Open Source For Business podcast with Gil Yehuda (Ep 01)
Categories:

E01: Gil Yehuda on Open Source Program Office’s & Verizon Media’s open source software projects and strategies

Intro

Henry Badgery: Hello and welcome to ‘Open Source for Business’ brought to you by OpenTeams, the first B2B marketplace for open source support and services. I’m Henry Badgery and hosting with me today is Eunice.

Eunice Chendjou: Hi everyone, my name is Eunice and I am the COO at OpenTeams. Henry and I are very excited about this podcast series. This is our first episode with Gil Yehuda, the Senior Director of Open Source at Verizon media. Verizon media is home to media, technology, and communication brands that nearly 900 million people across the globe love and trust. Some of the brands that you may know include TechCrunch, Yahoo and HuffPost.

Henry Badgery: Gil Yehuda is widely known across the open source community and many consider him to be one of the pioneers of what we call an OSPO that stands for Open Source Program Office. Gil is active on social media and can be found on Twitter @GYehuda, and also on LinkedIn. I’d also recommend to anyone listening to go out and check the impressive work that he and his team have been doing to fight COVID-19. They recently released a potentially lifesaving tool using their Open Source Search Engine Vesper. So, since I don’t want to spoil the episode for you, we’re going to dive right in. Gil, thank you so much for joining us today.

Gil Yehuda: Thank you for having me. Thank you both. It was a very lovely introduction.

Henry Badgery: Eunice and I are really, really looking forward to this episode. So, let’s kick it off. I know that you’ve published a lot of articles, you’ve given presentations all around the world, and I definitely consider you to be a thought leader in the open source space. But when researching for this conversation, I learned that you didn’t always have a positive view of open source. So, can you take us back a bit to give us an idea of how you got here today?

Career Path To Verizon Media?

Gil Yehuda: So that’s true and it’s peculiar to look back and to think about it from, you know, from where we are today to where I was many years ago. But when I first heard about open source, it didn’t make sense to me. And I didn’t think it was a good idea. It sounded like, people were sort of a hobbyist or voluntarily putting code into some not particularly structured or organized, common area and the hope was that great things would happen and it just didn’t seem very strategic, it didn’t seem very reliable, it didn’t seem like you could run a business on it. And I sort of I just associated it with like this frivolity and hope.

The truth is I was wrong with that. I simply was mistaken, and I think there were a lot of people and maybe still are who hear about open source and they sort of assume it, to just not make sense to not work with them. So, what I did is I learned more actually I was exposed to some really great thought leaders in open source, and I just listened to them and challenge myself. Well, maybe I’m mistaken, and when I heard more and learn more, I realized that this isn’t frivolity. This isn’t like, you know, hey, let’s give code away and see what great things might happen. There’s actually a business strategy around it. So, I was impressed.

Eunice Chendjou: That’s awesome. Well, I think what I’m curious about is, you know, what was that big realization?

Gil Yehuda: Okay, so there was this interesting realization I had that I look back and it sort of like concretized for me. When I was growing up and when many people were growing up, we were told that you know, you don’t want to trust strangers from the internet and you certainly don’t want to get into a stranger’s car because that’s dangerous, right? I mean, strangers from the internet and a stranger’s car seem like a really bad idea, and yet getting into a stranger’s car on the internet is actually a multi-billion dollar business proposition right now, right? And staying in a stranger’s house on the internet is a multi-billion dollar business, right? We are doing things, right, we do things that completely made no sense way back when, because strangers were untrusted entities and interacting with them was something that you incur a certain risk. And what we realize is that if we change something about that dynamic, so there’s a trust model, reputation system, a way to compensate if there’s a problem, it doesn’t mean that all problems go away.

Occasionally, there are real problems, but a bias of saying; oh, that’s crazy. A stranger from the internet, I’m not going to get into their car is Oh, no, I’ve no, you know, punch in and say; hey, a stranger from the internet please pick me up. I kind of know enough about you to trust that this transaction is going to work well. And that’s what, that’s what open sources, a stranger from the internet is giving you code that you’re putting into production. And you’re able to do that because you run a test case. And there are all these other people that are testing that code too, so It isn’t just like, hey, a stranger is going to write the code for me, and everything’s going to be good. I have a business model that regulates what was previously a risky transaction and now it’s a business transaction and I’m okay with that.

