Software Misadventures - Build the scary stuff | Bryan Cantrill (Oxide)

Starting point is 00:00:00 The most difficult moment for me in the history of Oxide, when we got our first term sheet. I'd been working so hard for so long on getting that, and we were raising a lot of money on nothing. I mean, you're out of deck, three people in a deck. I'd been thinking so much about get the term sheet, get the term sheet, get the term sheet, that when we actually got it, I wasn't ready for what was emotionally on the other side of it, which is like, oh shit, I got to go do all this. And I don't know how to do any of it. We talk about like how great it is to have a team.

Starting point is 00:00:34 We didn't have a team. We had three people. I didn't know how to build any of this stuff. You know, like I knew how to build aspects of it, but like, oh my God. You've done a lot of hard things. In fact, I would say out of school, you wanted to work on operating systems and kernel development.

Starting point is 00:00:51 Not every engineer's dream, I would say. Like, hey, I'm going to work on the operating system or the kernel, spending time at Sun and then joined. And now Oxide, that's also a very hard thing to do where you're actually building a new server and integrating software hardware together. So for many people who are out there who would be interested in doing hard things but may not have that slight push that they might need, do you have some piece of advice?

Starting point is 00:01:16 Piet Hein was a mathematician and poet and so he's got these little poems that I think are really important to any technologist solving a hard problem. And one of them that I love is problems worthy of attack, prove their worth by fighting back. Welcome to the Software Misadventures podcast. We are your hosts, Ronak and Gwan. As engineers, we are interested in not just the technologies, but the people and the stories behind them. So on this show, we try to scratch our own edge

Starting point is 00:01:54 by sitting down with engineers, founders, and investors to chat about their path, lessons they have learned, and of course, the misadventures along the way. Okay, so Brian, you're the co-founder and CTO of Oxide Computer Company, where you're building a new type of server. We definitely want to get into that, but I thought a fun place to start would be your Twitter description. So you have the nom de gré, pardon my French, of being the kernel of data corruption.

Starting point is 00:02:24 How did this come about? You know, that is a good question. So I, you know, I think for whatever reason we were, this is when Twitter was, of course, much younger and a lot less serious. Although I actually guess Twitter has now again become, has become much less serious in a much less illusive way. So we were kind of talking about what it would look like to have, you know, a Nondegare is this idea of, and this is especially when you get these kind of small insurrections,

Starting point is 00:02:55 where you get the, especially when, I mean, actually these are tragic wars, where you have these kind of these youth armies that have, that adopt these kind of florid names of war. And we were kind of joking about what would look like for. And I think we were we were arguing between, I think, general protection faults and kernel data corruption. So the general protection fault being a kind of fault on an x86 CPU and then kernel data corruptions. And I think we were in the midst of actually a pretty bad bug. I wanted to say that OS 1028, which is a Jira ticket number. You know when you can memorize the Jira ticket numbers

Starting point is 00:03:35 that stick with you, you know what I mean? You always want to be like, take someone, a software engineer in their dotage and be like, okay, rattle them off i know you got them rattle off every jira ticket whose number you can every bug i need whatever you know whatever banking system there's always a couple and os 1028 was an absolutely brutal data corruption issue that we had and i want to say that was about the same time that would have been like in 2012 and actually I had I had first encountered Twitter as a when they were initially a Sun customer way way way back in the day this is like 2007 okay yeah and so I actually spent time

Starting point is 00:04:18 with Twitter in its earliest days because they the the performance, they had this thing called the fail whale that would come up pretty frequently because they really deeply did not know what they were doing. Like most companies. Like most companies, but more so. I mean, I think that you had a, and this is, again, this is in 2007. So you kind of set the scene. You have this, there had been this huge dot-com boom and then bust. So the dot-com boom, still, we have not seen anything exactly like it. I mean, of course, this is when I was coming up, when I was first out in Silicon Valley.

Starting point is 00:05:05 But, and you know, hard to know if we will ever see something that has that exact complexion again, because there are certain, in particular, you couldn't, there was no working remotely, right? So that dot-com boom was very much a, was very related to the geography. And, you know, the, I think the analogy that many rightly made is with the gold rush out here in 1849. And, you know, one of the things you appreciate about California is this has been, this is an omnipresent kind of cycle in California where people kind of rush in and most of those people leave in disappointment. Or actually, I think that we kind of have this idea that people get kind of wiped out. I actually think that the results are just kind of mediocre for people.

Starting point is 00:05:58 I think it's like no one is destitute after the dot-com bust, but people also are struggling to find work and are moving elsewhere. So you had this kind of big rush out here and then this big rush out, kind of departing Silicon Valley. And kind of in the echoes of that, you know, the people that lower part of the stack, semiconductors, network slipping, computing servers, operating systems. I mean, that was kind of where that expertise was. And so when you had folks that were building an application on Rails, which is what they were doing in 2007, that was not like a Silicon Valley construction, even though they were geographically co-located with Silicon Valley. The level of technical aptitude, they did not have a high degree of technical aptitude. And Jack Dorsey was, I was super unimpressed with Jack Dorsey.

Starting point is 00:07:01 Not many people say that statement openly, by the way. Oh, God. More people need to know. He had absolutely no idea what he was doing, but had total confidence, total confidence. And in particular, he was certain that this was an issue. The Twitter issue was an issue with Sun, which is part of the reason I was there. I was there because I was there working at Sun at the time. So I spent 14 years at Sun Microsystems.

Starting point is 00:07:30 A customer of Sun's was this company, Joyant, that ironically I would later go work for. And then Joyant's customer was Twitter. So we were... Joyant had asked me, it's like, hey, can you come in and help them develop these performance issues? And it was taking hundreds of milliseconds to process a request. It was taking like, I mean, the number was like 450 milliseconds. And I remember thinking like, how is that even, what? 450 milliseconds. I'm like, well, okay, obviously there's like, we just need to find the IO there. I mean, clearly there's like, that's not compute, like that can't be. And they're like, no, no, it's all compute. I'm like, it's, it's all, that's a lot of instructions to execute actually. And what

Starting point is 00:08:15 we, and the thing that was neat is we were able to really use some of the, the, the tooling that we'd built this facility called D-Trace, in production at Twitter to understand pretty quickly what was going on. And Twitter was spending all of their time in mem copy and b copy. And what was happening was actually a bug in Rails. And Rails in a pretty hot path was instead of, you know, there was this kind of dry idea of don't repeat yourself. And the idea of like, it was somehow like quotidian or pedestrian to iterate over an array by indicating like the index from zero to the bounds of the array. It's like, no, no, no, you shouldn't do that. Like that is somehow inelegant. What you should actually do is just iterate through the array

Starting point is 00:09:10 until you fall off the end of it and then catch the resulting exception. That's a terrible idea. It's a really, and it's like, this is one of these ideas that's like, this is an idea from someone who is very much over their skis right this is someone who they think that this is clever because it involves like fewer characters or whatever but just because you can do it doesn't make it clever and it's like did you take a stop watch to this because if you did you would you would see that this is really slow. Because when you hit that exception path, exceptions should be exceptional, right? And everything about a runtime environment is going to be steering you towards those exceptions being exceptional.

Starting point is 00:09:59 And anyone writing the code that deals with that exception is going to be within their rights to be like, hey, this is an exception. Like, we got a lot of work that we can only do right now. And in particular, if you want to have any prayer of knowing where this exception comes from, I've got to walk your stack frame right now. I've got to walk your entire stack frame. And we've got to actually resolve that symbolically because you presume we want that to be in Ruby, not in hex decimal. And so the B copy that it was doing was the very suboptimal symbol table iterations that it was doing and copying all this information into this galactic payload

Starting point is 00:10:36 that it would then pop up one frame and then be like, oh, great, I ran off the end of the array. It's like, okay, well, you just, do you realize that you just burned a forest to the ground to actually figure this out? Yeah, next time maybe we can use the array beds. But I came out of that experience. And part of the reason I mentioned the geography, I came out of that experience being like, you need a lot more Manhattan technologists here.

Starting point is 00:11:06 Because at the time, I was spending a decent amount of time in New York and also very geographically focused, right? And if you were in financial services, like that was all happening in New York. And the level of sophistication in terms of technologists in New York with respect to this problem was incomparable. Oh, wow. Because money's on the line. Right, right, right. Yeah. Money's on the line. Also, this looks like a ticker plant. This is not that complicated. I mean, it is, but this is a problem that has been solved before. And what you, Twitter, need to go do is you need to go find someone at JPMorgan Chase, at Bloomberg, at Goldman Sachs. Pick your large financial institution where money is on the line and you need to convince them to move out to California and you go solve this problem actually properly. Which is ultimately more or less-ish what they did, but they got there much slower than they should have because they had this have because they had this kind of residual arrogance that like, we must know best. It's like, you

Starting point is 00:12:08 really don't know best. You actually don't know what you're doing. And that, I would say that, that kind of inappropriate confidence really came from the top with respect to Darcy. So I came out of that being like, I am never using this thing ever. These guys are a bunch of jokers, but by 2010, it's like, all right, I actually am going to create an account on Twitter. Unfortunately. There are secondary usefulnesses of it. I'd say confidence, necessity, but not a sufficient skill to do things well. I mean, it is definitely a problem. This is part of the reason why I really encourage folks to, you know, I think we,

Starting point is 00:12:47 formal education, it can be a mixed bag. So it can be extremely expensive, but I, part of the reason that I really encourage people to complete formal education, and this is not the only way to get this, but the value of formal education is that you get your ass handed to you. It's really, really important that you, everybody needs to get their ass handed to them intellectually. And like, if you haven't had your ass handed to you yet, then you simply need to seek out a more arduous domain, or you need to seek out a center for people that are going to attract. You need to go to where the people are who intellectually kick your ass. Because that humility is a really important part of your own kind of emotional development from like 18 to 22.

Starting point is 00:13:38 And again, it's not the only way to do it. There are other ways to do it. But if you come out at age 22, being like, I understand everything. And I am, I am the smartest person in the room. It's like, oh, shoot, we screwed up. Got to go back. Sorry. You got to go to grad school. I don't know what else to tell you. We didn't do our job yet. And you at some point in there from like 18 to, you know, your early 20s, you need to have a moment of like almost breakdown, right? You need to have a moment of like, I understand nothing. I am, I am a mouse. I don't under, actually, I'm not even a mouse because I don't even understand

Starting point is 00:14:19 how mice work. Like I, I don't, you know, everything all, you know, through up until, you know, I was 18, people are handing me all of these abstractions. And now I've been taking all of these abstractions apart and learning that I actually don't, that there's so much to not know about them. And, you know, when I, the world is so complicated and how can I possibly ever contribute? It's like, okay, good. Now you're in the right frame of mind. You're in that early 20s mindset. It's like that period of time, whether it was formal education or otherwise, has done its job of absolutely breaking you, and now we can rebuild you. Yeah, that's very relatable. I think that's how I got out of pure math after very fast and realized,

Starting point is 00:15:02 yeah, the game is a bit rigged in a sense right that you're only competing on just pure intellect or that has like such outsized impact on the final result and i realized you know what i think i gotta go find something that's you're competing on like multiple things where like hard work is like you know an important vertical instead of just you know how smart you are um sure but then that whole experience you know that like for the rest of your life you know it it changes the way you carry yourself right you don't you're not going to walk into the room being like well these are the smartest people in the room because like actually go back to like that like you know my

Starting point is 00:15:40 abstract algebra class and you know or whatever it is you, whatever it was in pure math, I'm out. So I think that's an important part of your development. Interesting. So I think in other interviews you mentioned, so you have teenagers. How are you thinking about the education path for them? Everyone's got to find their own path. My kids are, you know, they, they're going their own way. I'm secretly jealous, or maybe not so secretly jealous of, you know, there are kids that are super into, into tech at a young age. That's the way I was, right? I, I, I loved horsing around with computers and writing programs and so on. And that's not who my kids are. You know, my kids have got their own thing going on. And so on the one hand, from a spectrum

Starting point is 00:16:33 of a parent, you really do need to encourage kids to go their own way. I think too many parents really put their thumb on the scale and they create someone who's like later in life miserable because they are like, I actually, in particular, I definitely worry about this with computer science. Right. Where I think that with computer science, you know, there's been so much economic opportunity, which is not the case, by the way, when I was coming up, like when I was studying computer science in 1992, a country was coming out of a pretty deep recession in 1991. I knew a lot of people that couldn't find work. And certainly you would not study computer science if you were interested in work, because there was this very deep idea that all of that knowledge work was going to be done offshore. That would not be done in the US. It's amazing to me, the echoes with AI today. And it was such a deeply held belief, by the way, that I didn't like question that. But I also, this is what I love. I had

Starting point is 00:17:41 fallen in love with computer science at that point. And that's what I wanted to go do. And I'm just a big believer in allowing people to find the thing that is going to speak to them at that level. And then you kind of need, as an adult, and this can be really hard, as an adult, you really need to then encourage that. And you need to balance it at some level, but you know, it's like my, my oldest, it plays, it plays college baseball, right? And that is like, his orientation is 100% around playing baseball. This is obviously insane. I mean, at some level, when he was much younger, I spent a lot of time kind of trying to talk him out of this. And when at some point when he was young, you know, sixth or seventh grade, he's like, Dad, I know at some level it's not very realistic, but it's my dream. And you know when a kid tells you something like that?

