Current Time 0:00
/
Duration Time -:-
Progress: NaN%

How did you get to where you are today? What incidents and experiences shaped your career path? What inspired you to work on this startup idea?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
Um Well, that's Ah, a lot of questions all rolled into one s O that I think that that I had, um you know, kind of Ah, a fortunate timing in my life and that I was, um, growing up right as thean er net was becoming something that was that was starting to be used universally. Um, and at the same time, I had kind of an interest in computers. And so in the nineties, I I was Ah, I was a I was learning how to program and discovering websites. And, um, when I was a teenager, I realized that I could that I could make web websites for companies and charge money, and it was a pretty easy way toe thio make make some cash. I did it for very cheap. You know, I was I was, like, 13 or 14 years old. I didn't realize that I should charge more than 200 or $300 for a website, but it seemed like a lot of money you want to me. So I was really popular with with different businesses. Toe get them online. And from there I got interested in kind of how the Internet and the World Wide Web worked as a whole. Because, you know, once you build a website or Web application, you have to run it somewhere and on. I started Thio Thio kind of look into servers and databases and all of the things that happened behind the scenes. So that was that was really how I ended up kind of getting into a tech career. I actually was was planning to study music and, uh, and have a career is in music somehow. But, you know, I just kind of, uh started down the tech path and then continue to find more and more things that I was interested in. And that's how I how I got started, um, for the cloud, specifically, what inspired me and my co founder Todd at the time is you know, Todd was a designer and Web developers well, and we wanted to be ableto build websites really quickly to get them online really quickly. And and this is in, you know, the 2000 and three and 2004 timeframe. And at that point, you basically had two ways Thio put a website online. The first way was, um, by acquiring servers and installing everything that you needed to have on the server and then managing that server on a daily basis going forward, which was, ah, lot of work and took a lot of Ah lot of expertise that that, you know, not everyone had or the other way was you could go to what was called a virtual hosting company and, uh, basically virtual hosting companies. Um, they ran the servers, but they would put on each server as many sites as they could fit. And so you might be on a server with 10,000 other sites and eso there was. There were different problems with that. If someone else on the server was running a website and it had bad code, it could crash the server or slow your site down. If the server failed and the the virtual hosting company wasn't diligent. Then, um, you know, your site could go down, you could lose all your data, so those were really the two options. But we felt like there that we had learned these techniques at Rackspace for building load balanced websites with backups and redundancy, and and, uh, and is the way that major corporations designed their websites and we should find a way to make that available toe everybody. So it was kind of, you know, the inspiration came from the experiences that we were having. But you know, all their desire toe kind of improve the state of affairs for for everybody who was who was still trying to build websites, build web applications, get online and kind of make the digital economy. Really, we thought that it was important that the that the infrastructure not be so difficult.

What was the elevator pitch of your startup? What problem did it solve? How were your customers solving their pain point before your startup?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
um, the elevator pitch. What was Ah, it was, um, code it loaded and go. And we actually had, um we had several kinds of phrases or taglines around that, um, one of the the campaigns that we did was we said no more servers, um, or the system beats the server because it was it was a hosting system rather than a server. So a lot of tie ins to to basically running technology without running servers, you know, and and even today, you know, here we are in 2020 and one of the one of the big movements is serverless. Right. Um, you here functions as a service or serverless computing like aws lambda. And it's the same concept where people want to write their code. And they wanted to run somewhere reliably and without without having to be responsible for all of the hardware and the networking and storage and the system administration s. So that was really what the value proposition was around was giving them this capability tohave reliable, scalable websites, but without having to take on the responsibility of managing infrastructure. And like I said, you know, before that the two ways that people did it were by managing all that infrastructure themselves or by going thio what was just going to virtual hosting, which was, um, really kind of ah, low cost and also, you know, low, usually not very scalable and somewhat unreliable way of hosting.