Eunice Chendjou:  Wow, that’s powerful. That’s really powerful, especially from the change that we’ve seen over the last decade from not just open source as a whole but also so many other companies have emerged to welcome strangers. Obviously, there’s some differentiation you know, in things happening. But I think for those listening today and from the transition that you’ve seen from, you know, not being so pro and open source on like you know, you talk about open source is so much passion. I know today, you’re an open source, you manage an Open Source Program Office, I want to know what exactly an Open Source Program Office (OSPO) is and you know, why would any company want one?

Open Source Program Office’s & Their Importance

Gil Yehuda: Okay, that’s a great question. And so there’s open source which is you know, this code. A program office is an organizational structure inside of a company sometimes it’s just one person in a role and sometimes it’s a small team. But it’s a part of the company that is the center of competence, the center of gravity, or the plate, the go to place for the open source related issues the company has to deal with. For companies that write software, it’s the person or group of people that help them understand, what are the license implications of the software they are interacting with and for groups that publish software for, you know distribute software, then they sometimes handle compliance issues around the distribution of software with respect to the license.

In many cases, they help with getting stuff onto GitHub or you know, to publish open source code or to participate with foundations or to work with other communities around growth or fixing code or launching new projects or strategically deciding if to open source projects, which should our company use and do we know enough about the dynamics of that community to help us decide how we want to strategically use one of two open source options or open source versus a commercial option. These decisions are made by companies. The Open Source Program Office is the organizational structure that says we’re the group of people that can help you because we do this all day like we understand the company and the nature of open source to help make those decisions.

Eunice Chendjou: Well, and in your perspective, do you think every company that uses open source should have an Open Source Program Office within their organization?

Gil Yehuda:  Yeah, I can. Yeah, I do. I mean, it’s I think that a lot of companies should have. I think that more companies should have Open Source Program Offices then do it. It’s no longer like this you know, luxury good for the big tech companies. If you’re a company that deals with software, you are most likely in fact probably guaranteed to be dealing with open source software, it’s simply impossible to deal with software today and not interact with open source.

So, somebody is making decisions about your open source strategy and sort of recognizing that framing it, documenting, and sort of putting a bow on it and saying, Yeah, this is their Open Source Program Office seems like a great way to get accountability and consistency in those questions. So, I think that even smaller companies that maybe didn’t think that they need an Open Source Program Office, they do need to have that role. Maybe it’s a halftime person, but it’s somebody who should understand the behaviors of open source and how it impacts the company’s goals.

Being Open Source Friendly Attracts Great Talent

Henry Badgery:  I think that’s an imperative and it’s definitely a change that’s happening it seems within these companies, not just the large tech companies like you said, but the smaller ones because I think people are starting to realize that we need to take open source seriously. Everything is reliant on this how critical infrastructure relies on open source software and we don’t even know anything about licensing. So, that’s great that you guys are all filling that role, guys and girls, what I found is that for a lot of these software companies, there seems to be this shared belief that an open source first approach attracts great talent. So, can you tell me please, what does it mean, first of all, to be open source first and why does it attract talent?

Gil Yehuda:  Okay, so it’s a good question. I’m going to be a little literalist about this. I don’t know about the word first. I don’t know. If I mean, I don’t know if open source first, is the most strategic way of saying it. I would say open source is friendly because everyone says for security first, mobile-first, customer first, everything so first, how do you know which is first? Right? Can’t have everything first? So I think that open source friendly is important and open source aware and open source strategy is important. And yes, it does attract talent. And I think that’s an important spillover benefit. So, we do open source because it’s we’re sort of compelled by the industry, we’re in.

We simply can’t not be involved in open source if we’re involved in software but once we’re involved in software and we’re operating in an open source aware, open source friendly, strategic, we’re conscious of it, you know, we’re not relegating open sources we’re not dismissing it, then, then we get this very interesting benefit. Engineers want to have engineers want to be consequential. You could be a hero at my company. If you’re a great engineer, you could be a hero at every company if you’re a great open source engineer, right? I mean, think about the reach that you can have if you’re the kind of person who can write the software that everyone in the world wants to use. I want to hire you, right? Like, your software is so awesome and not only do we want to use it, everyone else wants to use it. That’s exactly who I want to hire. You want to be hired by the kind of company that recognizes that.