Starting point is 00:18:39 Like that's kind of a, like that should be a wake up. You know, from when a kid is telling you like hey you're like this is my dream and you know you're it really makes you kind of question a lot of things and like in fact people did try to talk me out of being in computer science and you know i mean the way that one worked out is I had won the generational lottery and people were trying to convince me to rip up the ticket. And, you know, boy, that would have been a mistake. So I think when you give advice to young people, as an adult, you very much want them to be able to feed and shelter themselves, you know, but people will figure it out. And I think that you, you've got to let them find

Starting point is 00:19:33 their own way. And you don't want them to enter adulthood thinking like, man, I had this other dream that I really got talked out of. And I let someone tell me to do something that I actually didn't want to go do. And like that stuff was like, that can, you know, really chip away at the way you think of yourself, at the kind of the essence of who you are. And I just think that like, people should be able to figure it out on their own. And of course, it means like, as a result, like that means like there's going to be some, you know, some big potholes along the way. And as a parent, it's really hard to watch that, but you know, you, that's also a big part of, you know, of learning and letting people kind of figure it out on their own. So long answer to the question,

Starting point is 00:20:18 but I really think that, that people need to be given the space to figure things out on their own. I'll send that to my Chinese parents, you know, see what their follow-ups are. But sorry, by joke, by joke. They're actually very supportive. They're very non-traditional in that sense. But anyway. Well, and I think that, you know, and there was this battle hymn with the tiger mom a couple of years ago.

Starting point is 00:20:43 And I mean, obviously, it's like a very different approach. It can be a very different approach. And I get some of this, I mean, there are aspects of that that I think are like great, you know, of having those high standards and so on. Not that I don't have high standards for my kids, but like my high standards for my kids, I got to tell you, are all around character and decency. There I am like really, like I'm pretty flexible on them finding their own route in life. I am not flexible on how they treat other people. Like that is extremely important to me. I'm not flexible on how they, on their own sense of integrity, like not flexible. And they know that.

Starting point is 00:21:21 Like they know that like if I am, if I'm taking a test, if I'm taking a surprise pop quiz, if I flunk the quiz, like there will be some level of disappointment. But if I cheat on the quiz and get an A, I will be metaphorically dismembered. Like that will be like not acceptable. I would much rather you like take the, do the best you can. And I do worry that when we emphasize results over all else, that we have told, you run the risk of generating adults that don't behave very well.

Starting point is 00:22:02 And we got a lot of adults that are not behaving very well right now, honestly. This is, you honestly. It's embarrassing how many companies have had really serious scandal and how many companies are, we've got a lot of people who are at Oxide. I mean, it's almost tragic to me, the number of people who apply to Oxide, not because of what we're doing technically, not even necessarily because of what we're doing in terms of organization and so on, but simply because we are a company with integrity. People are like, oh, this is such a relief to have a company that talks about, that believes in integrity. You're right, that should not be a competitive differentiator. That

Starting point is 00:22:42 should be all companies. That really... Were there specific events in your life that shaped this sort of non-negotiability on integrity? That's a good question. I mean, yeah, there were. I mean, not something I've really talked about publicly, but suffice it to say that like when I was growing up, there were secrets in my house that and they were really, really corrosive. And, you know, ultimately my parents are divorced and that. But in my case, that was like great news. I mean, it was very much in vogue in the 80s but in my case it was like all right thank god um you know people would be like oh god i'm so sorry about the divorce like no actually the divorce is terrific that's that is actually and you know that did for sure shape like my i i think that i do believe in honesty and transparency. And I think at a very, very deep level,

Starting point is 00:23:51 I would rather people be honest about who we are and what we are. And I think trying to create these facades ends up being really deeply corrosive. And, you know, I think that what that facade can be covering can ultimately end up being completely corroded, right? I mean, it's like, there's no ultimate like integrity there and the integrity in kind of like a structural sense, you know what I mean? And then when the whole thing comes apart, like there's nothing. And so I guess I've always believed and probably is forged to a certain degree from having kind of the what happened to my kind of nuclear family. And it's just really important to me as a it's important to me from a family perspective.

Starting point is 00:24:41 It's important to me from a company perspective. And I think also, I think the other kind of element that is to me, and this was very much, this does not come out of kind of childhood trauma and much more from just the being instilled in these values from my own parents, the kind of the idea of a societal responsibility in terms of what you're doing and societal engagement. And like, you're not, your objective in life is not self-maximization. Like you need to serve, everyone should seek to serve a broader community and society. And like that element is a very important element in terms of, you know, the way, certainly the way I think of myself. It's like I've got responsibilities that are much bigger than myself.

Starting point is 00:25:29 And we all do. And, you know, I guess it's a sad commentary on our divisive age that that's thought of as like old fashioned. But to me, that's really important. Both the parts that you mentioned, it's important to, in other words, like struggle a little bit to give, like get your ass handed to you. The importance of integrity in this case. How do you go about instilling this within, let's say, for example, your kids or even within the company? Because you can say these statements and they make sense logically, but it's very easy to go take the easier path, which is not necessarily what you want. But then how do you go about

Starting point is 00:26:05 encouraging to a point where it becomes practice? Yeah. And that's a great question. And the reality is it is going to be, that's going to be forged in difficult moments because that's where like the rhetoric is going to actually meet, you know, facts on the ground. You know, it's like just talking earlier about, you know, encouraging kids to find their own way. That is way easier said than done. If you talk to a parent of a toddler and you say, and this is, let's, let's say before toddlers have become difficult. So let's say this is like, you know, you talk to like a parent of an infant of a six month old, right? And you ask them, you know, what do you want for this newborn that you have? Like virtually any parent will tell you, I want them to be happy. That's what I want. And I mean, they're not going to say like, boy, I really, I want them to get into

Starting point is 00:26:56 Bowdoin or, you know, I want them to like, you know, work for Morgan Stanley. They're not going to have these kind of, they're going to be like, I don't know, maybe some truly sociopathic parent. I guess that would be a true warning sign. Like, actually, I want them to get into Bowdoin. You can actually call Child Protective Services. But mostly you're going to get, like, I want them to be happy. And I'm like, write it down and,

Starting point is 00:27:25 and put on a piece of paper where you can find it when they're 17 years old. And they are doing something that, that where are they telling you like, Hey, I actually want to take a gap year after high school because I'm miserable right now. And I don't know where I'm going. And I really, I want to go work in construction for a year, which is like, you know, the parent that probably aspires for the kid to go to Bowdoin, this is probably going to be something that they're not expecting to hear. And then you're going to be like, go find that piece of paper and listen to that kid, you know,

Starting point is 00:27:59 because it's hard. Like that's the moment, like that's the moment where, and it feels like the, when it feels like the stakes are high. And so, you know, those are the moments in which you really, you know, who you are in those moments of difficulty. And it's hard to say like, well, you should, you need to go through these moments of difficulty, but that is what's important. And you need to get in, and then that kind of needs to happen repeatedly where, and you know, in that moment of, you know, where the kid is contemplating, you know, do I, do I cheat on the pop quiz or do I kind of like take the low grade? It's like, that's the moment, like that's the decision point. And if you make the wrong decision when it's low stakes, when they're in a, you know,

Starting point is 00:28:47 a seventh grade math class, like that's an opportunity to now like really correct that and say, you know, the kid gets caught cheating on the pop quiz. You can really explain to them, like, here's what's important to me and here's what's important to us and here's who our values are and that moment can kind of shape them right and but it you know part of that adolescence is the those kind of moments where you can shape that as people get further and further on like they're getting more and more baked from a character perspective i i believe very firmly that it is, it's never too late in almost every dimension, right? It's like, I have discovered that like in my, you know, my junior year that like, you know, I was studying chemistry and I'm kind of miserable, but I just took my first computer science class. Oh my God, I love it. It's like, it's not too late. You know, it's not too late. You, you met at a data

Starting point is 00:29:41 science bootcamp, right? It's not too late to say like, hey, this data science thing is super interesting to me. Like I want to, I want to go back and I want to learn about, like, it's not too late. It is never too late to learn something new, learn about a new domain. That said, I think from a character perspective, I don't know if it's like, if it's too late, but things get like baked. And I,, you, you get to the point where, you know, you've got someone who hasn't what hasn't necessarily lived with those same ideas and has been, you know, whether it's taking a shortcut or whether it, just like losing track of, of who they are or why they are, right?

Starting point is 00:30:25 I mean, the example that I always use for this is like, you know, Uber was a company that was, that really deeply lost its way, right? And that came from the top, right? And Travis Kalanick and the kind of the culture they had at Uber was very, very, very broken in a lot of ways. And there were a whole bunch of things that were like deeply, deeply wrong. And if you go reread like Susan Fowler's, you know, blog entry on her experiences at Uber, you know, what is going through the kind of the mind of the people that

Starting point is 00:30:58 she is describing in there? What is going through the mind of the software engineer that is implementing Grayball, this thing that they had developed to evade regulators in Portland? Right. Like, how does a software engineer end up there? but I at that point like you really need you know is it too late for those folks like I guess I believe in redemption so probably not but holy moly do you need to have a moment of like I am down the wrong path and I need to deeply correct the way I think of what's important. And, you know, how did I, it's like, you know, Anthony Lewandowski who was at Uber and the, you know, Anthony Lewandowski should, should be in the slammer, right. And was pardoned by, by Trump, kind of the, another like an embarrassing moment for tech where peter teal was lobbying for him to be pardoned but you like you look at levandowski and the the kind of the arc of his life and it's

Starting point is 00:32:16 crooked he's very bright but he's deeply deeply deeply crooked and and i'm not to pick on him although like the justice system has already concluded that he's guilty of crimes. This is the guy that took stuff from Waymo and then brought it over to Uber. That's right. Okay. Okay. Yeah. He took stuff from Google and he lifted a bunch of Google IP. But that is just the tip of the iceberg. He very famously or infamously inside of Waymo at Google had a safety, safety is job number two. And they put an engineer in a Prius that, and he on the highway at a time when he'd been explicitly forbidden from doing it and damn near killed the engineer.

Starting point is 00:33:04 And like, and this is what you're talking about. Like, okay, now we're actually talking about real lives, real people, real lives. And that is someone whose priorities are really deeply screwed up. And it'd been deeply screwed up for a long time. Like this is not something that, this is the thing about all of these scandals.

Starting point is 00:33:23 Like it's never a surprise to the, to people who've like known someone for a long time, because it's like, no, you look at that. You can find these late, these like much smaller transgressions a long time ago when they're growing up, when they're in school. And then when they are, and the transgressions are just getting larger and larger and larger, and nobody is coming in and saying, stop, this is wrong. And I think that we in software engineering, it's kind of gutting to me that if you had told me in 1992, that software engineering would be an epicenter for people behaving badly. I think I would find that I'd be gobsmacked. I think I'd be just shocked by that. I'm like, how? But it's like, unfortunately, like the good news generational lottery ticket, right? I mean, there's a lot of economic opportunity, but that economic opportunity has attracted economic opportunists and people who aren't in it for the same reasons that I'm in it.

Starting point is 00:34:27 And it shows up in their behavior and misbehavior. So in this case, like you said, back in 92, getting into software engineering or computer science was not a thing that was very popular. In fact, people would be dissuaded of doing that. It's like, hey, don't do this. This is for someone else or we'll offshore these jobs. Today, it's quite the opposite. And along the way, there have been these cycles of culture changes in Silicon Valley where, like I said, initially it was a lot more about semiconductors and hardwares. And then it became a lot more about just apps on tops and then mobile and whatnot. You've seen this all the way through. What are some things today

Starting point is 00:35:05 that you would want to be different and you would want to bring back like the, let's say, the old Silicon Valley culture? Yeah, I think it's, well, I really think that it is important to bring back the why. Why are we here?

Starting point is 00:35:23 Why are we doing this? What's the purpose? What's the meaning? And not that, and I don't want to lionize or romanticize kind of Silicon Valley's past, but they're, because I think it was easier then too, in a couple of key regards. But I think it's very important that we have that why. And my big belief has always been, I'm not a huge fan of Halt and Catch Fire. It's a little bit mixed. It's fine. I mean, there are aspects of it that I like, aspects of it that I don't like. I'm a much bigger fan of HBO Silicon. But there is a great line in it where one of the protagonists says to the other, of computing, we are not the thing. We are the thing that gets us to the thing.