Can you walk us through your first few weeks when you started working on this project? How did things change over the next few months?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
Yeah. So they were kind of two. I would say maybe two phases to it. Um, that that we're both the beginning and away. The first phase was Todd and I were doing this is kind of ah, side a side hustle. While we had had day jobs, we would we would go over toe his apartment and work on it. And, um, you know, we What we did is we way we were able to get a couple of servers. Um, I did all of the system configuration and, you know, setting up. I wrote a bunch of kowtow automate the set up of the sites for customers, databases, automate backups, like all those kinds of things. And And while I was doing that, Todd was busy creating, you know, our website, our marketing material at the time we called it site lingo. That was the original name and eso in those first few weeks. What What happened is we actually had through through Ah, we lived in San Antonio, Texas, and through a friend, we had the opportunity to get a very large national brand running on our on our system. But, you know, we needed Thio. We need to look professional. Eso we We were scrambling for working all night for a couple of weeks. Toe put together a web presence that looked really and we printed out data sheets and, you know, put on button down shirts and went and had a meeting with them and pitched them and showed them a demo and all this kind of stuff. And the customer ended up signing up and and then, you know, that was kind of how we got started. And we went along slowly like that for a while until we ended up basically selling it back to Rackspace. And that was when we relaunched it as moso. And, uh and, you know, at that point, what we what we were doing was really trying to prove out that there was that there was more than just kind of a a local business here, you know, we had we it had grown to a point where Rod and I were covering the bills with the customers that we had. But, you know, I think the Rackspace executives wanted to know if there was if there was really market potential for it. And so, um, you know One of the things that's that's great about the Internet is it's pretty cheap to produce things and throw them up and see what kind of response you get. And so we created a a a new website for this this product moso that we're going to be launching and, uh, and described what it would be and what it would do kind of this next generation of what we had already built. And we put up a sign up form and, uh, and invited people Teoh a to an invite Onley Private Beta. And, um, you know, the goal was to try to get, um, you know, like 100 to 200 people to sign up and indicate that there was some market interest out there for this kind of product. And if they came through, then we would spend the next few weeks building it out and then, you know, like inviting them to the beta and and go from there. Eso we we built out that website, we put it up and we posted it online and a couple of forums and we ended up getting I think something like 5000 people signed up so it was, you know, 20 times what what we had kind of hoped for. And at that point, I think, um, we were all, like, Wow, there seems like there really is a market for this. And, uh and so, um, you know, we we got started building out that that the prototype of the next generation of all of those systems we invited in, you know, I think about ah, 1000 initial users to be beta customers and then went from there. So those were the two. Those were both kind of like starting phases and in both of them. What? What I think we were focused on was kind of bootstrapping to an initial set of customers and proving that there was market interest and somebody who was gonna pay us for this concept we had, um, rather than thinking about, you know, how do we go spend, uh, years building out a product and secret and then, you know, like, launch it and see if it if there's interest, I would say it was very much that kind of lean startup bootstrap approach for for for getting it off the ground

What were the challenges in building the initial team and how did you overcome them? How did the team's composition, dynamics, time, and resource commitment evolve?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
I think that the challenges that we faced in the beginning the initial challenges were really around finding people with technology experience that that was relevant for what we were doing. Um, you know, this was in 45 6006 on DWI grew the team from me and Todd Thio hundreds of employees over those those years. But what we what we needed were were technical people with with with experience and distributed systems. Um, you know, now this would probably be what people would call on SRE or a site reliability engineer. And and it's the It's a thing that has now been pretty well defined as a role. I think Google pioneered some of the the definition of that position, that role, that kind of set of responsibilities and and has really taken the the the concept of, like managing systems at scale to new levels. But back in 2004 you know, even Google was still figuring a lot of this out. So the biggest challenge for us was finding people who had who had experience in managing large scale networks of kind of distributed systems. Andi. So you know we had toe way hired. A lot of we actually ended up hiring a lot of fairly junior people on giving them the opportunity to learn and grow. And there was a lot of fun. I mean, I think that we all had way. We're all doing something that was pretty new. Some of the vendors that we worked with we would take their their hardware systems, their storage systems or networking systems and put them in our environment. And then we would push them sometimes, you know, 10 times farther than what they said they could dio in terms of the number of of IOP per second on a storage system or the amount of bandwidth it that they said they're low balancer could dio So it was. It was a fun time because of that. But it was challenging because, you know, we were doing things that even Theo engineers who built these these systems weren't sure that they they should be able to dio