Pragmatically, speaking from a training perspective, I would rather you already know the technology than have to train you on something proprietary, so it’s just easier, you know. So, from a hiring and retention perspective, you want to stay at a company that recognizes and allows you to be a hero everywhere. So, this absolutely benefits from a hiring perspective. It also helps to determine who to hire, like if I could see your code and I see that you’re participating in a project that I care about, and can harvest you and basically say; Hey, I see you’re really awesome in this community. By the way, we’re building out a team of this community, would you mind if we paid you to have a full-time salary with benefits and you can continue to work on this project but you know, you’ll have to, like, allow us to maybe assign the priorities, you know, because, but we’ll pay you like there’ll be a salary.

Henry Badgery:  It’s a win-win.

Gil Yehuda:  It’s a win-win. And if you don’t want to do you have to do you can continue to do it on your own independently if you if that’s your gig, fine. But if you wanted somebody to pay you, there are companies who care about the success of the project and they’ll pay you to continue doing your work if the company is open source aware, open source friendly, and gets it. So, I think it’s important for the companies who really want to have those positive outcomes to sort of step back to challenge their assumptions about open sources being trivial and sharing and whatever and like hope, saying no, this is actually how we’re going to operate and compete.

Eunice Chendjou: That’s really powerful. I think any developer today, hearing and listening today’s will probably be considering Hey, maybe I need to invest more time into contributing to building up themselves you know, software out there. But what I want to think right now is I want to step back for a second and ask you, you know, maybe for a developer that is within a company. Imagine if you were someone who worked at a tech company right now, you know, how will you pitch, you know, an Open Source Program Office or face, you know, to a group of executives or accounts, your executive team about, you know the power and potential of open source?

Gil Yehuda: That is a very important question. It’s a very important question. You know, I would explain to executives in less of a pitch or more of a, you know, let’s just take a look at our reality. The reality is, is that decisions are made on a daily basis that either incur additional risk or additional cost or additional, you know or give us an advantage. And these are decisions that are being made because of open source considerations. You are happening today. And as an executive in a company, you need to recognize what they have some sort of accountability, manage, course-correct if they’re the wrong decisions, at least know who you know, know what they are.

An Open Source Program Office is a management structure that allows you to do that. Because without an office, those decisions are being made anyway, you just don’t know by whom, you know, those companies who tell me Oh, we don’t contribute anything in open source? No, actually you do. You just don’t know that your engineers are doing it on the side. So, you actually are you’re just ignoring what and you don’t know where or why, or we don’t use any open source here? No, you do. You just didn’t know it. So, I would say that the pitch is you have to be responsible for your technology portfolio. And an Open Source Program Office gives you the ability to have that kind of management control and responsibility for those decisions, figure out what they are measuring whether they’re good or not, course correct and then communicate that so that as an organization you make those decisions strategically and wisely.

Measuring The ROI Of Your Open Source Program Office

Henry Badgery:  One thing that comes to my mind when I’m thinking about this setting up an Open Source Program Office, based on what I’ve learned from you today, I assume that is a successful Open Source Program Office is one that returns a positive ROI. Is this the case? And if so, how do you measure the ROI of your Open Source Program Office?

Gil Yehuda: That question comes up quite a bit. It’s a very important question. And I’ll take I don’t know maybe two approaches to it. 1. Most of us think about ROI in a very transactional way. Like if I build a product, it will cost me a certain amount of money, I will get returned from the product. Open Source and actually most platform investments are infrastructure. Investments, infrastructural investments return ROI in a very different way because you don’t monetize infrastructure directly, you monetize it through spillover benefit, right? That you what, you know like you build roads and highways and bridges. Governments build all these institutions and there’s no ROI of the road, right? There’s no ROI for the bridge, the bridge improves commerce and commerce provides the ability for people to travel across the valley to the store to then improve up tax revenue for the community as a whole. So, the spillover benefit is the enablement of other things to happen.