Starting point is 00:36:14 But computing is not the thing. And I think that that's an important thing to not lose track of, that what we are building isn't the thing. It should be the thing that allows us to get to the thing. Now, I think for some folks, you can easily make an argument that like, okay, but for folks that are at that application level, it's like they're much closer to like the thing, right? And if you are at, you know, an Uber or a Lyft, you know, providing mobility for people that don't otherwise have it, like to allow them to do, you know, to get to a job or visit a loved one or do the grocery shopping, whatever it is. So, I mean, it's like, it's easy, I think, in a lot of these to begin to make those connections. But I think it's important that we do

Starting point is 00:36:58 make those connections. And I think that that is what we lost for a while. I think we have it in pockets for sure. We don't have it universally. And I think that there are plenty of companies that don't know why they exist. And there are plenty of people that don't know. I actually don't know why I'm doing this. It's like, all right, well, go figure that out. Because I think we collectively do our best work when we have that deep sense of purpose. I understand the need to build a business. I think that's great. I think that there's a danger when that actually drowns out everything else. And I think that when we are only interested in building a business,

Starting point is 00:37:46 then it's like, okay, well, what if a shortcut presents itself where you can book this revenue in this quarter, even though you know it's not sustainable, or you've got these kind of opportunities to take these little business shortcuts that are actually undermining the broader meaning of what you're doing. And when you are only focused on the kind of economic outcome, you don't make the right long-term decision. And I think that that's what we saw. We have seen, and maybe this is another way of phrasing it, we have seen episodes of absolutely rampant short-termism around here that I think are really problematic. And it's like we actually need to be focused on the long-term and on innovating for the long-term and on serving our responsibility to society, which is to put these incredibly powerful tools, like the most powerful thing humanity has ever built is computing, I think.

Starting point is 00:38:50 You know, I think that you can, I mean, obviously, like we have built lots of amazing stuff. Our ability to build a lot of that amazing stuff at its root has our ability to build computing devices. Like that is what we have needed to build better everything. And I still very much believe in that. And I'm very long humanity. I'm very pro-humanity. I think it's, which is, it feels like a little ridiculous to say, but I feel that because another thing that I really don't like that I've seen several bouts of, I don't like the kind of this doomerism thinking. And we have seen it in, it was offshoring in the nineties.

Starting point is 00:39:38 It was Y2K in the late nineties. It was actually, there was surprisingly little doomerism in the dot-com bust because it actually just felt like, well, this is actually doom. Doom is actually here. So it's like, there's no real point in being a doomerist because it's like,

Starting point is 00:39:53 we're all living it. And not that there's not legitimate stuff there, especially when you look at things like climate change, it's really important to treat it very seriously. But like, we're going to figure it out. I mean, we actually are going to, and it will be, it'll be hard, but it will also be meaningful to go figure a bunch of this stuff out, so let's not, you know, I, I think sometimes we get, we underestimate our ability to figure this

Starting point is 00:40:18 stuff out and that's a real, a real problem. That's it. That's it. Okay. Talking about computing, we wanted to switch gears and talk about Oxide computers. Before we talk about Oxide itself, well, one, why start Oxide computers? Yeah. Why start a computer company?

Starting point is 00:40:37 Yeah. Great question. So the origin there is that I knew, I've always known that I wanted to start my own company at some point. And just because I felt like it would be something that I wanted to do, I felt like my career would be incomplete without starting my own thing. And not just for building a business, but that too, but also, and especially as I, so I've used Sun for 14 years. I was at Joint for another nine and Joint was bought by Samsung in 2016.

Starting point is 00:41:16 And I knew that the next stop for me would be doing my own thing. I also knew that I would be doing my own thing with my co-founder, with Steve Tuck. I had worked with, and this is where getting kind of your earlier questions around character. Steve and I had worked together for nearly a decade at Joyent. And we felt like our collective character was tested all the time in a startup. I mean, part of what's interesting about a startup is that like you have it, there are just crises all the time. Right. And those crises really everyone gets a chance to reveal kind of who they are.

Starting point is 00:41:57 And I knew that Steve and I deeply shared principles and values. And I that's who I wanted to work with. And we had a great, shared values, shared curiosity and enthusiasm, but also really different experiences. Steve came up on the go-to-market side. I had worked in sales at Dell and a joint. And obviously on the engineering side, this is what I want to go do. I want to go do something with Steve.

Starting point is 00:42:25 So that's actually what I knew before it was a computer company, certainly before we had a name, before we had an idea. I had a co-founder and a desire to start something. And we would talk about the things that we wanted to go start. And we had just a bunch of very bad ideas is what I would, it's kind of how I would characterize them. I mean, it's just like. Can you share a few that you could, that you can remember at this point? You know,

Starting point is 00:42:54 I did promise to Steve that we would never speak of some of these bad ideas. Let's call Steve and get his permission. No, no, no, no. Okay. I feel that I can, I will not speak of Steve's bad ideas because I will leave. That's totally fair. That's totally fair.

Starting point is 00:43:11 Okay. I do feel that I can speak of my own bad ideas though. I feel like that. I feel like within my, but to give you kind of a concrete. So if you have kids or maybe you don't, if you do any kind of like recreational athletics, one of the things that you deal with is this program called Team Snap. Team Snap is, so this is just a way for organizing a team and often using youth sports, but you'll also use an adult recreational sports. It's free. It's terrible. It's absolutely terrible. It's like,

Starting point is 00:43:47 and like, we see these things, right? You see things that are like, and again, often free, often have, they have forced into extinction something that came before it. So like, this is like Craigslist and classified ads, right? Like Craigslist is terrible. I mean, it's just awful, but it's also like, just, it can't be dislodged because the, the price is right. And Craigslist itself is like, we're kind of happy being like bad. Like we actually don't care. Like, well, I, okay, I guess this is just going to be bad. I get it. Like, you know, and, and I mean, what has actually happened over, over time is that the, for different elements of classified ads, different pieces of it became different products. Yeah. It's crazy. It's like, if you want to sell a car,

Starting point is 00:44:34 it used to be that if you sold the car, that would be in the local classified ads way back in the day. Then there was a period of time, like if you want to sell a car, you really had to do it on Craigslist. Like there wasn't really another way to do it. And now if you want to sell a car, there are a bunch of apps that you would go use to sell a car. So fine. But the TeamSnap is one of these, TeamSnap is like the Craigslist for, it's just, again, it's just bad. It does this kind of like the, and I'm sorry for anyone that works on TeamSnap, but also

Starting point is 00:45:01 like anyone that works on TeamSnap, like. Please fix it. No, it's like, you know, that if there's someone who works on TeamSnap, but also like anyone that works on TeamSnap, like... Please fix it. No, it's like, you know that if there's someone who works on TeamSnap, and first of all, part of the problem is like, I don't think there is anyone that works on TeamSnap. So I would almost like love for you to get the letter from someone like, no, I am a software engineer at TeamSnap. It's like, okay, we've discovered one. I think what's much more likely is like, no, they got rid of all their software engineers

Starting point is 00:45:21 like eight years ago. And I just feel if you were talking to someone at TeamSnap, they would say like, oh, my God, it is so much worse than you can possibly imagine. This is one of these things where it's like, you know, when it like looks bad on the outside that it's like way worse on the inside. So I'm sure there are wild stories about TeamSnap. It's like, no, it's like it's owned by this person who's just like just interested. I mean, they're clearly not interested in delivering a quality experience. And in particular, as someone who, you know, at that time, especially coaching kids in like youth baseball, there's a called Game Changer, which allows you to follow games online, which is great, right?

Starting point is 00:46:10 So you can have like a grandparent can follow their eight-year-old's baseball game kind of like pitch by pitch. Like that's great. That is a great experience that they can have that was not possible when I was a kid, right? But Game Changer like overly monetizes this and charges a fortune for it.

Starting point is 00:46:26 And it's like, there's a whole bunch of things that's a mess about that. If you could take like Game Changer and Team Snap and then you combine that with like league management, like umpire scheduling and stuff like this. It's like, you could like, I thought you could like, there's a real problem to go solve there. But this is ultimately a bad idea, right?

Starting point is 00:46:46 This is a bad idea for a bunch of different reasons. First of all, this is where you get to like, okay, so it's like you can go solve this problem that would be meaningful that people would love. But you start doing the math on it and it's like, you know, how much, you know, there are, and part of the value of this is that you would like want to tightly integrate it to the very kind of specific sports. You'd want to focus, I mean, I know baseball, you would focus on youth baseball. Maybe you expand that to youth soccer, but you start doing the math. I'm like, well, how much is there out there and how much are people willing to actually pay for it? And you kind of go, you know, what does the sales process look like to kind of the

Starting point is 00:47:21 league? And then like, also like how technically interesting is this? Like, is this enough to hold my own attention for a decade? No. Is this enough to be able to, because the other thing I kind of realized is that the best moments of my own career were working with like an extraordinary team. Like I, I love heist movies and that like the heist movie is my organizational model. Like that's what I love. I love I love heist movies and that like the heist movie is my organizational model. Like that's what I love. I love watching a heist movie. I love being a part of a heist. Like I want the safe cracker, the getaway driver, the demolitions expert. I love that, you know? And I, and I love the like, also like keep it weird. You know, this is like, I grew up watching the A-team when I'm a

Starting point is 00:48:02 kid. Right. And if you look at the A-team, it's weird. You know, you've got these folks that are, you know, B.A. Baracus is afraid to fly. And you've got these folks that are kind of coming from different backgrounds or whatever, but they're all bonded by this, like, this mission that they're going to go on. Like, I love the A-team as an organizational model.

Starting point is 00:48:19 Like, that is what I love. But like, you can't, like, you don't call the A-team for, you know, to go like grocery shopping or whatever. You know, and there was a great, you know, in the 80s for the television show, which was, by the way, of course, much worse than it is in my memory, like, 18, Knight Rider, you know, 1984, watching those two back to back was like, I mean, you would have like, you know, as a whatever, an 11-year-old. You're just like, I am just absolutely in heaven. I go rewatch those like with my own kids. And in particular, like sat through like the premiere of the pilot of Knight Rider, which I remember as being like a transcendental religious experience. Like I touched God when I watched the pilot of Knight Rider as a 10 year old or whatever I was. My kids watch the pilot and they're like, we get like four minutes in. And I'm like,

Starting point is 00:49:16 am I watching the right thing? Like this is terrible. Like this is so bad. And I remember my kids who were the same age I was when I watched it. My kids, my older was like, dad, I know this is important to you, but do we have to keep watching this? And I'm like,

Starting point is 00:49:32 we do not. We can turn this off. Very polite. Very polite. I'm super polite. Super polite. But I'm like, you know what?

Starting point is 00:49:40 No, you do not need to keep watching this. This is, this is awful. They'd be much less polite to me now as teenagers. They'd be like, bruh. But so anyway, the A-team as an organizational model, though, I loved. And in that kind of intro of the A-team, they would say, you know, the intro was like, if no one else will take your problem, if you can find them, you can hire like the A-Team.

Starting point is 00:50:08 Like, that's what I want. Like, I want, you know, if you can find them, if no one else will solve your problem. That's not like a team snap replacement. You know what I mean? Like, that's just not, that is not an A-Team moment. Like, that is that, and, you know know there's nothing wrong with that and like maybe at a different like stage of life or whatever i would that'd be totally fine that'd be great like no no i want to go solve this problem and i'm happy to do it with like two people and not you know

Starting point is 00:50:37 20 or 30 or 40 or 50 for this period you know this what hopefully will be multi-decade period, I wanted something that was big, that needed the A-team. And I knew I wanted to do it with Steve. And then at the same time, we were enduring these problems at now then Samsung, where we were hitting all of these problems running at scale and realizing that what we were trying to do, which is to deploy infrastructure on top of Supermicro and Dell and HPE. And we had very sophisticated software, right? Sophisticated software that was at the kind of the... We had our own OS kernel, like I've done kernel development my entire career. We had sophisticated software. There are real limits on what you can do even with arbitrarily sophisticated software when you don't

Starting point is 00:51:31 actually control the entire system. And if you look at what Google and Facebook, Meta, Amazon do, they control the whole system. They are not Dell and HPE and Supermicro customers, right? And you kind of add all of this up and, you know, especially the rejection of the bad ideas. And then I think also the other thing, I mean, in fairness, the other thing that as we were beginning to think about like what we want to go do, and I was reminding myself of the email address of a venture capitalist that I had dealt with over the years. And his last email to me was, as a reminder, great getting lunch with you today. As a reminder, I will fund literally anything you put in front of me. And that's important, right? To have that sense. And I

Starting point is 00:52:20 think it's actually really, really important for a startup. How would you go big? What would it look like to go big? And what would it look like? What are you afraid of doing? And what does that look like? And I think that you, because I think sometimes, I honestly think that this is like part of the challenge, part of how I ended up in kind of the team snap rut, is ironically focusing too much on the business side of it.