How did your venture get its first professional funding? What were the challenges and how were they overcome? How'd your fundraising efforts change in subsequent rounds?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
are. Funding was kind of Ah, a different than what you know, you might think of as a traditional start up D C model. Um, we self funded our our first foray into it. And then, as I said, you know, we were kind of bootstrapping it. And then when we brought it into Rackspace, um, our biggest expenses were all around data center space hardware. You know, that kind of thing. Bandwidth on DSO because that was what the parent company Rackspace did as a business. Uh, it was kind of funded through them and through them, giving us allocations in their data centers and and with their managed hosting systems s. So it was kind of a, uh we bootstrapped and self funded up to the point that we rolled it into a kind of a subsidiary. And then at that point, and it became this kind of internal, um, start up within within Rackspace. Um, you know, I think that that that had its advantages and disadvantages, you know, there was, um, there are There were a lot of advantages in terms there. Um, we were We weren't just a customer of a data center company or managed hosting company. You know, we knew the network architect's who were running the whole data centers, and we're able to work very closely with them in ways that if we had just gotten 10 million or 20 million of VC funding, we would never been able to do because, you know, there aren't that many network experts out there at that level. And so being able to work directly with them had huge advantages, you know, at the same time, then they were I think some of the disadvantages were be be because of, um, the difference between what we did. And Rackspace is traditional product portfolio. So, you know, a traditional Rackspace customer at the time paid 600 or $800 per server per month. Rackspace wanted to sell large deals with multiple servers and cell kind of this high touch support concept called fanatical support. And what we wanted to sell was something that was very automated and cost less than $100 a month. And so internally, within within Rackspace, there was always a, uh way were kind of too low cost Thio be of interest to like the Rackspace core sales team and and sometimes, you know, the people would say, Well, why are we Why are we doing this Mosul thing over here? Why are we doing this cloud thing? And so that that would I would say it was one of the disadvantages was, um, trying to innovate inside of a large growing organization is always difficult because you're trying Thio change it from the inside. So, you know, our what we did was different. And if I were going to do it again, I would probably do it the same way. Just because I think that ultimately, Like what? What this kind of business, um needs is that access to infrastructure and, uh and, you know, that was that was something that what we did gave us access to some of the best infrastructure that existed in the mid two thousands.

How did you set the scope for your minimal viable product? How did you get to product-market fit? How did your product evolve over time?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
this was again getting back to our philosophy. Um, we really believed in in this bootstrapping type of model. And so we we got customers onto everything, Um, as quickly as we could and, you know, again that has advantages and disadvantages. I think the advantages come from that product market fit in the product. Evolution. Uh, you know, once you once you put customers on their you're you're getting lots and lots of feedback about Is it useful? Is it good? Is it right? What would they what's missing? And s o you know, your you don't have to guess before you have customers. You're really just guessing about what they want. Even if you go out and you dio you do use a research and surveys, um, without people using it, you can't really know for sure if if it's what they actually want it or not. So I think that that was our approach was very much getting it into, um like our minimum viable product was scope down to to the what we thought was the bare minimum of deploying a Web application, deploying a website and doing that in a way that was really seamless had a very, very automated, smooth experience, but but didn't have, like, a lot of a lot of bells and whistles around that core workflow. Um, now, the I think the disadvantage that we faced, especially because we were running an infrastructure service, is once you get those customers, then you have to take care of them. And they are, uh, you know, they are putting load and there changing the dynamics of the system because of the code that they're running within it. You know, the cloud allows you to go execute any arbitrary code that you want to upload. So you have to think about how do you isolate customers from each other for security, for performance, for reliability? How do you make sure that, um, you know that if they end up on CNN or Oprah, which is something that we used to have quite a few of our customers kind of run into those scenarios? How do you make sure that you can scale up for those traffic spikes and that ends up taking, um, those operations end up taking effort away from product development? So, you know, I think I think again you have to balance the the benefits and and the disadvantages the benefits. You're going to get great feedback and be able to really align with the market. But the disadvantages or you're taking on a lot of, ah lot of responsibility. Just toe operate, um, the and maintain the base level of service once you bring customers into it.

