On a Tuesday not long ago, I was having my usual afternoon espresso and scrolling through my twitter feed without many expectations when I stumbled upon this incredible photographic project: Faces of Open Source. đź If it isnât at the National Portrait Gallery yet, well it should be! It has a simple yet engaging premise to capture the development of an entire field through its people. Peter Adams, the artist, writes that his âown journey with open source have fueled this photographic project over the last three yearsâ.
Open sourcing tools is pretty tough work. The idealist in many of us might think that the community is the answer to maintaing our research tool, but itâs never that easy đ§. Making code public requires thinking, planning, iterating, and continuous development and discovery for all aspects: the code itself, the community, governance, and yes even financials.
I wrote briefly about a concept called âMinimum Viable Governanceâ a couple of months ago. And most recently I was thrilled to join the CZI Essential Open Source for Science meeting and hear about many other challenges and great developments in open source for research.
đThe most arduous task is by far the growing and managing of open source communities (see this great keynote from 2015). Itâs not only hard to engage users and contributors, itâs also difficult to know whether youâre doing a good job. The nature of open source means the contributors donât have visibility on who is using their tool, how often, what features are most used, which ones are being redeveloped independently, whatâs the user flow or what specific use cases are common or less common, who is a beginner and who is using the tool prolifically, who is about to become a contributor, etc etc. CHAOSS is working really hard to improve and standardize these types metrics for the open source community.
đ°Funding is fairly important. Often, projects that develop open source tools for research focus on solving the problems at hand, and are struggling to continue post-grant funding phase. These are exactly the type of projects that CZI has been funding. Over just 4 cycles, the organization has been supporting more than 100 proposals with $23M; paving the way for other funders in this space. NSF is also investing in projects that are looking to 'transition to sustainability', especially infrastructure which would be supported by open source software.
âźď¸Back in October, a Finnish startup called Aiven reached a valuation of $2 BILLION, after raising another $60 million in funding (in addition to the $100M that it got in March this year). Thatâs at least three times (or 8x if you count total funding) the amount that went to support the hundred open source tools for research. The startup is building a service that manages open source technologies in the cloud; it currently supports Kafka, Cassandra, OpenSearch and Grafana.
While this is pretty cool, itâs not unique; commercializing open source via open core or other models is definitely on the rise. There is even a public spreadsheet with all the Y Combinator companies that build upon a free and open source core technology. Across the 68 startups on this list, they already raised more than $1 Billion. No joke! And if you want to hear something that could blow your mind: the estimated revenue of the top 50 commercial open source software companies (aka COSS companies) is more than $23 Billion đ¤, many of them acquired and/or gone public for an aggregate price tag of $90 Billionđ¤Ż. This index is tracked by OSS Capital.
OSS Capital is the world's first and only early-stage COSS company investor and platform. COSS companies are defined as "Any company that would not exist without the parallel co-existence of a given Open-Source technology"
đđťââď¸If you are thinking about commercializing your open source software, or are just curious, some of the OSS Capital gang started the COSS community with a weekly series of posts that are approaching this topic from a product and investor perspective. This one, for example, very clearly illustrates the difference between an open source âprojectâ and âproductâ, including the non-technical aspects and what it means to sell the âpackaged experienceâ.
In other news:
A new accelerator and/or venture fund SciFounders is giving more power to the scientist startup founder. This reminds me of what Andreessen Horowitz did for tech founders more than 10 years ago now.
Since we are talking about open source, Neo4J raised $66 Million in series F (that means it got more funding before, so a total of $582M).
This one didnât raise any funding, but itâs just brilliant, I can see librarians getting excited about it: Filestar converts, merges, transforms any file formats đ¤Š
Benchling develops software tools for researchers in academia and pharma companies (so not social science!); it raised $100M and itâs valued at more than $6B, not surprising though.
Have a most wonderful week, and be sure to check out Peter Adamsâ Faces of Open Source, the art and the tremendous achievement captured here.