That’s the challenge that infrastructure people have all the time, which is you can’t really ROI the infrastructure, you ROI the capabilities that are enabled by the infrastructure. So, infrastructure needs to enable those possibilities and that means that attaching the ROI directly to the infrastructure reduces infrastructure to product and then it no longer is infrastructure and it can’t be useful. So, I challenge the, what’s the ROI of it and instead look for what the spillover benefits are. And the primary spillover benefit for an organization to have an OSPO, an Open Source Program Office is to have the capability to do all the things that an OSPO does as a company, to be able to bring in open source properly and know what they’re bringing in and in a secure and supported manner to contribute to open source projects in a smart way, to acquire companies that use open source and to incorporate that into their you know as part of the integration, to participate and change the economics of open source in those communities that are important to it and to both reduce tech debt and improve the option to sell product through open source.

You can directly reduce risks, increase opportunities, and bring financial benefits. But in order to do that, the company has to be able to it has to know what these licenses mean and after knowing what an open source foundation is, they have to know what they can or can’t do and a GitHub repo, and the Open Source Program Office that gives them the infrastructure to even participate in those money-making or money-saving activities.

Henry Badgery:  That was fantastic. I think you really gave some great metrics that companies can actually base the measuring of their Open Source Program Officer success off of.  One thing that stood out to me in particular about Verizon media when researching you and your team was that you had published guides and videos all over the internet, teaching others how to set up an Open Source Program Office.

I love this mentality. It’s very open source in nature. And so for those of you who are listening and interested in learning more about actually starting an Open Source Program Office and running one then I really implore you to check out one of their guides, which I will leave in the description below. But otherwise, I’d actually like to shift gears a little bit now, as I know that Verizon media manages quite a few open source projects. So, can you tell me a little bit about these projects, please?

Verizon Media’s Open Source Projects

Gil Yehuda: It helps to understand Verizon Media as a company. It’s a fairly new entity. It’s really a combination of a bunch of other companies but primarily Yahoo and AOL and the 50 companies that Yahoo was composed up and the 50 companies that AOL was composed of, and a couple of Verizon entities. So, it’s like a mash-up of like, 100 of these companies under the guise of these two larger companies. And all of these companies have been participating in open source for many years.

I came into the company through Yahoo and I think that Yahoo was one of the more prolific participants in open source. So, most of the things that we have really come from that pedigree although not all you know, obviously the other companies that are part of our family have participated in some very important ways. We have a lot of projects and a lot of them are in the big data space that seems to be an area that we participated quite a bit in and we take a little bit of credit for maybe even creating it and popularizing it with our work on Hadoop, and the greater Hadoop ecosystem, bringing it to light, publishing it to Apache Software Foundation and then all the other projects around it. But that’s not all right, we do operate in other areas, obviously we’re very interested in security technology as is everybody and we publish a whole bunch of security-related technology projects.

We’re involved in mobile technologies. We publish a ton of mobile apps and we’ve open-sourced a bunch of things in the mobile space. We do a lot with DevOps and with you know, folks who build software and monitor software that’s in production and you know, manage large scale systems and we’ve published open source projects in that area as well.

We basically look at all of our investments and say, if we felt the need to write the software because it was a novel problem that no one had really solved the way we felt that needed to be solved and we invested, why not share that? Like, why not allow people to see whether we did a good job to give us feedback, to use it? I mean, it’s we’re not competing in this space. It’s not, that’s not where we make money. And by offering it, they may find uses of it that we didn’t expect, they might contribute some of that back. Or they might come back to us and say; this is wrong thinking like you shouldn’t have done it. And that’s important feedback to have because then we can course correct and actually ask yourself, so maybe this is wrong thinking maybe we should rethink the strategy and we wouldn’t be able to do that without open source of the company.

Henry Badgery: Could we talk about Hadoop as I know Verizon Media is a contributor. What do you think it would look like? What are the benefits of Hadoop being open source and what it would have you noticed that say if you met imagine that they became closed source like what’s the stock difference between the two?

