Insights from our OpenSource roundtable v.2
DataSeries < > Jeremy Garcia|VRT220420
In July, DataSeries, an OpenOcean led initiative hosted their second Virtual Roundtableabout the “Future of OpenSource”. The session was moderated by Jeremy Garcia, Founder of LinuxQuestions.org and Director of Technical Community/ Open Source at Datadog.
Have a look at the summary insights from our first OpenSource roundtable here.
INSIGHTS GATHERED:
INITIAL NOTES:
We need to rethink how we are engineering today’s software. Software is still oftentimes built with the philosophy that it needs to be owned & maintained by the creator. Modern approaches via GitHub with various amounts of contributors prove to set a new path (monolith software).
There is a fundamental assumption that OSS is more secure, but we must still be cautious as the landscape has changed dramatically during the last 10 years. Now much more is embedded in larger pieces of software and much of it cannot be vetted very well — mature projects are strong but are maintained less (e.g. heartbleed). Recent Twitter hack shows force of attacks.
There is no lack of money for widely deployed OSS. Instead, there is a representation issue. Many OSS communities are not that inviting although many contributors are keen to join. This proves the popularity of projects such as Tidelift & SQLite.
Multi-cloud adoption means a single vendor can’t become too big and too powerful — like when Oracle dominated DBs. MongoDB is heading to be like Oracle. Kubernetes provides an OS approach, but it is not as mature as the cloud vendors’ own offering (it is catching up). Cloud vendor threat is real and complex.
- Cloud Management Platforms have not been very financially successful vs the LaaS players
- If you are losing to the major vendors then maybe you have the wrong business model
OpenSource companies pioneered the distributed working fashion. The current pandemic’s silver lining is that we are all working distributed and it enhances the fact that this work ethic is inevitable. A key challenge that lies ahead is to turn this into normality and design a new working approach that will prove to be effective.
OPEN SOURCE BUSINESS MODELS:
- Automattic is operating in a 8–10bn market and aims to make 4–5% of that. So ratio of core company to market should be 19:1 or 20:1 -> $1 in Windows95 if there’s $20 in the ecosystem (a rule for platforms)
- Dual licensing can be an easy route
- Revenue models can be complimentary
- Don’t charge for something that is as easy as a plugin. Open Source companies put all their weight behind their core project and use it to launch their product. The reality is that the product needs to be fundamentally differentiated — so that it can’t be just incremental on the OSS project
- The aim is to create an ecosystem with many tools that bring the bigger vision to life (e.g. Red Hat — their OS is just a combination of products)
- Not a platform if there is not a $1bn ecosystem on top
OTHER KEY INSIGHTS:
- The biggest challenge is “short-termism”. Always think about the long term, 10–20 years ahead rather than 5 years
- It is very clear that OSS methods are not the de facto standard for writing software
- Netflix used OS to become successful and are now creating their own OS project to give back to the community (circle of life)
- If you are building a business from an OSS project, then give back to the community (doesn’t have to be money, can also be time). WordPress is giving 5% back.
- Everything that is old is now new again (dual licensing came about in the 80s). We can look back to find inspiration
- Is OSS is what technology wants?
List of participants:
Moderator:
Jeremy Garcia, Director, Technical Community & OpenSource at Datadog
Startups:
- Thomas S Hatch, CTO at Salt Stack
- Peter Zaitsev, CEO at Percona
- Aingaran Pillai, CEO at Zaizi
Corporates:
- Matt Mullenweg, CEO at Automattic (Co-Founder of WordPress, etc.)
Investors:
Ekaterina Almasque, General Partner at OpenOcean
Mike Reiner, Venture Partner at OpenOcean