Starting point is 00:52:52 Like what is a product that people will pay for? And then you kind of end up with something, it's like, well, this is a product people will pay for, but not, but this is not a world changing thing. The upside is pretty limited in that case. Like you're speaking- The upside is limited, exactly, right. And so it's like, what would you do? So you kind of want, as you're kind of doing the thought

Starting point is 00:53:10 exercise of like, you know, what does new company formation look like? And I think it's totally reasonable to have, because this is obviously the route that I took, it's totally reasonable for new company formation to start with, I know the people that I want to work with, and I don't know what I want to do yet. Like that is, to me, that, you know, I don't know how often people kind of talk about those very kind of first moments of a company, but there've been a lot of great companies that have been formed exactly that way. That's the origin of Compaq, by the way. You can argue that Compaq is the most successful startup in history in terms of the speed at which Compaq hit not just revenue, but hundreds of millions of dollars in revenue, the growth they had. I mean, they hit right space,

Starting point is 00:53:49 right time, right team, right everything. But when they were starting Compaq, all they knew is like, we want to do something together and we want to all quit TI. They were all together at TI. And there's a very funny documentary, Silicon Cowboys, where they actually interview Rod Canyon and some of the other founders of Compaq. And I mean, I'm sure this is somewhat apocryphal, but it also feels earnest where they're like, yeah, we were kind of debating that versus starting like a burrito joint. I think that there's something great about the fact that they knew they wanted to do something together. So I think there's a lot of value in kind of coming to the problem that way of we know we want to go it. And there's a very, I mean, if you're fans of HBO's Silicon Valley, there is, as there are in so many aspects, there's a very funny, unspeakably vulgar scene in the way that Silicon Valley is, but where Richard is encouraged by Russ Hanneman to do the thing that he actually wants to go do and not doing the thing that you that you think other people want you to go do but what is the thing that is like in your heart I think that that is really important and they and that can take some time to figure out and but this problem and again it was kind of that email back from the VC that kind of catalyzed this. I'm like, Steve, we need to go do, I mean, if we can get literally anything funded, if

Starting point is 00:55:30 we take out that fear of not getting funding, let's go do the thing that we actually want to go do, which is start a computer company. And, you know, we had endured so many of these problems together that he's, I mean, it's like, yeah, we should obviously go do that. But do you think we can actually raise, do you think we can get it funded? I'm like, yeah, I think so. I mean, this is when I was a dunce and thought that VCs would only, it's like this venture capitalist is only telling us the truth, obviously. And the funny kind of epilogue there is that, so I started talking to VCs and realized that like, this is interesting. Like it's contrarian, but the way as a venture capitalist, the biggest outcomes are when you're contrarian and right.

Starting point is 00:56:26 And that doesn't mean, I mean, contrarian and wrong is not great. So you want to be, but contrarian. Well, it's also not terrible because the other thing about venture capitalists to know is that like zeros are endemic. And a venture capitalist's greatest fear is not a zero because those happen all the time. A venture capitalist's greatest fear is that they don't fund. They pass on the company that is the next multi-billion dollar outcome. fear that you can go into any VC firm and in any VC firm, you know, once you kind of get the kind of the Wonder Woman truth lasso on them, it's like, what's the one that got away around here? Because there'll be one at least. And that one will come up pretty frequently because, you know, they, as they're passing on a company, some partner's going to be like, is this another Pinterest? You know, I worked very closely with a VC firm that had like, no, we had, Pinterest had told us

Starting point is 00:57:31 like, we actually want you. We were not going to raise from anybody else. If you get us a term sheet, we're done. Like we could have had all of Pinterest and it would have been firm making, but they talked themselves out of Pinterest. And it's like, forget the zeros. Like they forget their own zeros. Like they don't even know their own zeros. They, what they all remember is Pinterest, contrarian and right. They all remember, again, those ones that were contrarian and right that came through that were just too contrarian for the firm. And so when you are contrarian, like that gets VCs interested. And to a degree that can be really actually its own challenge is that VCs get too interested too early when you're like, no, no, no, we are not a good fit for you. And they think that like, you're like negging them and that you're, you know, it's

Starting point is 00:58:30 like, no, this is not a negotiation tactic. I'm not trying to, I'm not trying to get you excited about Oxide by telling you that you're not, by saying that you're not a fit. And they're like, okay, but now we're very interested. I'm not literally trying to tell you you're not a fit and so many of these firms would insist on going deep and then they would realize like oh you know we're not a fit we're scared i'm like i know i know you're scared i know like stop it like we're going whaling we're going whaling you want to fish from the pier there's nothing wrong with fishing from the pier actually they don't even want to fish in the pier like they actually want to go to the fish section in Safeway. Like that's what they want to go to. And it's like, great. I honor it. That's not what we're doing. We are, we're packing

Starting point is 00:59:15 the gear and going on a multi-year voyage to a distant ocean. And the success is very much not assured, but if it's, if, if, if we can deliver, if it's a win, it's a big, big win. And so we're kind of discovering that VCs are interested, especially so. The VC that actually sent me that email, though, by the time I actually got back in touch with him, I'd already talked to a bunch of VCs. So I already knew there was a lot of interest out there. And, man, I get maybe 30 30 seconds 45 seconds in describing oxide he's like wait a minute if you're starting if you're talking about starting a computer company i want absolutely nothing to do with it no everything except hardware and biotech

Starting point is 00:59:59 no and it was like okay yeah he's like He's like, I'm like, okay. And actually it was funny because I'm like, okay. One, like you sent me an email saying that you would fund literally anything I put in front of you. So that's obviously not, I mean, he was literally like, don't know what you're talking about. Like, just like, no, I don't, that doesn't sound like me. I don't think I would say that. I'm like, do you want me to send you the receipts? I mean, I got them.

Starting point is 01:00:29 And actually it was funny because I was like, okay, it's a phone call. I'm like, okay, sounds good. Like, bye. And even a VC, he's like, no, no, no, wait, wait, don't hang up. Don't hang up. I'm like, oh, okay, interesting. He's like, no, no, no. And he did, I mean, to his credit, allowed me to expand on it. And he's like, all right,

Starting point is 01:00:50 look, there's no way I'm going to fund this, but I do want to help you out. How can I help you out? Which is not, this is not an unusual or unreasonable, by the way, when people are starting a company and talking to venture capitalists. VCs will, if you are off thesis for them in some way, or they just don't want to fund you because you're just like, I'm just not going to fund this idea ever. And in that particular firm's case, they had just lost their shirt on a company that was like not wholly dissimilar. And a company that no one had heard of, which is part of their problem, but was not wholly dissimilar. And so there was no way they were going to fund us because it was just too close to a very recent zero.

Starting point is 01:01:32 On the one hand, like VCs, like a zero is not their greatest fear. They're not in the business to get zeros. And in particular, if they've had a recent zero, they're like, I actually, there's aspects of this problem that I am aware of that I would not have been aware of previously. And I actually know how hard this is. And you kind of need, there's a degree to which like, you got to hit that sweet spot of being able to delude yourself into thinking like, I can solve this. If you know how hard something is, you'll never do it, right? There's a sweet spot that you need to hit. With them, it's like they had had this very recent failure. But when he

Starting point is 01:02:11 asked, and I do tell people that are like, as you, if you, because I know you guys talk to founders and people that are looking at new company formation, if you're in, as you're talking to a VC that's asking, how can I be helpful? Have an answer to that question, you know, and know that like the answer to the question is not going to be like, well, fund my company. That's how you can be helpful. But because especially if like, look, I like you or I like this idea. It's just, it's not a fit for me. It's like, it's not, and VCs won't always ask, by the way, how they can be helpful because sometimes they're like, actually, like, I don't like you and I hate this idea. So like, I actually want to get, you know, so fine. They're not going to ask and you don't need,

Starting point is 01:02:51 you don't need to ask. But if a VC does ask, how can I be helpful? Like, how can I answer that question? And then in this VC's case, I'm like, you can be helpful by introducing me to the founders of this company that was zero. I want to talk to them. I want to understand. Like, I want to understand what happened. And that was a very, very helpful conversation. And in particular, at that moment, the thing I was most worried about was that I actually do want to understand technically what I'm getting into. And I want actually that particular founder, when I described what we were doing, he was also like a little bit,

Starting point is 01:03:31 a little bit, you know, grizzled. He's either, you gotta, this guy's coming back from the front. He's got the long stare, you know, he's about to like spend the rest of his life being a peace advocate. He's, he's definitely done. He definitely wants to get off, you know, he's off the, fresh off the battlefield, not to overplay the metaphor,

Starting point is 01:03:50 but he's like, what you're describing is a suicide mission. Wow. And I remember writing down suicide. And, but he's like, absolutely, you'll never succeed. Like, okay, well, this is inspiring. Thank you.

Starting point is 01:04:06 But in talking to him about kind of what had happened there and there were a bunch of things that were interesting there. One of them was the actual mistake that they made, which is a mistake that many startups make, by the way, is they thought they had product market fit before they did. So they had a very happy customer building something. Again, this is like not, this is like Oxide adjacent, right? So they were not actually, they kind of had the vision of doing something that was the size of Oxide, Raxco computer. But that vision had gotten pared way down as they were in, and they were implementing. And they were basically down to something that today you would call a smart nick really focused on security and they had a customer which is great they like great we have product market fit and they built out this pretty elaborate go-to-market team they actually didn't have product market fit and the what they had was

Starting point is 01:05:03 one customer that was great even that one customer didn't want to market fit. And what they had was one customer that was great. Even that one customer didn't want to buy that many of them because they were so happy with it. I mean, they're like, I need like five of these things. I mean, I love it. But I don't actually need, you know, I just don't need very many of them. And then like that was a huge bank. It's like, well, how many of those are there? If they need like five a piece, like.

Starting point is 01:05:28 A handful, yeah. It was only a handful. And what they needed to realize, and again, it's hard to do in the moment. It's easier said than done. But like, you actually don't have product market fit. Like, you have some stuff that's promising. So, one thing we want to talk about, what you said. Wait, can I make a quick joke for some comic relief? So, once, you know, relief so once you know oxide you

Starting point is 01:05:46 know goes public right like a few years down the road and then you become a venture capitalist and then you know you have all these people coming to you you know pitching you really questionable ideas and you're like instead of saying like oh the tam is too small you can actually tell them the story about the uh about the baseball app you know everything i think it's gonna be excellent so um it's i think i would like that much more than being told the tam is too small so oh okay that's interesting so you because you hear like just a vc saying the tam is too small i think that was so i did a little bit of pitching when i was kind of working on like a startup idea a couple years back and then yeah so that was like a common thing. He's kind of like, yeah, the TAM is too small.

Starting point is 01:06:25 Like we can't invest in it. Yeah, interesting. And I think that this is where like, be helpful for venture capitalists to explain a little bit more about, you know, what their own economic model is and the kind of outcomes that they need for their LPs. And because I think also you've got businesses where it's like,

Starting point is 01:06:49 this is a, you know, there's potentially a business here. It's just questionable whether it's a venture. Or lifestyle business, as they say. Right. And I think that, you know, lifestyle business, I think it's kind of ridiculous. When I first heard that term, honestly know, lifestyle business, I think is kind of ridiculous. When I first heard that term, honestly, I did not know what that meant. And so I was like, well, it's like a mom and pop shop. Like, so like, my dad was a businessman, and like, he worked really hard. Like, so if you ask lifestyle business, like if lifestyle wasn't

Starting point is 01:07:23 easy, by any means, he was working super hard so why absolutely absolutely like what lifestyle are we talking about like we're talking about the lifestyle where i mean it's like you talk to anybody's like you know running a restaurant or you know and i i've got just total reference for those folks that are working hard and trying to you know provide a service to folks i think it's like yeah that lifestyle is like super hard work. I'm not sure what lifestyle we're talking about. Like, I think it's so insane. And I think it's also a disservice, honestly,

Starting point is 01:07:52 because when we talk about, oh, it's a lifestyle business and you have this idea like, oh, like this is easy. And then when it's not easy, because by the way, it's not easy, people begin to think, like, I'm a failure. Like, this is easy for everybody else. And it's hard for me only. And it's like, yeah, no. No, it's actually, like, hard for everybody.

Starting point is 01:08:20 It's actually just super, super hard. It's also really gratifying. And there are lots of reasons to do it. But it's super, super, super hard. It's also really gratifying. And there are lots of reasons to do it, but it's super, super, super hard. And I actually was just thinking about this recently. A friend of mine started a company in 2018 and we were talking about raising money. And he's like, oh, it's like, it's easy. Like, I just think it's easy. I'm like, it's easy. Okay. I don't think it's easy. Like, that's not the first word that comes to mind. And he's like, yeah, like, you know, venture capitalists can basically throw in money at me. And I'm like, okay. And this, this is a guy with like crazy pedigree. This is a guy where like a VC would look at the resume and be like, I'm funding this. Like, I don't even know. I don't need any other data.

Starting point is 01:09:08 It's like, well, you probably should get some. It's like, Nope, I don't need it. Like, because you know, this person, like super young faculty member at very prestigious institution with like the most prestigious background you can basically have like in a domain that's hot. Right. Like, yes, I'm finding this, no question. So he's like, yeah, this is easy. And I'm like, yeah, I guess it is easy. All right. And I remember at the time telling him, I'm like, be super careful about taking a lot of money Because the objective is not actually to raise money. That's not like the objective is to like have a good or service that people will pay for.