Gil Yehuda: Gosh, what was the Hortonworks acquisition by Cloudera? I don’t remember the number offhand but it was a multi-billion dollar acquisition. This is a multi-billion-dollar business there, right. So, the difference between Hadoop being open source and not open source is, you know, a few billion dollars of market opportunity. And the same with a bunch of other businesses that were launched from a former Yahoo is in some cases launched directly with the support of the company.

One benefit is just the financial benefit of the existence of these technologies in the open and the support infrastructure for others to use, Hadoop in particular has been so revolutionary because the prior to that there was this model, this operating model that was limited to the size of your database. And, you know, by reversing it and saying no, we can actually have this much greater expansive view of data, we now enjoy these amazing products and services on the internet that, you know, feed us billions of tweets and tons of stories on our walls of all types of social networks. Those you can’t build something like that on a database. You know, you have to know that an expensive system, a big data system that has replication, and that has really the features that the big data ecosystem provides.

This has revolutionized the way we use the internet and it’s revolutionized the way we manage our software infrastructure. I mean, imagine if we still had all of those entities on our payroll, having to fix every single thing for years and years and at some point them leaving the company because they wanted another job, we would have to somehow own all the talent in order to manage our own infrastructure and huge expense without the value for it.

Back to your question on ROI, what’s the ROI of keeping your code proprietary or what’s the cost of keeping your code proprietary. Like why have you strapped yourself down to that tech debt that forever on after you’re going to have to be the only entity in the world that’s going to be able to fix your code? Why would you want that? Why wouldn’t you want other people to work on your code with you?

Henry Badgery: Definitely, I couldn’t agree more.

Open-sourcing A Company Project

Eunice Chendjou: That’s powerful. That’s powerful. And I think this is, you know, what you just said, it’s like, you know for a lot of people today the idea of giving away, you know code for free sounds, you know perplexing even for me it took me a long time to wrap my head around the notion that if you open source your code, you can actually gain many benefits.

I’m curious to find out because I know you mentioned a few times, why would anybody give it away your code, but how does it actually work I know that Verizon made up with a lot of you know, of code out there as open source but you know, how do you like, what is the strategy behind it?

Gil Yehuda: So I’m with you Eunice, by the way, I also thought it was a hard thing for me to wrap my head around which is why I didn’t think it made sense at first, but you know, it took time and when I saw what it did, I realized that it can, it doesn’t always work out well, but it can work out well. So, the mechanics are a strategy around it. We don’t just open source everything. And we don’t just, you know, put stuff on GitHub repository and help because that’s not a strategy.

It isn’t the abandonment of process but it’s a refinement of a process where a group will come to us and say we have this. So, a classic case, as a group comes to us, they say we have this project, we want to open source it. And we’ll challenge them and say, well why do you want to open source it? What’s the benefit to you? What’s the benefit to the company? And in that challenge conversation, we’ll learn things like why did you write this code in the first place? What unique problem does it solve? Was there something else that solved it? In many cases, we try to have the conversation before any code is written sometimes it’s a little too late. We’ll ask them, well, do we use this code? And sometimes we’ll say no, we don’t use this like, wait a second. How are we going to convince people in the world to use a project that you couldn’t convince your teammates to use? That’s not going to work so well, right. So, we ought to be able to prove that this is useful internally before we ask people whether they think it’s useful to right so let’s hold them to, to measure.

Sometimes it goes the other way. Sometimes I’ll go to a team and I’ll say, I do this like a little bit of a trick. I go to a team and I’ll say, we’ve identified your project as a candidate project to open source, be prepared. And about a month, we’re going to put a version of your code on GitHub, and they freak out. They’re like; oh my God we are not ready for this. They add test cases, and they start documenting the heck out of the code and they fix the code and I tell them three weeks later, just kidding, right? But look at how much better your code is. Look, you actually have cases you’ve documented it. We can add value, even if we didn’t open source the code if we merely threatened to do so because it forces us to think, wait what if other people look at this code? It’s got to be great.

So… we up-level our engineering capabilities by recognizing that code could be public. People could see this. Of course, they see the binary. They see the application, and they have opinions about whether it does the right thing for them. What if they saw the source and they had opinions about whether we covered it right. You know, that refines us so the mechanic is they come they ask, we negotiate, we figure out, you know, are we encumbered by it as a strategic, are there patents that we have to that was filed on the code that we’d need to talk to the patent lawyers, what licenses would be appropriate for this, are we forced into any sort of licensing schemes or do we have options?