Who were your early users? What marketing channels, approaches, and marketing tools did you use to contact users? What worked and what didn't?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
So our early users were were all, uh, professionals that were very similar to Todd and I. We called them D d I or designer developer and integrator. So it was Web designers who were building websites. Um, Web developers who were writing Web applications are integrators who are, you know, consultant companies that that worked with with companies to kind of manage their online presence and eso, you know, way identified a specific market segment really early on that we we could identify with ourselves and that we thought, you know, we understood very well and then because we were that market, we kind of knew where toe ahead Thio to gain access toe Teoh. You know, the channels that they were reading or, um, you know, forums that they were in publications. They were reading that kind of thing. Eso So that was how we got it started. We we That market, I think, is often drawn Teoh kind of like new things. So we we played up the newness of it, the kind of, um, forward leaning approach. Todd's graded at marketing and branding, and he created these different brands I think blazer out or something like That is what he called our network architectures. Um, you know, we had we had all of these different, different kinds of terms that we would put around the technology pieces that we built. And I think that was effective at at getting people toe to sort of think about it is more than just hosting, um, in terms of channels that that worked and didn't work. Um, you know, we Yeah, we did everything pretty much in on in the form of digital marketing. Eso we were We were doing banner ads and Google AdWords and, you know, search engine placement. And that kind of thing on dat was that was fairly effective. But what? What I think ended up being really, really valuable. Um, to us, the most valuable was we also invested in You know, what you had what people now would call influencers. You know, there were there were certain web developers who who are very high profile and, uh, people who were well known online they might have blog's or, you know, later Twitter accounts are, uh be, uh, people that were that had podcasts and all that kind of thing. And so, uh, one of the there was a set of early customers that we went after who were, who were people who covered everything that Apple did. So that was kind of one very specific niche that we went after because, you know, people loved the Steve Jobs keynotes they loved every apple product reveal and and these these kind of independent bloggers and journalists who recover that would get massive amounts of traffic on the days of of those keynotes. And so, you know, they often, um, were graded covering that news but didn't really have the expertise to manage a really high traffic site. So we went and and, you know, offered several of them hosting and and in exchange for some recognition on their site, you know, powered by moso kind of logo. And, uh and that ended up being a really effective channel because, um, you know, people were used thio websites crashing because of of traffic spikes. And so you know that that was actually a thing that people ended up commenting on when they did the iPhone reveal. And when they did the you know, all of these reveals and and these websites were running on most So you end up getting us a lot of coverage and kind of visibility. So, you know, I think I think the the that whole that that whole approach now has an entire industry built around it, you know, influencer marketing Onda and I think that if you have a specific enough market that you're going after, it can be a really effective way to, you know, to get exposure. But also, you know, you can you can prove that your that your product, um, works because a lot of times those influencers are also power users of these products.

How'd you hire, incentivize, and track the progress of your sales and marketing team including agencies and part-time workers to scale user base?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
you know, we had obviously, ah, lot of tracking of kind of our vital stats around sales. One of the things that that I learned from Rackspace and that Rackspace was very, uh, just laser focused on was was turned. Um, turn, you know, is is through the loss of revenue and the loss of customers. And the business that we were in was A was a recurring revenue model business. So, you know, if we brought on a customer and and they paid us, they pay us $100 a month, $300 a month, you know, it usually would cost us, uh, something like 3 to 6 months of revenue toe. Bring that customer on. So if they left before we got to that that payback period than we never made any money on them. But then every month after that, you know, it was like it was a great return on them and, you know ah, lot more money flowing down to the profit line. Eso You know, if you can keep a customer, it's actually in some ways, you know, more valuable than than a sale, because they paid themselves off. And, you know, they're more profitable. So the things that we focused on, we're obviously the standard, you know, customers coming in sales and everything. But we also did have a big focus on turn and and on, you know, making sure that customers, uh, we're stayed around for for a long time. Eso we had we had bonuses in place for our sales team and for our support team. And those bonuses were tied thio two new sign ups to customer retention to keeping turn very low. And we did, um, we did a lot of surveying of our of our customer base, Um, surveying after every interaction with with the sales or support person. You know, these are optional surveys that people could opt into eso those those metrics. Those ratings ended up playing into the kind of evaluation of those people. I think that that the one thing that I think was in a recurring model that sometimes gets overlooked is that is that turn. You just have toe really focus on that turn. Um, because it's just it's how you end up really growing your customer base over time is by not not losing customers, not not giving them a reason. Thio, toe leave and walk out of the door