Starting point is 01:09:55 And by the way, your customers don't care at all about your pedigree. They don't give a shit. Like you can't be like, but wait a minute but i've i i this product is the result of our finest educational institutions and you're like this product doesn't work it's junk you know it's like the humane pin or the rabbit r1 like this is like and like i don't i don't care what the pedigree is of the team that built this. And like, that's where it's hard. And so he's like, I think it's easy.

Starting point is 01:10:37 I'm like, okay, it's not easy. And I'm like, what, what happened to them? And it's like, what happened to them is they raised a bunch of money. Then they raised a bunch more money. And we're talking hundreds of millions of dollars. And the thing is a total, clearly an absolute impact creator.

Starting point is 01:11:01 And their last remark is talking about how they are so excited to join name of much larger company. This is what people, companies always do when they're being acquired. It's like, oh, like we are joining forces with Google. It's like, you're joining forces with Google. Like, I think Google is acqui-hiring. I don't think that, like, I've always wondered, it's like, if you ask Google that question, what's Google's take on this? Is Google like, we are joining forces with this nanoscopic company that no one's ever heard of that took too much VC money. It's like, I don't think that that is going to be, that's not exactly going to be Google's

Starting point is 01:11:40 take on it. And I guess I kind of admire the optimistic thinking, but that's actually not what's happening. And it was pretty clear that this was not a great outcome. There are elements of this problem that are easier. Yes, he's been able to raise venture capital, but when you raise a lot of VC, you have now VC expectations for a VC level outcome. And now, you know, you've raised hundreds of millions of dollars. And now you've got a company that wants to buy you for $40 million, which should be great. It's like you've created $40 million for the value. But a VC that is so far below the preference stack, a VC is going to be like, would rather take the zero. And would rather not have a deal at all. And then you're trying to do right by your employees. I think that it's a mess.

Starting point is 01:12:38 And it's hard. It's really, really, really, really hard. And I think don't let anyone, we should not let others tell us that it's easy when it's not easy. It's actually hard. So you're building, I don't know exactly how to define it or how to say it, but you can correct me here, but like oxide racks, which have a bunch of servers in it, I think you call them sleds. Building hardware is super, super hard as opposed to just writing software on top of it. For example, were there times where people dissuaded you from actually going ahead and

Starting point is 01:13:11 doing it from saying, Hey, Oh yeah, for sure. I know for sure. For sure. I mean the, and yeah, when we were doing, we we've done our own so we built a rack scale computer and this is a rack scale design is something that various folks have taken swings at but the only people that have really pulled it off are the hyperscalers are amazon facebook google but it's not in a way that they can actually prioritize it so we yeah we did our own hardware design we did our own compute sled design we're not doing our own silicon so we're using amd silicon we did it we've done our own switch, which was also a huge, I mean, it's also a very complicated, and then we've done all of our own software from the lowest level

Starting point is 01:13:51 all the way up. And there are people that tried to dissuade us at more or less every level of that for everything at Oxide that we are doing doing you can find someone who's like doing that yourself is a terrible idea so you certainly on the hardware i mean that is like in some ways like that's almost like too easy because you it was easy to find folks who are especially in vc in VC who want to, they like just the margins of software. And they are kind of addicted to software companies and SaaS companies because they ultimately, VCs are funding a company to sell to someone else. And the high gross margins that you get with software companies are something that are very attractive to kind of that next investor. Are you building enduring companies that way? Maybe, maybe not. If you go to Marc Andreessen's Software is Eating the World piece

Starting point is 01:15:01 in 2011, and this kind of peace that's hailed as being very prescient. I remember at the time being like, are you kidding me? Of course, I mean, this guy like writes an essay that software is important. And all of a sudden he's like a genius. It's like, it's 2011. It's like, yes, software is important. Like, really? We needed Mark Andrews in the telescope. All right, fine. But the, and then I'm like, all right, maybe I'm being a little too hard and like you making some important points about like the kind of the shift in thinking. And then you kind of go reread it now. And I'm like, I'm kind of back to where I was when I first read the piece, which is like a bunch of it is obvious. software companies that address and dreeson tells us in 2011 that these are the software companies that are representative of the new world in which software has eaten the world do you know what

Starting point is 01:15:52 those companies are no zynga groupon foursquare you're just like zynga groupon and foursquare oh right okay yeah so somehow know, people talk about Fang. I don't hear anyone talking about Zynga for whatever, you know. And not to take anything away, I guess, from the hard work of the folks involved in that. But like those are ultimately easy in, easy out companies. And they didn't build enduring value because, and this is the challenge with software. Yes, it's easy to build software. You can build a high gross margin business. It's actually much harder to build a high

Starting point is 01:16:31 net margin business. So gross margin does not consider like the cost of sales. It's like only cost to get sold, right? It's price versus cost to get sold. And it's like, oh, we're just going to ignore the fact that we actually had like an enterprise sales team of, you know, 30 people pursuing this deal for nine months and, you know, going to conferences and all the AdWords and the marketing and all the, it's like, that's all like the kind of that SG&A, all like that, you'll see that in the net margin.

Starting point is 01:17:03 Nobody wants to talk about net margin, especially VCs. We like net margin is really important to us at Oxide. And, and every once in a while, you will, we'll get a VC that will mistakenly like leave the door open for my co-founder, Steve, to go on his extended rant about, about why actually we should be focusing on is net margin, not gross margin. And VCs are always like, okay, I actually, I don't want to hear it. But net margin is the thing that you should care about because that's the profitability of the business, is net margin.

Starting point is 01:17:36 But when you're not trying to build an enduring business, but just trying to kind of flip it to the next investor, you actually care about the metrics that show that growth, which is often gross margin. And you show gross margin and you show growth and you're like, well, this is like an obvious business. And it's like, well, maybe not, depending on what do these sales actually look like, especially in enterprise software where the sales can be pretty normal. So I think in terms of Oxide, in terms of where we had people dissuading us, we absolutely had people dissuading us on hardware because of

Starting point is 01:18:10 margin. It's like, no, but like, but margin. And it's like, okay, but if you actually, we believe that we can build a 45 to 50% gross margin business on this. And we think that we can also build them. We are going to be focused on net margin and we can build an actual profitable business this way. And VCs that really did their homework came to that same conclusion. But VCs that only care about those kind of software margins, like, no, I don't want to hear anything about software

Starting point is 01:18:41 because like, or hardware, because I only want, and we had firms tell us we only fund software. And it's like, you only fund software. Okay. So, so you, if NVIDIA comes to you in, in whatever that would have been, 1999, if, if, if Apple comes to you, if, if AMD comes to you, it means like, so you are not going to find any of these companies that we now rightfully hail. Like people talk like, oh, you know, TSMC and NVIDIA and AMD and Apple. Like these companies are amazing. Again, you know what they all have in common? And not just hardware. Like software is a very important component of all those companies. But those companies have

Starting point is 01:19:25 built systems and they built hardware and software together and that is the way so we for sure had people on the heart but then even going up the stack i mean we had plenty of people who are like no you can't do your own hypervisor or no no you can't do your own control plane like you should run this thing should run esx vmware esx and we're like yeah no it's not going to do that like that's not we are not seeking to become in part because we knew that if we can't build the whole thing we actually can't get to that margin that is a viable business because we're not generating the value you've've got to generate, like you need to generate more value than you capture. Right.

Starting point is 01:20:07 You've like, what you want to do is generate a product where your customers are like, wow, this product is so great. It's worth its price 10 times over. And the price with respect to COGS and the business is something that is viable and sustainable. Like that's a business. And that's what we were, in order to do that, you've got to be able to create that value all the way up. You can't stop it. Hardware is important. It's an important foundation, but you need that software to actually take the hardware and turn it into a real system. And so you really want to design hardware and software together.

Starting point is 01:20:51 That co-design is very core to what we're doing at Oxide. But we, you know, ironically, you've got people trying to talk us out of the hardware and then other people trying to talk us out of doing the software. It's like, OK, no, no, we actually were going to do the whole thing. And of course, we ultimately felt very lucky to discover our lead in terms of our lead investor at Eclipse Ventures. And Eclipse very much saw the like, no, hardware and software should be built together. And when you get to the right investor, it is great because you're completing one another's sentences. Because we were beginning to talk about the importance of hardware software together.

Starting point is 01:21:28 They literally stopped us. They're like, you know, we looked at the 25 biggest companies ever built and 21 of them involved both hardware and software. And it's like, OK, that's a good stat. Didn't know that one. But yeah, OK, that sounds good. Good, we're on the same. So, you know, they very much shared our worldview. The other, importantly, who else shares that worldview? The actual customers that want to deploy this thing. Because they are like, the problem that Oxide is solving is the fact that the hardware and software have been delivered by separate entities. So, yeah, like you take your VMware ESX and now you've got to go buy Supermicro boxes to go.

Starting point is 01:22:12 Or you're taking OpenStack and then deploying that on HPE or you're taking whatever it is. I've got my software's coming from over here. My hardware's coming from over here. And then I've also got a switch. I've got a switch operating system. I've got a control plane software. I've got all this software. I've got all this hardware. They're not coming from the same entity. And everybody is pointing fingers at someone else when the thing goes sideways. And from a, the perspective of someone who's deploying infrastructure on-premises and is that is extremely frustrating to have like boy

Starting point is 01:22:50 nobody seems to have my interests at heart like everybody is like when i'm on a call with vmware now broadcom i'm on a call with cisco i'm on a call with dell they are trying to prove to me that it's not them and show that it's someone else's fault. It's like, can anyone take responsibility for this problem? Like I'm left as the, the, and like, why is no one solving this problem? And I think that like, that is one of the things you could talk to Sandhill and the venture capitalists and they're like, we don't think this market exists. And like, why would anyone solve this problem? And then you go talk to the people that actually have this problem. And they they're like my house has been burning down for 30 years like why is no one calling the fire department and they're like yeah oh sorry they they think your

Starting point is 01:23:34 house doesn't exist they're like what and so we actually did have some of those earliest folks you're talking to god bless them we're very very grateful for because when you know another kind of question is like how do you start something that's gonna take you so long to, God bless them. We're very, very grateful for it. Because when, you know, another kind of question is like, how do you start something that's going to take you so long to build? I mean, it took us three and a half years to build, to build the oxide rack, the first oxide rack. And how do you, like, when you have customers for that, what does it mean to have a customer when you don't have a team and don't have funding and don't only have a deck? And like, what you need to, what you are connecting venture capitalists to are not like people that are going to buy your product today it's people that are going to buy that see the need for your product

Starting point is 01:24:13 and will buy it in you know three years what happened so but we have let those earliest folks know and we had a bunch of them who were very interested in what we were doing that like listen like venture capitalists don't think this market exists. And would you be willing to speak with them? And maybe we overshot the mark because customers are like, yes, send them to me. Oh, wow. Okay. Like, send me the venture capitalists. And one of the messages we definitely heard from more than one firm is like, yeah, talk to your customers they're very animated they're very upset yeah they're yeah they're upset like they're angry i'm like they are they are angry they are angry and like it's like it kind of felt like they were

Starting point is 01:24:57 angry at me it's like well they kind of are angry you not you not you per, but like the entire capital allocation class. They are disappointed that they feel like they've been denied modernity and that they are being left to suffer and they're being ignored by everybody. That Dell, Cisco, HPE, Juniper, now HPE, VMware, now Broadcom, they're not actually interested in solving my problem. In fact, they actually don't really understand what my problem is. They don't understand what Elastic Infrastructure is. They don't understand what cloud computing is. And they have kind of pitched themselves in opposition to cloud computing. So I'm fighting them every inch of the way,

Starting point is 01:25:49 and no one seems to be giving me what I actually want, which is a product where I can run that elastic infrastructure that I run in the public cloud. I want to run that in my own data center. Is that so complicated? And the answer is like, I mean, no, I mean, it's like, it's not, it's, it's complicated to implement, but it's not actually like the request is extremely reasonable. And, but people feel frustrated because it hasn't been, it hasn't been served and it hasn't been served for a bunch of reasons. It hasn't been

Starting point is 01:26:20 served because existing players, it's not in their interest to go do it. They don't see the whole problem. This divide between hardware and software is really deep. And so this part of the reason that this problem really needed new company formation and in order to have new company formation, now you got to go do the Sandhill Shuffle and you got to find a venture capitalist who gets it. And like a lot of them don't. I mean, some of them do. God bless them. The ones that do. And, you know, when you talk to them, the VCs that get it, which is great. You actually talk to them about like VCs that don't get it. And they're like, no, it's great. I'm glad that I'm the only one that gets it. Are you kidding me? Like, I love the fact that like, you know, in fact, send me entrepreneurs that are solving problems like this, because I love the fact that like, you know, in fact, send me entrepreneurs that are solving problems

Starting point is 01:27:07 like this, because I love the fact that I'm that, that others won't fund them. Less competition. Like, okay. Exactly. Less competition. I'm like, okay, but it would be like better for humanity if we funded more of these companies. You gotta acknowledge that. But that's in terms of the, so it is very helpful.

