Impact of Fragmentation on the Open Source Community – Hilary Carter, The Linux Foundation

This is Textron TV. Hey guys. Thanks for the throw.

We're here with Hilary Carter who's senior vice president for research and Communications for the Linux foundation. And we're talking about a new report. They put together looking at fragmentation in the open source Community Hillary welcome to the show.

It's hard to imagine that there's any issues with fragmentation in it anywhere, but be that as it may what's going on in the open source world and why is this becoming an issue? Well, that was what we were setting about to uncover was is it an issue in the first place? I mean there are many people in the open source community that actually don't believe that we have any issues fragmentation among them.

And when we reached out for comment the sad I don't really think it's a problem others in the community. However, see fragmentation is a real thing and I'll break that down. And we're talking about geographic fragmentation or a duplication of Technology projects or any kind of inefficiency talent recruitment language.

And how do we improve the ecosystem if there is fragmentation? If so in which ways and what do we do about it? How do we make open source more of a truly collaborative Global phenomenon and a lot less compartmentalized or siled or fragmented?

It seems to me at least that some of these issues just come up with somebody takes a piece of Open Source Code and then they extend it in some way and it becomes something of a fork or a branch off of the main project and then we get a bunch of these things going on and then nobody ever seems to come back and say can we just roll all this up into one project and kind of make this easier for everybody. So are we the victims of our own success here in some ways? Yeah very much.

So, I mean working and duplication are a natural outcome of innovation and people wanting to do different things that are almost identical but not quite and we're not here to pick the winners, but rather let the Natural Evolution of a project and it's Forks evolve and see where the you know, which may have the greatest impact what we can do. Help create efficiencies where they make sense and try to avoid some of the needless duplication or help align resources behind some of the more prolific more mature projects. Generally what we found was that there's a lot of overlap a lot of fragmentation at the outset of a project or new technology area.

A blockchain is a really good example of that. So And we have a highly fragmented blockchain development community. That doesn't interoperate terribly.

Well contrast that with the project like the Linux kernel, which is incredibly mature prolific and highly ubiquitous So that's a pattern that we see is just being natural to the way open source develops. Course the Linux folks have been at this for a while. So does some of this fragmentation equal to the maturity of the project that the maturity of the technology Yeah, absolutely.

As I was saying before we find it most acute when when a technology is initially open source, and we might may find examples of it in different communities and slightly different forms. But over time the longest standing open source technology projects tend to consolidate or you know, no one wants to compete with the winners essentially. So there's no point in Reinventing a kubernetes or making it slightly different or trying to create a new Linux kernel create something new that doesn't exist.

And then you know, we'll see we'll we'll see where we are in five years time. But yes Linux foundations been at this a while. And so these These are the patterns that we have seen over the course of the Linux Foundation from our namesake project to some of the new technologies that we're that we're hosting today.

To open source projects benefit from something you might consider a benevoline oligarchy of you know leadership at the top that kind of looks at all these issues and says, yeah, let's not reinvent this wheel. We already have it over here and there's a little more centralized approach to managing things. Yeah, I mean there's certainly efficiencies we're very mindful of the fact that resources can only go so far and so there is there's a need to attract resources attract funding attract talent and you know position your projects as being the most viable ones to work on.

And where we can where we can. You know make some gains is in those projects where they're they're still questions out there about which are the the winning Technologies or where? Which of the projects that are most likely to become the Linux kernel of tomorrow and help organizations and vendors and and end users pick their project and Technologies more efficiently align their resources more efficiently get behind certain projects that are likely to have the most impact and if we can not compete with each other and that's a really good thing.

Everyone's hosting a winner of its own kind then that's a we think that's a good a good thing. How do we balance Innovation against efficiency? Because some folks will say if you Throw all your weight behind one project then over time.

You're not really going to see the opportunity to kind of totally disrupt that project with something new and different because everybody's going to be so focused on contributing to the existing project versus looking for something that might totally eliminate that need or create some entirely different approach that no one had thought about but maybe we put our blinders on and won't see if we're all focused on one project. Maybe the joy of of the projects that proliferate is that they have been open and receptive to to changes or innovations that the community needs and wants and so they're they're, you know proposed as pull requests and the project becomes stronger and more robust as a consequence of innovation itself. But on its own Innovation is generally responding to a need so it's either because a large project isn't meeting a specific need or because that project doesn't actually exist and they'll propose it out.

