Personal Reflection on the Open Core Summit

It's been a week since the inaugural Open Core Summit took place -- a conference focused on commercializing open source. Having attended both days of the event, and with enough time passed since, I'd like to share some personal observation, learning, and reflection.

Day 1 had a long string of impressive keynote speakers. I attended all of them except for two (needed more caffeine), and pushed myself to engage with them by tweeting out snippets from each presentation. See the whole thread here:

Positive-Sum Thinking

One narrative that emerged was "positive-sum thinking" when approaching the question of commercialization, in an age where hyper-scale public cloud vendors dominate computing. It's a dynamic that the industry has been struggling with for a few years now. The message was delivered quite eloquently by both Deborah Bryant of Red Hat and Sarah Novotny of Microsoft, and reinforced later by Andi Gutmans of AWS (perhaps unsurprisingly, all major cloud vendors). Deborah had a humorous but memorable line, comparing the phrase "commercial open source software" to "salsa sauce" -- it's redundant! Open source software has always been "commercial". Marten Mickos, the former CEO of MySQL, provided an insightful framing as well: cloud may have killed some business models and rendered open source's distribution advantage obsolete, but open source itself is alive and well, thus the struggle is searching for new/improved business models around open source.

In the "positive-sum thinking" narrative, cloud vendors and commercial open source companies can co-exist, establish mutually beneficial relationships (echoed by Jay Kreps, CEO of Confluent), and grow the pie for everyone without challenging or changing the fundamental meaning of "open source".

Not everyone saw things this way, of course. Salil Deshpande of Bain Capital Ventures shared his view that a new standard or meaning of "open source" is needed, the Open Source Initiative does not own that "meaning", and the licensing proliferation that has been happening in the industry will continue if a standard isn't established. A spirited panel between the co-founders of four open source database companies: Timescale, InfluxData, Neo4j, and Yugabyte, further underscored the reality that different licensing strategies and debate around them will continue to pop up.

To me, the positive-sum vs non-positive-sum debate bears some resemblance to the policy debate about free trade in the U.S., which has resurfaced recently as trade barriers are being erected. At its core, the free trade debate is a struggle to balance a mismatch in human impact between "diffused beneficiaries" (consumers buying cheaper cars, electronics, clothing) and "concentrated disruptees" (steel producers, automakers, and their unionized workers). In the open source case, the beneficiaries are individual developers, academia, new startups, and clouds, while the disruptees are specific commercial open source vendors.

As a proponent of free trade (and, I suppose, a positive-sum-er), I tend to veer towards growing the pie rather than looking for my piece to take. Nevertheless, as a student and relative newcomer to this space, I appreciate the Open Core Summit as a neutral venue where I can hear, absorb, and process all points of view.

Project != Product

Besides talking about positive-sum thinking, Sarah Novotny of Microsoft made another important distinction between open source project and product, which I summarize as:

The project is for your community.

The product is for your customers.

Project != Product.

Community != Customers.

This distinction was echoed by Glenn Solomon of GGV Capital on a panel, when referencing something Aghi Marietti (cofounder/CEO of Kong) previously said: don't confuse your open source users with your open source community.

I've seen this distinction (and conflation of the two) play out both in my own company building experience and when observing others. It's a common mistake. A prerequisite of any company looking to commercialize an open source technology is that open source project be widely adopted in production. However, that doesn't mean you already have a product to sell; the classic necessary but not sufficient problem. A well-adopted project is the base from which you build products. The project itself is not a product.

I'm glad Sarah elevated this point during her talk, especially for all the founders in the room.

RISC-V

While open source is typically associated with software development, RISC-V, an open source instruction set architecture (ISA) poised to disrupt hardware, was also present on the keynote stage. This was rare and welcoming. I usually don't see hardware topics covered at open source conferences. (RISC-V has its own foundation and conference.)

Listening to both Navin Chaddha of Mayfield and Naveed Sherwani, CEO of SiFive, talk about the impact that RISC-V is already having on semiconductor was eye-opening. According to Naveed, a design to reduce power usage by 90% could be produced in a matter of months, because RISC-V is open source thus can facilitate more people to collaborate on the problem.

SiFive's story is also an example of "positive-sum thinking", where large traditional semiconductor manufacturers -- Intel, Western Digital, Qualcomm, etc. -- are all investors, and key stakeholders in the RISC-V ecosystem at large.

If you ask me what single topic or trend I gathered from the conference that's worth watching for years to come, RISC-V is it. RISC-V, or open sourcing hardware in general, is profoundly disruptive. It will take some time to unfold though and is an open question of how and who.

What Makes a Good Community Event?

This was the first community event I attended that explicitly focused on the commercialization aspects of open source. Having been to my fair share of other community events -- from a ~20 people meetup in Portland to ~8000 people mega-conferences like KubeCon North America -- intimacy trumps size for me.

Why is intimacy good? At least as an operator/entrepreneur, I've always found more intimate gatherings to yield more direct and actionable insights -- things I can apply tomorrow or this quarter. There's less need to "play to a crowd". You can have an exchange of ideas and experiences without the pomp and circumstance. A community's value has more to do with the quality of interactions than raw number of participants.

This first Open Core Summit could use more intimacy and less size. After listening to (almost) all the keynote speakers on Day 1, which was a 12-hour marathon in a huge hall, I was flat out exhausted. Even though I returned for Day 2 (not sure if many people did), I was too tired to engage fully in the breakout sessions, which could've been more intimate if the venue was smaller. Thankfully all talks were recorded, and I'll watch them all.

The thirst for information on building commercial open source companies is real. And as Joseph Jacks and Bruce Perens announced at the very beginning of the conference, the Open Core Summit concept is meant to drive the Commercial Open Source Software (COSS) Community. Since "community" is the focus, I look forward to future OCS's be more intimate (and require less caffeine).

###