Who would use the code, have we tested that, like, you say, an engineer says, oh, people use this code? How do you know, have you gone to a meetup? Did any like, have you propose to people would you want this code? If we open sourced it, then did they say yes like; yeah, sure or enthusiastically yes. Like, did we do any of that vetting to prove that there’s something here? If so, sure, what’s open source didn’t invest. If not, then at least we have this great conversation to refine. Why are we doing this and what are we doing it for and what can we do with this that maybe we didn’t think of. So, it’s always there’s always value, and sometimes there’s actually a publication too.

Henry Badgery: It is no secret though starting an open source project is difficult. As you said, you don’t know whether it’s going to be popular, you need to do a lot of due diligence and track a lot of things. But I think what people actually tend to forget is that managing an open source project once it’s started is even that much more difficult. So, can you please tell me how you and your team go about managing these open source communities?

Gil Yehuda: You are right. You’re 100%, right, that this stuff doesn’t come for free and it takes real management. I have a team. I’m very fortunate to have a team. These are like the most wonderful people I know. So, I’m lucky.

Henry Badgery: An amazing team.

Gil Yehuda: It really yeah. And they get it and they understand technology, and they understand people and process. The fabric of any good collaboration, the underlying fabric is trust and a sense of belonging and psychological safety, that I can contribute to something there’s this duality of the conference of publishing something out there and the humility of receiving a pull request an issue a bug, you know, being told that something wasn’t perfect.

We try to foster the kind of psychological safety and trust within the community that we have your back, we’re actually here to make software better together and to do so, like we can publish things that aren’t perfect and we can accept requests from people who are perfect, but together, we’re going to make things better because we’re looking for improvement. So, we try to foster that message. We work with the people who are asking to publish the project and really have them do the heavy lifting and saying, if you want to publish this project, it’s really upon you to create this community and make this happen. But then we become your strongest supporters.

You’ll have to run a Slack group or a Google group or whatever communication technology you want for your project and we support a lot of them because different groups, you know, tend to different modalities. Some people prefer meet-up, that sort of thing. Nowadays meet-ups are kind of harder to do physically in person, so you have the online thing, right? We tell them listen, you’re a representative of your community, you know your community better than we do, right? Because your community is like a detail, it’s like, you know, the machine learning people who deal with whatever or the, you know, we will help you will set you up with the websites you need and the social media accounts you need and with the infrastructure will also help will have your back when issues come up.

Sometimes issues come up sometimes there’s friction and fight and like uncomfortable or legal questions about how people participate and with what permissions and all that and we want our project leaders to lead the conversation but we are right there behind them to say; we’re going to support anything that really isn’t about the technology, but about that process, we got that. And then we, you know, we promote their stuff, we let them know when there’s a meet-up that they should be attending or there’s a conference, they should be submitting a call a paper for.

We track those things in advance and help them, you know, prepare their presentations, make sure that it’s what we want to share in public, and then track adoption, and then say, Hey, you know, do people listen to that podcast or blog post and what’s the reaction that we got and maybe we need to go back and, and clarify something or fix something. So, it’s a job like it’s a full-time job to make this work. But again, that’s how you get the benefit. If you just put the code out there and hope then it really isn’t. It really isn’t strategic. So, we invest in that to get out those outcomes.

Henry Badgery: Sounds like you definitely have it. A lot of it most of it probably I want to say all, but I’m sure there are things you’re still figuring out but you have what it seems like – everything figured out. I can understand though that because you were one of the first in the field and the fact that we do know that being a pioneer is difficult. You don’t have any blueprints to base anything off. You don’t have any predecessors to learn from. I can’t imagine that it was always so easy.

To benefit of those who are listening, what were some of the key challenges that you and your team experienced to get where you are today?

Gil Yehuda: You know, it’s a good question, Henry. I don’t know if I’m a pioneer. And I don’t know, like, the truth is, I learned a lot of this. I learned everything. Everything I know is based on a lot of lessons I learned from really smart people and from really good experiences with teams of people. You know, this isn’t like anyone. I don’t think people wake up and they know stuff. I think that you learn stuff the hard way and I’ve been doing this for a while, so maybe, you know, so it sounds like I have everything figured out, I don’t.