Right and they may find that somebody on the other side of the world has proposed that same project out right and You know there that's how you get that that fragmentation is that two people are solving very similar problems. You see that in fintech or with? digital credentials right now after covid a lot of innovation came forward that was trying to digitize credentials vaccine credentials and other types of medical records.

And there was all of this duplication coming at once because everybody wanted to be first with that. So again, it speaks to the life cycle of a projects probably not avoidable, but ultimately in an environment where there are constrained resources. You're going to look at the technology that has the greatest potential.

For implementation and and deployment and get behind that project. What's your sense of areas where there's a lot of duplication of effort. It seems to me perhaps in the realm of cybersecurity.

There's a lot of things that people are working on in different projects that they could share more readily because they have a common set of issues or as each project. So unique that maybe we don't get see that level of cross project collaboration that we might I think in the security around Mike, we you can never have enough alignment and conversation and I think that's what coordination is all about. And I think that's the outcome of This research is let's identify those cases where there's there's wasteful duplication on an issue is important as security and get behind a really good approach and apply standards and best practices that we can agree on that will have impact and therefore creating Seasons across the board.

and You know. It would be great if we could coordinate. across foundations more than we're doing to to work on issues as important as security anything in the report that left out at you that surprised you based on the feedback you got from folks.

I think the it wasn't a surprise but the themes are still very prevalent about diversity and inclusion and the need to radically localize. Communications are tooling so that we can attract more talent in different parts of the world. my colleague from Mexico Ramon Roche was talking about you know translation of of chats as a means to bring in new developers into the Drone code project that he's behind and we take for granted the fact that English is the Language of Open Source but on its own that is such a large barrier and we need to move beyond the idea that if it's not English, then it's not at all and create more opportunities for dissemination of information and documentation that is in multiple languages to really realize the benefits that come from many eyes and and many contributors.

So the language thing was on one hand not a surprise, but the extent to which it's such a barrier is a really good reminder. Her community is more inclusive across the board more diverse is one of those barriers that persists and we don't necessarily have all the answers on how we overcome that. I think it's a multi-generational multifaceted problem to solve but it warrants a dialogue and I think that's what's exciting about a report like this is that it creates opportunities to have conversations same with public policy.

You know. How do we come together as a community of different locations and and different Focus areas and solve some of the problems around creating regulatory Frameworks that are supportive of Open Source broadly and create responsible public policy around topics, like data privacy and security, you know having a piece of research be the focal point for dialogue is a very positive thing. That's what I'm hoping will come of this this research.

One of the things you hear a lot about from folks is they don't know how to engage with a certain open source projects particularly some of the smaller ones because it's hard to find that folks who are in charge of that and it's hard to get their attention and a lot of times people wind up Reinventing the wheel because you know that code or exists somewhere else that somebody saw for some proprietary thing that they're happy to share. It's just that we haven't really put you know, this one meet this one and get to know each other. So any thoughts about how to do that better how to create that community.

We are trying hard to provide those opportunities for introductions. And that's the reason why we host in person events all around the world and provide scholarships travel scholarships and other types of funding. To bring people into our community and to create points of access so that those questions can be answered those introductions can be made and I think meeting as a community at events at meetups is critically important.

So I would encourage people to go online and search out where it is that a project Community meets to have these kinds of conversations. Is it is it a Meetup? Is it a formal event?

Is there a travel funding? What are the what are the social media channels and chat rooms as a Discord. Is it WhatsApp, you know, there are lots of opportunities digitally to to you know, find the answers and and access the folks that you want.

You really want to talk to that. I'm encouraged that we have the the right tools and I think anybody who's motivated will find their way using one one platform or another. All right folks.

Well as cool as all these virtual platforms are there's still no substitute for a little face-to-face time and get some things done in ways that you probably couldn't do before in a way that has a lot more empathy for the other person on the other side of that conversation Hillary. Thanks for being on the show. Thank you, Mike.

All right back to you guys in the studio.