Starting point is 01:27:23 I mean, this is a long answer to your question. Like, did people think we couldn't do it? And yes, at every level. And they would say some pretty discouraging things, honestly. I mean, the, like, I had one venture capitalist who, as I was describing what we're doing, stops me. He's like, are you talking about starting a computer company? It's like, yeah, that's exactly what we're doing. He's like are you are you talking about starting a computer company it's like yeah that's exactly what we're doing he's like no no you like you absolutely you it is irresponsible to your family

Starting point is 01:27:53 for you to start a computer company you're like okay whoa okay dude let's like well like let me worry about the fam okay like i think maybe I think maybe we can dial that down. But he kind of went on this extensive, like, soliloquy. And I mean, this guy's like a bit of a, just bluntly, like, a dick. And actually, the person that made this intro was like, you two are either really going to get along or really not get along. And we definitely are in the not get along category but now i'm like personally offended that like you think i would get along with this guy like dude am i a dick like am i i'm a dick like like i and of course the person

Starting point is 01:28:40 that made the interest like well no i was like pretty sure you weren't going to get along i'm like okay this does not feel like you're this you're backpedaling here. But I did, yeah, definitely did not get along. Not just because of the discouraging words, but then he was like, look, you are never going to get this far. And I'm like, okay. And, you know, I will acknowledge like you understand technology. You don't understand venture capital. I understand venture capital. No one's even going to look at this. And I'm like, okay, well, actually, like, firms are looking at this. It's like, not prestigious firms. I'm like, actually, like, very prestigious firms. I'm like, and in particular,

Starting point is 01:29:18 like, I name dropped, and he name dropped a bunch of names that would never look at this. And I'm like, I'm talking to most of those people. I'm like, I've got a meeting with one of those that you just named. Like I've like, we've had three different meetings about this. He's super interested. And this guy was like, okay, you should know that my child goes to preschool with his child. I see him several times a week. And I'm like, are you accusing me of lying to you? Like, he's like, I will ask him about this. I'm like, you should ask him about it. Call him right now.

Starting point is 01:29:59 I mean, are you just like, are you? And it was funny. It's like, if somehow he like took that to break through to them, he's like, huh, he's looking at this. And I'm like, yeah, he's looking at this. He's really serious about it. He's like, oh. Okay, well, I'm a small fund. I would love to participate, but I am like,

Starting point is 01:30:19 are you joking? Are you a parody of yourself? Like, are you a parody of yourself like are you are you kidding me it's like 20 minutes ago what i was doing was irresponsible to my family 10 minutes ago i was a liar and you were gonna like fact check me at the preschool drop-off and now you want me to accommodate you and you're a small fund that you wanted i mean are you like are you it's like oh my god i'm actually working hbo silicon valley like i'm in it like i'm just like you people are an absolute parody of yourselves that's a great insult line i like it i'm to start using it. You're a parody of yourself. Oh, that's good. Oh yeah. Good luck with that one. I'm not sure. I'm not sure how well,

Starting point is 01:31:15 how receptive people are. You'd be like, you know what? You're right. I am a parody. I do think the, the one thing I do think it's interesting to ask people is have you watched Silicon Valley? And because an amazing number of people be like ah i can't watch it i'm like you can't watch it it's too real it's like it's too real whoa it's too real oh i'm so sorry is the satire too real for mr tough ventureist can't deal with satire. It's like, man, you guys like, you know, you need to spend a little bit more time in like the mom and pop restaurant, like working the line and like busting their hump.

Starting point is 01:31:54 If this is like, if it's too real to watch like satire, I'm gonna be like, oh my God. But I do think that like, then you do get people who, so I actually find kind of an immediate kinship with people in Silicon Valley who like and watch the show. Because to me, it says like, OK, you're able to actually like to take half a step away from this. And we are all parody in that. It's brilliant satire. Everybody in the industry from software engineers, I mean, like, come on, every software engineer, you know, you've got some Gilfoyle in you. Every software engineer's got some Gilfoyle in you. Every software engineer's got a little bit of the

Starting point is 01:32:35 Nash. Maybe you got a little, you may not like it, but certainly like Gilfoyle. I think like we all got some Gilfoyle um and it's important to be able to kind of like laugh at that and to treat that as the satire that it is and yeah it's like there are definitely moments that are extremely extremely real and like that's good that's not bad i mean it's actually i like it i you know watch it with my my teenagers and being able to like pause it be like okay so this actually happened like let me explain how you know, watch my teenagers and being able to like pause it and be like, okay, so this actually happened. Like let me explain how, you know, what they're describing is like a very real thing. So, okay. On that note, I want to switch gears a little bit. So one thing I wanted to just read

Starting point is 01:33:20 off your website on oxide.computer slash principles. There's a statement in our mission, kick butt, have fun, don't't cheat love our customers and change computing forever and it says right below that that this comes from the scott mcnally's quota for sun microsystems and you worked at sun now if we look at sun's history it's been renowned for a lot of innovation a bunch of products came out of sun they open source a bunch of those as well But towards the end there were certain financial struggles, too Now coming from Sun working at that company for I think 14 years What are some of the cultures you want to keep at Oxford are at least not the same but be inspired by and what are the things?

Starting point is 01:34:03 You want to intentionally change that's a great question and it's a great question because i think many people see the the first half of that question i think is clearer to people of like oh and there are things that you clearly want to emulate in sun which is true i think what people forget is like there are also things i don't want to emulate in sun and And so I actually, it's the back half of that question that people don't see as much, but it's also very important to me because Sun did a bunch of things right and did some things wrong too, and missed some opportunities that we're trying to rectify. The things that Sun did right, and actually it really is, Kick Butt had fun. So the motto of Sun was Kick Butt, have fun. So that was the, and when you went to, you know, I went to a new employee orientation in 1996,

Starting point is 01:35:00 and you've got, you know, I'm there with the 40 new hires on a Monday morning and we're watching, you know, the video from McNeely saying kick butt and have fun. Like this is what we do at this company. So it's like it was that kind of a mantra and which, you know, I think for its time was very iconoclastic. The idea that you would have fun at a company was out, like that's not the way it was in the eighties, right? Like that's not the, the idea that you would love your work and love your colleagues and have fun doing it was very, I mean, now we kind of take that for granted.

Starting point is 01:35:37 And the, and I, I believe you can prove me wrong, but I, I, I'm like almost certain that a i mean sun there are lots of things that were developed at sun that had influence way beyond sun that we don't always appreciate one of them is on the cringier side of things but started with the best of intentions is this idea of like corporate april fools pranks that started i am virtually certain at And Sun had, and that started in like this kind of, and it was very much this kind of impish idea of like a little bit of engineering mischief, you know? I mean, there's some mischievousness that is a hallmark of great engineering. And like, and there's a very clear line between like mischief and malfeasance right and you know and the april fool's prank which was and then it became kind of like increasingly institutionalized at sun over the years and became kind of frankly cringier at sun and then sadly escaped sun and now any kind of cringy april 1st you know some company doing some cringy announcement on april 1st is like that you

Starting point is 01:36:43 can trace that back to sun and by the way it's so pervasive that if you have something important to say, you actually can't say it on April 1st. There was actually a company that looks a lot like Oxide called Nebula that shut down on April 1st, 2015. And I remember being in the office of my CEO at the time and he and I were reading this and I'm like, this is either the darkest and most screwed up April Fool's joke in history. Like, no, no, we pranked you by telling you we were shutting down and the right day. It's like shut down March 31st or April 2nd. Like you just like April 1st is just off the menu for shutting down. Like, sorry, but you got to wait. Anyway, so that kind of impishness definitely comes out of Sun Became Cringier later. But that kind of idea of like having fun at work. But it was it was deeper than that. And like both the kick butt and the have fun had like, there was actually

Starting point is 01:37:46 something deeper that was understood by every Sun employee, but it was really left unspoken until the actual, the end of the company. So at the end it was, and what I kind of quote there, the kick butt, had fun, didn't cheat, loved our customers, changed computing forever. That was the actual full expression of what Sun was about. And that on the one hand, it's like you can be a little bit, it's a little bit unfortunate that that wasn't like completely articulated until the end. On the other hand, people knew that so deeply that it didn't really have to be articulated so i mean i mean to take those kind of apart a little bit like the not cheating was really important for sun and the

Starting point is 01:38:31 you look at you know sun for 30 years no one went to jail for sun there was no there were no i mean there were people who were you know there were you know people who, I don't want to say jerks, but like, that's actually like, that was pretty rare, honestly, at Sun. There were people who I really disagreed with, but we were both actually trying to do the right thing. And we disagreed on what the right thing looked like, but there was not really malfeasance. There was not, you know, and Sun did not cut corners. There were not financial scandals. There was not an option backdating scandal. There was not like the company was really pretty squeaky clean in a lot of ways. I mean, squeaky clean relative to its own age, especially relative to the age that came

Starting point is 01:39:15 after it. So that was a very big deal. Loving our customers was really, really, really important. And it took me years to really appreciate this. But the reason that Sun did not perish in the dot-com bust is because we did love our customers and our customers loved us back for it. And customers viewed us as an entity that they could trust. And that trust is really, really, really important. Because I think us, I just, I love the fact that like when Sun comes in here, they don't try to pretend this isn't their problem. They can own, they'll own the problem. particular counter example of this where sun had like really uh i mean had incorrectly chased a very a brutal quality problem called the e-cash parodyer but the fact that like the customer was telling me about this tells you that like how exceptional it was that and it was indeed a

Starting point is 01:40:20 horrifying story and the the sun did not know what was causing this problem. Just to finish the story, son did not know what was causing this problem and was doing things that companies shouldn't do but do do when they've got a problem they know is important that they don't understand. What you should do is say, I don't fully understand this problem. Here's what I'm doing to understand it. I don't fully understand it. People think that customers I'm doing to understand it. I don't fully understand it. People think that customers don't want to hear that. And I mean, admittedly, like a customer would rather have the problem resolved. Like it's a customer's not like, you know what I would love to hear.

Starting point is 01:40:56 I would love to hear you have no idea what's going on. Like, that's what I would love to hear. It's like, nobody like wants to hear that. But if you have no idea what's going on, customer would really appreciate the transparency of like, okay, I get, you don't idea what's going on, customer would really appreciate the transparency of like, okay, I, I get, you don't know what's going on, but you are trying to understand what's going on and here's what you're investigating. Companies don't do that. Generally, they generally will confidently tell you it's something that it's not.

Starting point is 01:41:18 So Sun didn't engage in this as much as a lot of others did, but like, this is what was so frustrating about being a Dell customer. Dell does this all the time. Dell spends all their time trying to find something that they can blame on the environment. And, you know, we, but Sun was not immune from this. And as one customer was relaying to me, they were seeing this, this really serious CPU defect. And Sun was kind of, didn't know what was going on and was trying to blame everything. And they're like, well, we think it's like, it might be because of vibration in the data center and your proximity to a tube station. This is, or to the tube, this is in London. And like, it was a quarter

Starting point is 01:41:55 of a mile from the tube. There is not, I don't know how familiar you are in London, but like, you are always a quarter of a mile from the tube in London. Like that whole city is vibrating by that, by that standard. Right. So it's like, go fish. And they're like, well, we think it's like actually the there's dust in the data center. And so there was like, you know, there had been some dust issues. Was that directly connected to the problem? Sun had no evidence for that, but wasn't saying that. Customer like spent spent a million pounds

Starting point is 01:42:26 doing a totally new HVAC system to make the data center very clean. But the problem's persistent because the problem wasn't due to the dust, right? In an effort to save the customer, they were invited up to the factory in Glasgow, Scotland, that actually made where this machine came from. Customer goes in and sees this environment where this big server comes in and is de-boxed and is then burned in in the same room. De-boxing creates a ton of dust. And you cannot de-box in the same room that you're going to run infrastructure if the objective is to minimize dust. And this is like a bunch of sun top brass there. Customer, as they described, ran their finger along one of the surfaces, looked at the finger.