Who were your competitors when you started and how did the competition evolve? How did you create a competitive advantage and a unique selling proposition?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
so the competitors were really different approaches. At the time that we started, they were They were again. You know, you go get a server and you're responsible for it, and you have to take care of it or you go to virtual hosting. And the experience is maybe not as good, not a scalable, not a secure eso. So when we first started, we we were really that are unique. Selling proposition was all about the type of service that we were offering because it was different enough that that it didn't It didn't look like it wasn't really. You couldn't confuse us with the other options because the whole experience was very different over time. There were other other companies that started to introduce similar things. And then where I think we ended up having, um, you know, a big shift in the industry overall was with the launch of of Amazon Web services, um, eso we had we had launched in in 2004 and then and then had kind of done our relaunch with Mosul in 2005. And in 2006 early 6 2008 Ws launched S three, which was their storage service. And at the time, everybody is kind of like What is this s three object storage. Like, why is Amazon doing this is such a weird thing. Then in October of that year, I believe they launched E C two, which was there there, you know, kind of virtual virtual private server. And again, I think everybody in the industry was like this. There must be something here because Amazon is very disciplined and, you know, and but, you know, like, what is this exactly? And, uh and I think that it was it took up a few years before, um, you know, maybe not. If you probably took a year and a half before, I think easy to really just started toe skyrocket in terms of its usage and its mindshare especially. And through that time, I think we were able to see that that this was this was again with something different. You know, it wasn't a dedicated server that you were solely responsible for. It wasn't just low in virtual hosting. It wasn't what we were doing either, which a lot of people would consider probably platform as a service. It was somewhere in between platform as a service and just running a server. And the attributes that we that we looked at and identified that we both had in common were very high levels of automation. Um, elasticity in the sense that you could you could start small and scale up and scale down. Um, it had ah lot of self service eso It wasn't something where you had to call and talk to human anytime you wanted a new server or new website, and so we could see that this was that this was taking a lot of the advantages that we brought to it. But applying that to servers and we acquired a company called SLICE Host Thio to fill out the that kind of automated server portion that we didn't have in our product portfolio. Um, and around that time, we re branded from Oso to to Rackspace Cloud. And I think that that when we did that, then our competition really became much more, um, kind of head to head with, uh, with with AWS and eventually Google Cloud and Azure. Um, you know, I ended up moving away from the company around that time and and, you know, I I think that that some of the remaining leadership at Rackspace um you know, after a couple of years, they sort of bought into the idea that there is no way toe really compete with Amazon. You know, if you if you look back to the first few years after after AWS launched, we actually were We're very close. In terms of the scale, Amazon was investing billions and billions of dollars. And so I think it was kind of scary to see that happening with with one of your competitors. Andi, I think that that, you know, the the company just wasn't sure it wanted toe to continue Thio try to invest in a head to head battle with with with a W s. I think that that, you know, my belief is when we started all of this, we were in a head to head battle at Rackspace with companies like M C I and Sprint and A T and T and others that were much better capitalized. But we won because of a of a specific focus. And I think that that if you look at the marketplace now, you know AWS is obviously huge growing It's a great business, but there are a lot of a lot of offerings out there that that, you know, are very focused in terms of how they provide an email service or a database platform or, um, you know, hosting services like metal if I or others like that. And so I think you know that that it's been a very interesting 15 years since we launched to where we are now. I think that the things that applied back when we launch still apply, which is that you know, you find your spot in the market and you have a super, um, super strong focus on experience for a specific group of people. And if you do that, you can compete with with anybody. Google Amazon Whoever is out there, um, if you just build a better experience for for ah, segment of the market where there's where there's real demand

What were the major exciting and memorable moments? Were there also any moments that almost got you to quit? How did you get past them?