We run into challenges all the time. I have a lot of people that I’m grateful for who have advised me and much like our software gets better together, our Open Source Program Office gets better together. You know, I’m part of a group of Open Source Program Offices and other tech companies and I communicate with my peers all the time and I learned from them every day and that I think, learn from me too. I mean, it’s collaborative, because we believe in this stuff. Like we believe that collaborating together makes things better not just for code but for everything we do.

Some of the early challenges, ironically, the early challenge that I had was dealing with people like me, when I was on, you know, opposed to open source was to then just deal with other people who are like, Oh, this is frivolous, and it’s a waste of time and why would we want you know, people to publish code in their basement over the weekends for free and that’s just a liability to the company. And part of it was like, no, we’re not talking about people publishing code in their basement for free as like, that’s not what we’re talking about here.

In fact, some of that may be a liability, like you should figure out what they’re publishing and just make sure that it’s not a problem, right? I mean, we have a social media policy, if it’s not part of the business, then whatever, it’s their private stuff and if it is part of the business, well, we should know what they’re doing, because it’s part of the business, right? So maybe, like, but that’s not what we’re talking about. We’re talking about something more strategic. So, shifting the conversation from the naysayers was part of it. Ironically, when I got to Yahoo, I found myself dealing with a little bit of an opposite problem, which was folks who wanted to open source everything without really, you know, challenging well, why are we open source and what because we wrote the code and shouldn’t die here.

It shouldn’t like to be locked up here. It’s like, okay, yeah, but until then, this is going to die on GitHub. So, let’s think smarter. Like why are we doing it? Who are we doing it for? What outcome are we trying to get? So part of it was like then fighting the fight, whatever, dealing with the people who are like, let’s open source everything and see if anything is good. You know, so it’s finding that balance between the fear of doing anything and trying to figure out, you know to shift that to strategy and the halo just opens for seventh because it sounds like fun and the like, let’s refine that into strategy because after all, we do have to answer some of those naysayers and they’re going to perceive what you’re doing is exactly what they don’t want to have happened.

It was finding that kind of balance of what can we do? What should we do? Why would we do it? How do we know that we did it well and then experiment a lot? In some of the experiments, a lot of them didn’t work so well. But after you do a lot of experiments, you’ll learn. And I think after doing a lot of experiments, I have a better sense of what might work.

Eunice Chendjou: Well, it seems like you’ve had a very interesting experience with open source. But what I’m really curious about is, you know, and what are you the most excited about with regards to the future of open source. What’s next?

Gil Yehuda: Wow. Okay, what’s next and open source here? Man.

Henry Badgery: This is a good question.

The Future Of Open Source

Gil Yehuda: It’s a good question. It’s also a difficult question because if anything we’ve learned in the past four months, predicting things is hard. Okay like four months ago wouldn’t have predicted May 2020 to be the way May 2020 year’s right we wouldn’t was five months ago, six months ago, like or three years ago, I wouldn’t have predicted. So, predicting about predicting the future is very hard. I’ll admit that what I am excited about. And I’m looking forward to it.

I’m concerned, frankly, I’m concerned about like the proliferation of so many open source projects that it’s just hard to manage and hard to know that there’s like, millions of projects out there and for the next you know, For the next person who goes out there, they just don’t even know how to find what they’re looking for and they create a duplicate of something that’s already there. And the duplicate, you know, might have been better or worse, but not compatible. So, I’m kind of concerned about the chaos.

I know people are working on some sort of registry or indexing and it’s a hard problem. So, I’m concerned. I’m concerned about locking, you know, open source was designed to prevent vendor lock-in. So, we don’t have vendor lock-in. But we do have ecosystem not really locked in, but ecosystem biases. So, if you’re in an open source world, in a particular world, it’s kind of hard to get out of that world, just sort of in that sub-world. And I want to make sure that we don’t create accidental inflexibility so that you can pivot from one space to another.

So, if you’re using one cloud provider, you should be able to move to another cloud provider or if you’re using one persistence technology, you should be able to move to another one and if we lose sight of the lock-in that we were always really worried about, then we might find ourselves accidentally creating maybe not legal locking, but pragmatic locking. So, I’m concerned about that.