Starting point is 01:43:23 Of course, it's like black with dust. Customer pointed the finger to the executives and said, I paid a million quid for what exactly? And I was like, wow. I'm like, I bet. What was the reaction? He reaction he's like oh people were ashamed like people were there what you could hear a pin drop in there and i mean that is kind of like on the one hand the fact that son got this kind of wrong for so long that's a real problem but that kind of moment of truth we were talking earlier about these like difficult moments, like that's a difficult moment. And Sun realized at that moment, like we're down the wrong path here. And ultimately that problem, by the way, was like six different problems, including this would manifest itself as this, and there were a bunch of

Starting point is 01:44:26 things that were wrong with this, but architecturally it would always manifest itself as an e-cash parity error. But there were many, many, many, many, many root causes. One of which was our SRAM manufacturer had an impurity in their process and there was radioactive boron in the SRAM cells. So like that's bad. That's a, it's an alpha emitter. You can block alpha particles really easily, but not if they're in the SRAM cell. It can't be. It's like in the room. That was just one of the problems. There were many, there were signal integrity problems. There were architecture problems. There were software problems. There was a whole, like, basically you could view everything that was wrong with Sun's engineering at the time through the lens of that eCache parity error. something that I, you know, I've been very first, tried to be very first principles in our engineering approach. Sun was first principles at times, but was not first principles at other times. This is an example where they were very much over their skates. So those are like elements

Starting point is 01:45:34 that I very much wanted to emulate. Sun empowered innovators, which I loved. There was a moment when McNeely actually came to speak to the kernel Group is when I was like, you know, I'd only been at the company for a year or two, kind of starstruck the, the CEO of this 30,000 employee company that was there. And someone was asking him about like the coffee machine or something being like, you know, the coffee machine is terrible. It needs to be replaced. Something that felt like, okay, I don't know. Like, what do you, you know, yeah. I mean, it was like a slightly bigger complaint than that, but it was basically like, okay, I don't know. Like, what do you, you know, I guess you've got to endure all. Yeah. I mean, it was like a slightly bigger complaint than that, but it was basically like

Starting point is 01:46:09 I, I, I complaint about something in the organization and McNeely's response was like, look, I don't know what you, who you think I am or what you think I can do, but I'm not a magician. I don't have a wand. It sounds like you've identified a real problem there. Fix it. Get it fixed. And I remember thinking, like, being a bit struck at the time about someone who would so willingly kind of give up their authority, delegate their authority, to say basically, because what he was implicitly saying there was, I trust your judgment. I trust your judgment that the problem that you've identified is real, that the solution you've identified is valid, and I trust you to get it fixed. Get it fixed. And my expectation is that you, I don't want you suffering inside. I don't want you waiting for me to fix that. Like the way to get that fixed around here is to fix it. And I don't think I realized how

Starting point is 01:47:11 idiosyncratic that was until after Sun. And I don't think I realized how powerful that is. And part, like the great things that Sun was able to achieve all stems from that trust that Sun implicitly had in its innovators. And not to say that there wasn't friction and churn and a bunch of other things, but not to overly romanticize it, but ultimately Sun trusted its folks. It found the best folks and trusted them.

Starting point is 01:47:48 And that's a reason why people will say, I did some of, you know, many, many, many people will tell you that they did some of the best work in their career at Sun. And it is because of that. That is very much what I want to recreate at Oxide. And we have recreated at Oxide. That sense of we're going to spend a lot of time and energy finding the right people and we're going to trust them. I trust your judgment. You're here because I trust you. And my expectation is that if something is broken, you'll fix it. And that has generated something that is powerful beyond us as individuals.

Starting point is 01:48:28 So that's a lot of what Sun did right that I wanted to emulate. But there are other things that Sun struggled with. And Sun loved its customers, but didn't always generate the products that really reflected that love and in particular like sun would i think too often would generate kind of a would drop kind of a bag of bolts off at a customer be like you know good luck with it all well yeah i think that like not always and you know the the vision would be great but but sometimes Sun would like lose patience with the execution. Not always, but there were a couple of things that Sun was way ahead on, but took its eye off the ball. Sun took its eye off the ball, ultimately did the right thing with respect to open source, but should have open sourced the operating system in like 1999, not 2005. Obviously had the right thing with respect to Java, but was too late to open

Starting point is 01:49:34 that up, I think. And then invented some wild technologies. There's something called Sunray, which was this technology where you would have a workstation that basically had a smart card reader and your graphical session was on the actual server. So when you plug the smart card in, that local client served only as a frame buffer. And you can view this as kind of like really revisiting the mainframe and it kind of it's distributed terminals but in a modern at the time this is like 2002 kind of fashion and man the demo of that thing was jaw-dropping because you'd be playing a video whatever you'd have all this stuff up and you would pull your smart card and then you'd go to the next station, put your smart card in, all comes back up right as you left it. And we still don't have anything that's exactly

Starting point is 01:50:31 like that. And, you know, there are fields that should have been revolutionized by this, right? We should have like, this should be every hospital today. You know, every, you know, every nurse or doc should have that smart card around their neck. They should plug it in. Like they should be able to, you know, it's like they shouldn't have to be on a tablet or their phone. Like they should be. They should have a laptop around in a way. Basically just need a screen. Exactly.

Starting point is 01:50:54 Exactly. But this is where like Sun, great idea that Sun lost patience with the execution on and lost patience on the execution at exactly the wrong moment. And in particular, this is like every airline kiosk should be this, but this is in order to win the airline kiosks. And this is again, in the early 2000s, like 2005, maybe they really needed to have USB printing, which is a pain in the ass. It's awful. And who wants to go deal with this thing if you don't have to? It's like, well, we actually needed to do that. If they had done that, because American Airlines was going to go

Starting point is 01:51:39 deploy new kiosks. Everyone was going to follow American. And American was looking at Sunrays versus a bunch of other things and ultimately didn't deploy Sunray because they couldn't get the USB, like USB printing wasn't happening. Sunray was just kind of like disinterested in USB printing. And to me, like the larger issue there was we, I mean, had the love for the customers. And the reason I know this story is because of the SEs and the solutions engineers that were in Chicago dealing with American at the time. But we didn't always connect it, connect sales with solutions engineering, with support, with engineering, with marketing, and, you know, kind of holistically point at these problems. And as a result, you end up with like, it can feel to a customer like this is interesting and innovative, but chaotic. And it's like, oh oh you've done all this amazing thing like you can't get usb printing to work it's like that's the only thing that i need and then like and as i go so go all of the all the airlines behind me and it's like some just

Starting point is 01:52:59 couldn't summon the attention span to go do and and to a certain degree, it was kind of like, just, it was too chaotic. And so, you know, kind of the question that we have in front of us is like, I don't want to have that problem. I want to be sure that at Oxide, when you've got a customer that has a feature request, that it's like, oh, I love this, but like, here's this thing that I need. I want to be sure that we as a company and as an organization have the disposition to get that problem solved. And I think that if you can, if you can do that, if you can take not just this extraordinary foundation,

Starting point is 01:53:39 but get that level of polish and finish on it, you can do something extraordinary. Because I mean, this is kind of the level of polish and finish that like an Apple might do, but you're doing it based on the feedback from someone who's actually deploying it. You deliver something that's much larger than yourself because from a user's perspective,

Starting point is 01:54:04 it's like, oh, this is awesome. Like I was missing this piece and they added the piece that was like, they listened to me. They just loved that. Yeah. They listened to me and I had an impact on where this company went. And it's like, man, I love that. And I want us all to love that feeling. And I think as a company, we've hired carefully to get folks who want to be engaged with how the technology is used and get real wind in the sails from that kind of customer experience. And, you know, we feel really, really lucky that our, you know, because, you know, when you've got these, there was years ago, there was a motto contest in Silicon Valley, which I love. And actually, the winner was Quality is Job 2, which is a playoff of Quality is Job 1 from an old Ford motto back in the 80s. And Quality is Job 2, of course, like, ha-ha. There's definitely some truth to that, unfortunately, in Silicon Valley.

Starting point is 01:55:20 But the actual second place finisher was truer to me. The second place for the model contest was God bless the early adopters. And man, that is something I feel in my marrow because we rely on the early adopters. You know what I mean? Like the technologists that will suffer, that will be in pain. I mean, I know you had Kelsey on here a couple of times talking about those early days of Kubernetes, right? When Kubernetes, I mean, Kubernetes was a God awful mess. And I mean, I actually dispatched one of our engineers and this is like in 2015.

Starting point is 01:56:04 I'm like, I think, you know, I mentioned Kubernetes, like go experiment with it. Like, let's see what we find. And he came back. He's like, I, no one is using this and I feel I can prove it. Like the documentation is so wrong in so many different regards and so contradictory that I have to be the first person that's ever tried to use this. Like, this is like, this is, this is, to say this is unusable is to understate it. Like this is, so, you know, I think that, but we need those early adopters. We need those folks that are willing to suffer with a technology when it is really immature. So as a company, you're always looking like we are looking for those early adopters and we want

Starting point is 01:56:43 to deliver them a delightful experience. And I think we've been really focused on that at Oxide and it's been great to get, you know, we started shipping just last year, but we have, you know, getting more and more and more of those early adopters and, you know, watching them realize like, oh, this is like, this is potentially transformational for me. Like I'm going to be able to do some really new things. Or just watching their own love for what we have done and appreciation for just the beauty of it, which is something we've really focused on.

Starting point is 01:57:15 I mean, beauty and the data center are not something that you've really been in the same sentence, but that's something that we've really focused on and to watch our customers really resonate with that. And I feel like, God, normally I would be totally circumspect on saying this, but I feel like I can kind of say it out loud now. So Tobin, who's the founder and CEO of Shopify, had this tweet last week talking about- I saw that. I saw that. Oh man, did you see that? Yeah, that was amazing.

Starting point is 01:57:43 Oh, that was. That must have felt good. Oh, that felt real good. That felt real, real good. Because, I mean, it obviously felt great just because, like, obviously we love Shopify. We love what they're doing. Love their technologists. It's been, it was great to get it in, to get the rack.

Starting point is 01:58:04 I mean, everything, all of the specifics about it are great, but there's something much, much larger about that, about a customer that's excited to be a customer and that they, a technologist that's excited to use your technology to go solve a different problem. A problem that, you know, going back to, you know, that I was, VCs thought I was being irresponsible to my family by solving. It's like, those are kind of like the bookends, you know what I mean? Those are the bookends of like, you know, what you're doing is like irresponsible to your own family. And you've delivered something, you've delivered this beautiful machine and that you know hey did and i love the way he phrased it like hey you know if you want to come work on this thing

Starting point is 01:58:51 like come work for shop oh yeah yeah and this is beautiful oh man yeah so like uh did you see second tweet it was like just watching a server do server things it's kind of just beautiful and he had this console dashboard of sorts on the terminal just showing like how much CPU usage you have, et cetera. And like just seeing the CEO of a company actually appreciate pieces of technology and not just like, Hey, here's a fancy. Yeah, but no, no, no. Here's the actual machine running the software and seeing the appreciation for that was just

Starting point is 01:59:18 beautiful. Oh, it was beautiful. It was beautiful. It was beautiful. And like, you know, just obviously really deeply grateful about kind of all aspects of that. Deeply grateful for, you know, a company that's buying a product. Well, it's still early, you know, I mean, again, this thing has been in the market for, but we've got plenty of customers like, we've been in the market for a couple of years, like then come back, you know. But it's like, okay, well, I hope we can, you know, we're around. But, you know, to sort of have a customer taking their own kind of a risk, but then to be willing to talk about that publicly, I mean, it's just like such a, it's so deeply gratifying and deeply gratifying to everybody at Oxfam. Like that is something that, you know, you, we've all been working so hard on this for so long and to be able to deliver that and to have someone talking about it in that way is, it is meaningful at a level that's really

Starting point is 02:00:15 hard to express. It is very, very, very deeply meaningful. Yeah, totally. Well, we have way more questions to go through than we'll have the time for today. And this has been an awesome conversation. We hope there's a round two. We hope you come back. We would gladly ask you a bunch more questions. We purposely saved the shouting in the data center for a second. I noticed that.

Starting point is 02:00:39 I think we can't do a proper interview without mentioning that. Yeah. But one thing which I will ask before we let you go is you've done a lot of hard things. In fact, I would say out of school, you wanted to work on operating systems and kernel development. Not every engineer's dream, I would say. Like, hey, I'm going to work on the operating system or the kernel. That's not what most people say. It's actually one of the hardest things to do.

Starting point is 02:01:04 Spending time at Sun and then joined. And now Oxide, that's also a very hard thing to do where you're actually building a new server and integrating software hardware together. So for many people who are out there who would be interested in doing hard things but may not have that slight push that they might need, do you have some piece of

Starting point is 02:01:25 advice? Yeah. I mean, I think that hard things are hard. I mean, you don't want to, and I, you know, we have found this at Oxide that we have had people that are interested in the idea of working at Oxide, but when you actually get to the specifics of actually like building this thing and getting it working, it's like horrifying. You know what I mean? In terms of like the level of detail and like how hard it actually is. And, you know, one of the things that we did as, or do as this kind of a part of our own hiring process, if we have someone who we think we're interested in, we will open them, allow them to get into what we call

Starting point is 02:02:08 our requests for discussion, RFDs. These are our design documents. Oh, I've seen that on your website, by the way. That's an amazing concept. I mean, you see these documents internally within a company, like RFCs, requests for comments. I actually like RFDs, requests for discussions. Honestly, I like that take a lot.