Based on experience at: Founder, Mosso
Summarized By: Jeff Musk on Thu Sep 10 2020
eso major exciting and memorable moments. Ah, lot of them, you know, I mean, our I mentioned our team earlier. We built out a really cool team that had a lot of people that, you know, I just I loved working with them and, um, we were all of the time, you know, we were doing things that nobody else was doing. And so that's that's always a lot of fun. And we also tried to celebrate together. When we had winds, we had a bell in the office, and every time that that, you know, we closed to sell, somebody would go ring the bell and we had a counter on the wall that was hooked up to our our sign up system. And so is people were going through the website and signing up. You know, the counter was going, and whenever we crossed another 1000 customers, then we would we would like have have we're in San Antonio that we have, like, a margarita machine or something and, you know, celebrate those milestones. But, you know, they were also very tough, tough moments, one of one of our mentors. And we were doing that said that that starting a business is is a continuous cycle of of the highest highs and the lowest lows and the feeling like, you know, your your you always feel simultaneously that you're just about to fail completely and also just about toe breakthrough and have achieved total success. And it's true because when you're starting a business, you know, you you are responsible for for so many different things and whether you're talking about product design or, um hiring and managing a team or customer service or finance or, you know, whatever it is. And in any given day you can have great results in one area of the business and have some disaster that you're dealing with another area of the business. You know, one well, at one point, you know, we were making a lot of progress in our in our product and our product design and rolling out a lot of great features. And then we had we had a big failure in the data center and, you know, just spent 24 hours of solid, painful work, you know, trying to get everything back up and operating it, make sure we didn't lose any data and and it's those Those failures are very public because customers were impacted. They're calling in their angry rightfully so they're upset on dso. You know, you've gotta You've got to deal with that and talk to them. And at the end of that, you know you can you can feel very, very beaten down and and, like, maybe this is, uh you're not doing the right thing and you shouldn't continue. But, um, you know, I think that that's pretty natural state thio thio have Sometimes when you are, when you're when you're really out there creating a business and building something new, there's there's a lot of those challenges that come along and, uh, and, um, being able to push through those that perseverance is Sometimes the only differentiator between successful businesses and unsuccessful businesses is you know that that willingness and that that desire Thio not quit

What responsibilities and decisions does one handle in a job like yours? What are the challenges? What strategies are effective in dealing with these challenges?

Based on experience at: Executive Director, OpenStack
Summarized By: Jeff Musk on Thu Sep 10 2020
eso. Yeah. So for the last eight years, I've been the executive director of the exact foundation, and open stack is an open source project for cloud computing. We started with some of the code from our ex based cloud. We open sourced it under an Apache license, and then we we built a global community around it. And then we started a nonprofit thio, kind of manage that community, and so that that's what I what I've been doing and a z executive director again it xem ultimately, you know, I'm running the organization. So there are. There's a wide variety of things. Finance related, uh, employee management related. We throw events. You know, I never knew how much I was gonna learn about the events business, but we've thrown events, um, in China and Japan and Australia, Germany, France, Spain, Canada, all across the US, Ireland, all of these places. And eso um, you know, there's so much that you learn about along the way. And it's one of things that I actually love about. Business in general is, um, you know, I enjoy learning new things and getting the chance toe experience new things. And when you start a business. You definitely have lots and lots of new things that you're going to get that you're gonna find yourself having to deal with. Whether that's, uh, how how to do business in different cultures, how to work across time zones like all of these kinds of things. Um, there really fun. I think you know the challenges. Um, it's interesting. Thio compare and contrast kind of the commercial world, where with Moso in the Rackspace Cloud, the goal that we had was toe was to make money ultimately, you know, toe sell customers and get them to pay us, build a product that they're willing to pay for. Um, in the nonprofit world, you you need funding. So we have to convince companies toe support our mission and and provide us with funding so we can go out there and do work. But ultimately, that's not the goal of itself. The goal is not to not to just bring in as much revenue to the open stack foundation as we can. The goal is to actually build the most successful community of software developers that we can and and the way that that that that happens in open source is through community. So you know you in an open source communities. There's there's very little that gets done through authority and, ah, lot that gets done through influence. So even though I have the title of executive director, I don't really have the authority or the power to tell anyone in the community to do anything. You know, I can't say go right this feature because I think it would be a great feature. I can try Thio suggest, you know and say, I've talked to these big companies like Workday and CERN and China Mobile and, you know, steel companies. And they all think that this feature would be very valuable. Who who's interested in developing that, are improving it. But you know, I can't, like, go out there and make a priority list on the technology side. So I think that, you know, that's one of the one of the big differences in the roles is the way that work gets done? Um, in you know, really, in any kind of community, but especially in open source, where there's there's very much of a grassroots model to it, the challenges you know, building relationships, maintaining those relationships, learning from those relationships and then using that knowledge to kind of influence the direction of the community over the long term.