Henry Badgery: That idea of ecosystem bias is really, really interesting to me.

And I was wondering if you could tell me have you observed that in the space so far, you just think that it’s going to be something that’s definitely happening in the future based on what you see now?

Gil Yehuda: Well, here’s what I have seen is preventing ecosystem bias sort of upfront. So, we have a particular and I think other entities have as well. Strategically open source projects intended to prevent a lock-in situation. Okay, so we sort of see a play coming, say, oh, here’s an entity that’s publishing something, I absolutely understand why they want this to be published because it benefits them.

I’m going to publish an augmentation to that project that enables me to use it without having to use their additional services. In other words, I’m going to make sure that keeps it open, right. So, I have seen that kind of neutralization of the potential of technology locking. But I want it you know, I want to just get back because I don’t want to end on like, I’m concerned about the future I also want to. So, go back.

I am looking forward to more collaboration and more unification and you know I do have a positive view of the future where when people work together and create value, they reduce some of the barriers that maybe are naturally set up. Identify barriers, belonging barriers, who do I hang with? And who do I get to talk to who shares my passions, open source enables us to sort of expanding that to talk to people we would have never have met before, and to create value together with them in a way that we wouldn’t have been able to predict before? And I think that makes us better people and it makes the ecosystem better.

I think there’s so much more runway ahead of us in open source to do that and explore that and to get good at it, that I’m really looking forward to the future of open source because I think it’s, it’s the future of how we as you know, people on this planet can work together to solve hard problems. You know, we’ve solved easy problems and as the planet gives us harder problems to solve. I’m looking to open source as a fabric that allows us to trust to get in a stranger’s car right to trust each other to then do things that we wouldn’t have done before. So, I have a positive view of the future.

Eunice Chendjou: This is amazing. this is really, I really enjoy you know definitely looking not just at the positive side but also looking at a wide group of banks had to go wrong what’s the run this up again. What are some of the action items that people listening can take and drive the change within the company so that becomes the most friendly what’s open source software?

Gil Yehuda: Okay what can we do to drive for change in the company? I think the first need to start with themselves to challenge certain assumptions and to grow they are a mental model of what open source possibilities are, at least that’s what I did with me is like I looked in the mirror so while, I’m wrong let me figure out how to be smarter about this and learn more and just take an insight.

I think that people should recognize that the naysayers are not mistaken, they’re in the sense that they’re not in there simply missed focused and if we present open source as this hobbyist, you know opportunistic you give away because I feel good about myself sort of social signaling thing than what the naysayers are going to say; I’m sorry I just don’t see why I’m paying you for this like, sounds like a nice thing to do but I don’t understand why the company wants to rely upon this because I have to answer to somebody and I don’t know how to answer that.

I think the people need to refine that and say; yes I want to be able to interact and expose express my passions on the side as a hobby with people totally but when I come to work I also need to dress them up with a strategy and with better outcomes and with looking at what happens if I don’t open source or if I don’t work with open source communities or utilize open source strategically and what happens if I don’t participate.

It’s like what would happen to a project if I take an open source project, I fix something I don’t contribute back. What’s the end result would be with that kind of behavior, probably not good so how do we fix that let’s understand why you want open source everything and why you want open source nothing and figure out, what are you really saying so that we can figure out what to do as a company. I think when people come in more strategically they’re going to be received more strategically and they’ll be more successful.

Closing

Henry Badgery: One thing that’s been a common theme is this idea of being extremely strategic with your Open Source Program Office that was something that stuck out most from talking with you. So, Gil thank you so much for your time and for sharing these amazing insights. It’s been a fantastic episode.

Eunice Chendjou: Thank you.

Gil Yehuda: Thank you, thank you both.

Henry Badgery: To those of you who are listening, if you want more content like today, then please like the video and also subscribe on YouTube and if you’re listening to the podcast then please leave a review and let us know what you think. Safe everyone! Until next time. Thank you for listening.

Avatar image of Henry Badgery
Written by

Henry Badgery

Leave a Reply

Your email address will not be published. Required fields are marked *