Starting point is 02:02:22 And just seeing all of these discussions out on your website, that's amazing. Yeah. And we opened a bunch of them up so you can go see a bunch of them that are public. We've got a bunch more that we need to open up. But so when we have candidates, we open them, all of them. But if you just look at the RFDs that we've got that are public, people will get a flavor for them. And you will read these things and you'll be like, God damn there's like a lot of detail here like this is complicated i mean there's a level you're just like holy okay they like wow they've like thought about a bunch of stuff but this is like and if you dig deep enough

Starting point is 02:02:57 on any one of them you'll get to the limit of our own thinking and you'll see like okay this is like they thought about some stuff they're there's still some other questions here. So the, I think that the part of the reason we do that is because to show people like, yeah, this is hard. It's very, very detail oriented. It's hard work. You know, I, I think that we do our, I mean, you know, I would like to believe that this is changing, but this kind of like Instagram culture of showing people the result without the hard work and then pretending that it wasn't hard. Yeah. You know, pretending that it was easy. I think we do all of our ourselves a huge disservice when we talk about how easy it was instead of talking about how hard it was. And, you know, there's Ben Horowitz, the venture capitalist that wrote a book, The Hard Thing About Hard Things.

Starting point is 02:03:56 And, you know, I mean, there are aspects of the book that are a little ridiculous, but there's things I definitely like. And in particular, like, you know, this is a company, he was at a company that was going public to make payroll. And while he's on the roadshow to go public to make payroll, his wife has a really serious health problem. And he gets a call from his father-in-law and his wife is basically in the hospital. And it's like, yeah, that's hard. That's hard. That is what the kids say. I will say this. I do have a great deal of faith in Gen Z, in part because my own kids are Gen Z. They're the kids of Gen Xers. And I love how part of the lexicon of Gen Z is high praise is like, that's tough. That's tough. And it's like, I like that. That's good. That's good. There's some Gen Z slang. I also love

Starting point is 02:04:43 Riz, by the way, as a verb. It's phenomenal. I've never heard that. What does that mean? Okay. Yeah. Yeah, this happens. So yeah, you are obviously millennials. This is where millennials realize, yeah, you're sorry. You're actually age-danger. No, so Riz and I should be saying, this is like, I've got the unfair advantage. I got teenagers. And I, the way I came across this, this delightful word is the, I lead the kids scout troop. And one of my fellow adult leaders was talking to two other, to two moms in the troop. And one of the teenagers turns to the other, but very much within my earshot,

Starting point is 02:05:27 and this is talking about his father over there, talking to these two women, adult leaders, moms. He's like, look at that guy, the Rizzler. I'm like, the Rizzler? What is the, what? And so, and I was texting him later, I'm like, Paul, I don't know if this is praise or not, but you should know that the kids are calling you the Rizzler. The Rizz is derived from charisma. So to Rizz is to be charismatic. So if I'm Rizzing you up, I'm flirting with you. And the great thing about this is like the power of English in terms of the different parts of speech you can use. It's like it becomes a noun, like you're the Rizzler. You know, you're the, so Rizzler.

Starting point is 02:06:16 Anyway, I love tough as a Gen Z-ism because it's like that's tough. That's tough. And as praise. I think that's good praise to have. I think that we need, like the hard things are hard and expect them to be hard. And when you, like, you know, you're doing the right kind of thing because the problem is punching you in the face and you're still moving forward. You know, one of my favorite little, Piet Hein was a mathematician and poet who would have these little, that I love, right? I love it when you get people who are both technologists and artists at some level, because

Starting point is 02:06:57 like the art has a different resonance with people who are technologists. And Hein is a mathematician, was a mathematician, so Hein has that. And so he's got these little poems that I think are really important to any technologist solving a hard problem. And one of them that I love is problems worthy of attack prove their worth by fighting back.

Starting point is 02:07:19 Oh, that's beautiful. Yes. And another one he's got is the, and I won't quote it, can't quite quote it, but things take time, TTT. And just to remind yourself that things take time. Problems worthy of attack prove their worth by fighting back. So when you've got a problem that is punching you in the face and pushing you down, that's a problem that's worthy of attack. And it's hard and it's easy to say like, oh, like you should like get back up. But it's like, it's perfectly reasonable. Like, yeah, yeah, I know I have been getting back up for the last 30 times. It's but like this 31st punch, it's like I actually want to lie down for a while. And like that's also understandable, right?

Starting point is 02:08:09 And it's like you also like you need to get the wins under the belt. But when you get that win, when you grind, grind, grind, grind, grind, grind, grind, and then like you nail the bug or you get the thing and you solve the hard problem. My hard problem to me, a hard problem is one that you didn't think you could solve at some level.

Starting point is 02:08:33 And clearly different problems will be hard for different people. Right. So this is another thing that's important. It's like hard is relative, not absolute. And like, you know, and there is value in solving a hard problem, even if it's easy for other people. Even if it's easy for other people, cook yourself a meal. It's like, no, I don't know how to cook. It's like, cook yourself a meal, even if it's a mess and it's easy for someone else who's got a lot more experience in the kitchen, there's something gratifying there about like, wow, I didn't think I could actually cook myself a meal. And I like, this is actually pretty good. Like, why am I doing DoorDash again?

Starting point is 02:09:16 Like, this is actually good. You know, and so like when you can drive yourself and in the technical domain, when you can drive yourself to solve a problem, like I actually legit did not think I could solve that problem. And I did it. It's important to get that win because then you do get the confidence. Like, okay, I know this is possible. Actually, I know I can do this. And then you start getting more and more of those wins. I personally, you know, we talked about kind of the A-team as an influence for me. I love that when you are working on a team solving a hard problem. The thing that is extraordinary about that is that

Starting point is 02:10:01 when you might be having your own moment of low confidence of like, God, I don't know if we can solve this. And in comes your colleague and has the insight, solves it. And you get that moment of like, oh man, thank God we are on the same team. And God, I hope I can return the favor one day. And then you return the favor one day. You know, and it's like, oh man, that is lightning in a bottle. Like that is magic. And when you look at the times in the history of the industry where people have done extraordinary things, it's because they've solved hard problems as a team. So, I mean, that to me is what I would kind of the other recommendation I would have for people is like, don't solve hard problems by yourself. Go find someone else you can solve it with.

Starting point is 02:10:57 And that feeling you'll get from solving it together, that's unparalleled. And, you know, you start adding that up and you will get an increasing amount of confidence and you will pick on bigger and bigger stuff. And by the way, by picking on bigger and bigger stuff, like you will have the, I mean, the most difficult moment for me in the history of Oxide, this is kind of crazy, the most difficult, and it took me kind of like years to really talk about it because I felt like, bluntly, like I felt a bit embarrassed about it when we got our first term sheet. So we got our first term sheet and I'd been working so hard for so long on getting that. And we were raising a lot of money on nothing. We didn't even, I mean, out of debt. Three people in a debt.

Starting point is 02:11:48 We get that term sheet, and I was not braced for it at all. I'd been thinking so much about get the term sheet, get the term sheet, get the term sheet, that when we actually got it, I wasn't ready for what was emotionally on the other side of it, which is like, oh, shit, I got to go do all this. And I don't know how to do any of it. We talk about like how great it is to have a team. We didn't have a team. We had three people.

Starting point is 02:12:15 I didn't know how to build any of this stuff. You know, like I knew how to build aspects of it, but like, oh my God, I am. And I have just like, there was a level of terror that was pretty deep. And, you know, I have confided more or less absolutely everything in Steve, in my co-founder. But like, I felt I had a responsibility to him to not confide this sense of terror. That may have been a mistake in hindsight because I think it would have been fine. But I just couldn't quite bring myself to be like, by the way, I'm terrified.

Starting point is 02:12:55 Because he'd be like, what? It's like, okay, well, get over it. Because we just raised a bunch of money together with the idea that you could go build this or build a team to go build it. So, like, you've got to go build it, pal. Not the right moment to be terrified. Right, not the right moment to be terrified. It's just like, it's like we just took off. And, like, the pilot has told me, like, hey, do you, by the way, I've got a pathological fear of flying.

Starting point is 02:13:22 It's like, what do you mean you have a pathological fear of flying? Did you want to talk about that like when we were at the gate like why is this not like we're now like well we're we're flying now so you know and but fortunately and this is where you know and you know you'd had i and i can't remember which one of you mentioned this but talking about the podcast about how your mom was listening to the podcast. I thought that was great. Okay. I love that. I love that because, you know, I think that, you know, we rely on our parents for different things at different times. And, you know, you need certain things out of your parents when you're, you know, a kid, a teenager. But when you go into adulthood and have your own kids,

Starting point is 02:14:02 it's like you need different things out of your parents. And, you know, I'm super grateful to my mom because I was talking to her like during like I was like almost catatonic, which was weird for me. I was not not something I'd ever really experienced of like I am. I'm sick to my stomach. I'm so gripped by fear about this. And my mom was like, you've done things like this before. You've done hard things before. Like, you'll do it.

Starting point is 02:14:34 And I'm like, Mom, you don't know what you're talking about, which is true. I mean, she definitely did not know what she was talking about. Like, my mom doesn't know. I'm like, Mom, you actually don't know. Like, I appreciate the sentiment, but, like, seriously, I'm out of my depth. And, but I actually had the kind of the wisdom to not argue with her and be like, you know what, she actually is right. And I can't do it. And then it's like, importantly, you get those first wins. You get, you know, in those early days, just getting employees to come join us. And,

Starting point is 02:15:08 you know, our first engineer, Robert Moustaki, getting Robert, and Robert and I talked about Oxide a lot for like, you know, a year prior. I've worked with Robert, I'd hired him out of college. I've worked with him for, I mean, like there's like every reason to believe that Robert was going to come join us at Oxide. But it's also like plenty of reasons why you wouldn't necessarily. And so, you know, just Robert coming aboard, like, oh my God, you know, they're now, you know, and, and then getting those, like those, those next hires and then beginning to like, okay, figure out how we actually, you know, I, and I remember actually one of the earliest RFDs I wrote was on the RFD process because I'm like, I feel I can write about

Starting point is 02:15:52 this. I, there's so much about the specifics of the problem that I still don't understand, but I can actually make the concrete progress of talking about how we're going to talk about it. You know what I mean? And like just getting that as a win and feeling like, okay, I did something today. I, I, I, and that I think is what's really hard is when you're working on a hard problem, you have these days where you feel like, I don't think I did anything today because I just like came up with this, you know, this little spoon at the granite kind of face and like kind of beat the spoon at it for a while i can't even like there's not even a mark here the spoon is dang dang up i gotta go like get another spoon or whatever i'm just like what am i doing and like you you need to get find ways to get the little

Starting point is 02:16:35 wins and then when the little wins you get the the wins are a little bit bigger and then the wins are a little bit bigger and i think that you know you always have to celebrate the wins you you want to be grounded about it. And that's not the time to go on Instagram and be like, oh yeah, by the way, like, but it's like, no, you can know, like, this is gratifying because it's hard. This is gratifying because we got this win and it was hard. And those moments of gratification build on one another. And, you know, when you're at the point where, you know, your customer, the CEO of a customer is tweeting about your product and how they love it, like that is extraordinarily gratifying, but it has come

Starting point is 02:17:18 only after this extremely long progression. And I don't want to sugarcoat like what that process looks like because it's long and it's arduous, but you can do it. One can do it. It's one foot in front of the other. And, you know, that is the, and just making that the, the, the kind of the energy involved in attacking the problem is energy well spent, even if you're not always seeing it in the problem itself. Problems worthy of attack prove the worth, I find. That's beautiful. As you were describing this, I have like 15 more questions in my head, but I'm going to restrict myself and save them for another time. This has been an awesome conversation, Brian. Thank you so much for taking the time and joining us today. We'll link to everything we discussed today in the show notes to the Ocklides website.

Starting point is 02:18:08 You have an amazing blog and a bunch of talks as well. We highly encourage people to check it out. We'll also link the photo that Toby actually shared on Twitter. That's beautiful. Definitely worthy looking at. And again, thank you so much. Thank you for having me. I really love your

Starting point is 02:18:25 approach with the podcast. Love that your moms are listening to it. And I love the, you know, one of the things you'd said in an earlier episode is just doing your own homework on, and I love that about the questions that you ask. That's amazing. How many folks don't do that? But really, I certainly did my own homework on you coming into this. But really thoughtful questions that's shown in the results here. You guys, you're having terrific conversations.

Starting point is 02:18:52 A lot of fun to listen to just as a listener. Oh, thank you. That means a lot. That means a lot. All right. Well, I look forward to being back at some point for all the questions. Oh, yes, for sure.

Starting point is 02:19:04 We will take you up on that. We promise. All right. Well, I look forward to be back at some point. Oh, yes, for sure. We will take you up on that. We promise. Thank you so much. Hey, thank you so much for listening to the show. You can subscribe wherever you get your podcasts and learn more about us at SoftwareMisadventures.com. You can also write to us at hello at at softwaremisadventures.com. You can also write to us at hello at softwaremisadventures.com. We would love to hear from you. Until next time, take care.

Software Misadventures - Build the scary stuff | Bryan Cantrill (Oxide)

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.