How did this open-source project start and get to the current stage? How did the community around the project evolve? What led you to contribute to this project?

Based on experience at: Executive Director, OpenStack
Summarized By: Jeff Musk on Thu Sep 10 2020
So we started the project at Rackspace because we had, um e think they were kind of maybe 22 main ways to think about it. The first way it was was that, you know, we had, um this cloud that was growing. It was the fastest growing part of expected business. But, you know, we were we didn't have the billions of dollars that Amazon had to pour into their data center and AWS and all of their software development. And so, you know, I think that at some level, we thought if we create an open source project and get momentum around it, we can build a coalition of people to create technology that we can all benefit from the other side that I think was interesting. Proofpoint, for me at least, is within the Rackspace cloud. We had companies who were, you know, major Fortune 100 companies who were coming to us and saying, We like what you're doing, but we will never be able to run everything that we dio inside of anybody's cloud. So can you just give us the software that you use for your cloud and we'll run it in our data? center and we said, You know, for the first couple of years we said, No, no, that's not we're not a software company Were data center operations, you know, cloud Company. But I think that that we never we probably never would have gone to the point of licensing and selling our software just because we were not set up to do that. That's not at all what we did, but by open sourcing it, what it does is it gives people access to the software. Um, it gives people the ability to contribute to it, but it lets those those companies that want to run it, they can run it. And over time, you know, within the first couple of years we had a lot of other companies that got involved companies like red Hat and canonical who know how to take open source software, go sell it to an enterprise, provides support, provide patching and long term updates, and those kinds of things that an enterprise needs in order to rely on software. And they can kind of fulfill that role of the software company that that a Fortune 100 company, you know, they want to have a contract with in order to use in order to use a piece of software, whether or not it's open. Source. I don't think that we Iraq space ever would have would have been able thio do it as well as a za red hat in terms of all of that. Now, Rackspace did offer private cloud services and and still does, you know, based around open stack on day. And so you know it also creating new business opportunities for rack space. And so that that I think, you know, the way that we we saw. It is a way to create an entirely new market space around cloud by giving everybody this kind of base layer of software that we could all interact where improve you sell, build products and services on

How are feature requests prioritized? How are tasks assigned to contributors? What percent of your project-effort goes in providing support and responding to users?

