is going to go all philosophical.
Paul: Well, I mean, so for example, a lot of the things that had the biggest retweets and that sort of thing were fundraising announcements. But people don't follow fundraising announcements to buy your product. At least not at the rate, like... As content marketing goes, a fundraising announcement isn't all that good. A viral blog post, much better.
Edith: Yeah. I'm happy that we got to talk about this. Is there anything bad that came out of this article?
Paul: I don't think so, really. I don't think that there's any negativity from any of the people involved. There was a couple of haters on Twitter, but there's always haters on Twitter.
There's a couple of haters on Hacker News, but, you know, that's what Hacker News is now. Unfortunately. But yeah, you know, definitely I'm glad I wrote it, and I'm glad I followed it through, and I think it will be hard to ever top that one.
Thanks for listening to this episode of To Be Continuous, brought to you by Heavybit and hosted by me, Paul Biggar of CircleCI, and Edith Harbaugh of LaunchDarkly.
Edith: To learn more about Heavybit, visit heavybit.com. While you're there, check out their library, home to great educational talks from other developer company founders and industry leaders.
Subscribe to Heavybit Updates
You don’t have to build on your own. We help you stay ahead with the hottest resources, latest product updates, and top job opportunities from the community. Don’t miss out—subscribe now.
Content from the Library
How AEO/GEO Differs from SEO (and Whether It Matters)
Developer content expert Adam DuVander explains how founders should think about being visible in LLMs (and in search).
5 Social Media Strategies for Early-Stage Startup Founders
How should early-stage founders approach social media (especially “tech Twitter”)? The experts at Leadtail explain.
Why Do Developers (Actually) Hate Marketing?
Your startup needs to market to developers, but don’t developers hate marketing? Here’s how to reach them.