Based on experience at: Executive Director, OpenStack
Summarized By: Jeff Musk on Thu Sep 10 2020
um, within the Open Stack Foundation. We now have a number of projects in addition to open stack. Open Stack itself is a very, very large, open source project. It's one of the top three open source projects in the world by activity, so the most active is the most active to our Lennox and chromium, which is the open source project behind the Chrome browser and Chromebooks and all of those chrome related, uh, devices and kind of that ecosystem. And then there's open stack. And so we have, you know, thousands of developers that write code thousands of changes that come in every every year, every month, actually, and eso for for a project like open stack there are. There are a lot of processes that that are in place to kind of manage that workflow in that workload, open stack is divided up into um into smaller sub components like Nova handles, virtual machine management, cinder handles, storage device management and each of those sub groups has has a leadership team of core reviewers who who actually approved code changes. Um, there's also we do a six month release cycle and heading into that release cycle, each project kind of reviews and sets the priorities that they want tohave for for their project. And then there's an overall technical committee that tries to set um, community wide priorities like one of the ones that was completed. Um, pretty recently was was an upgrade in the in the version of Python that that was open staff is mostly written in Python code and the version of Python that we started out on. We started open. Stack was just end of life, and so you could no longer run that version of Python. So we need to make sure that the entire community moved to the newer versions of Python and did it well ahead of that deadline. And so, you know, there's a lot of structure in a project like Open Stack. We have other projects, like there's one called Kata Containers. That is, um, is a container related project for security and isolation, and it's it's smaller in terms of the number of contributors and the rate of contribution. Um, it still has dozens of companies to participate in it, but they have I would say, you know, um or fluid and informal process where people can make proposals that use get hub for their code management, and people can propose things in an issue on that. And there's kind of a core set of maintainers that review those issues and approved code s Oh, you know, I think that the one thing that I've learned is that there is no there's no single way to do open source, right? Um, ultimately, you know, open source works because, um, because there's an engaged community who's contributing to it. And so the I think that the right way to do open source is whatever keeps a community engaged around a specific topic and that might be get hub, it might be not get hub. It might be, you know, having ah, a single project leader who makes those decisions or it might be a committee. Andi, I think that you really just have to build that structure around, whoever the community is that that's coming to participate because it's all really driven by participation and interest in that level.

What lessons have you learned working on open-source projects? What would you do differently in the next open source project?

Based on experience at: Executive Director, OpenStack
Summarized By: Jeff Musk on Thu Sep 10 2020
I learned a lot of lessons. I think the one about influence to me is is a big one. Andi, I think that that that's not that. That's a lesson that's valuable, not just for open source. I think you can use that in, um, in commercial context as well. Uh, you know, we say that we have a model for for activity, which is you start with relationships and through those relationships, um, you you gather knowledge and that knowledge and those relationships give you influence. And then in a community, you can use that influence to to sort of direct activity and actually, you know, cause things to happen. And if you think about that as a model of relationships, knowledge, influence, activity, you can apply that in in any company, you can apply that in kind of like any context where there are people. Um, it's really about about understanding things and getting buy in from people as a whole. And so I think that Zatz something that, you know, I have I have learned a lot about through through working on the open stack project. Um, I think that in terms of you know what? What would I do differently in the next open source project? Um, I think that it's hard to open stack again. It's so big that it's it's hard to kind of, um to just apply the things that I would say. We've learned with Open Stack to every other project because they're like the dynamics could be very different. But I think if if I were to work on another project like Open Stack, that was very high growth and high scale. I think that one of the things that that that happened as open Stack was growing really rapidly. We started to put some of that structure in place that I was talking about earlier. And in some ways you know, that structure was their toe. Actually slow things down because, you know, if you have 5000 people who are all writing code at the same time, it's very hard. Thio. Make sure that it's that it's that's all code you want in the code base, that it's all good code, that it's all you know, secure code. And so we put a lot of processes in place to slow the too slow. That flood of of new contributors and code down to try to maintain a level of quality and focus, you know, to varying degrees of success. I would say I think that what happens, though, is in in any anything any, especially any open source project. There's kind of this this. It's like the Gartner hype curve. You know, there's like a huge peak and interest and excitement, and then it trails off, and then it kind of like goes to the level of productivity. And what I think we we didn't necessarily do is once we got through that initial flood of excitement, we kept some of those barriers in place that were meant thio help the project in the beginning and maintain the level of quality we kept them, even as as interest had had subsided from the peak. And that actually made it harder for new contributors. In some cases, you know, to come get involved to step up into leadership roles. Andi. I think that that's a that ends up being when you need new contributors the most is when when some of the old ones, they're stepping away and I think that way, you know, we didn't necessarily respond quickly enough. So I think that that would be one thing that I would probably do differently is try to make sure that that again you know, like the ultimate goal is to have an engaged community that's that's contributing, participating. And so if the dynamics of the community and the situation change where you're starting to slow down or make that hard, then you need to change the systems and the processes really quickly, so you don't just kind of run, run them off and lose all of that interest and engagement.