Open Standards and Licensing in Database Technology with Mark Stone
blank

DDC

Reading Time: 8 minutes

How about more insights? Check out the video on this topic.

open standard licensing video

Step into the dynamic world of database technology and open source innovation as we delve into a thought-provoking discussion between two seasoned experts, Peter Farkas and Mark Stone. In this blog post, we unveil the captivating dialogue that took place during their enlightening conversation, where they generously shared their wealth of insights and experiences.

The webinar series discussed in the sources aims to connect experts and users interested in the future of document databases. The speaker, Mark Stone, who previously worked at Microsoft, shares his experiences with open source and how Microsoft’s approach evolved over time, along with the challenges they faced. He emphasizes the importance of open standards in promoting innovation and discusses the need for a document database standard that can act as a mediator between powerful organizations and the community.

Highlights:

 

  • Open Standards, Licensing, and Open Source
  • Microsoft’s transition to embracing open source
  • The OuterCurve Foundation and addressing concerns
  • The Stargate project and symbiotic competition
  • The importance of an independent document database standard for growth

Peter Farkas: 

Welcome everyone to the next episode of the Document Database Community webinar series. I’m Peter Farkas, the co-founder and CEO of FerretDB, an open-source MongoDB alternative. Our mission with these webinars is to create a vibrant community where experts, users, vendors, and innovators in the document database space can come together. The world of document databases is marked by diverse approaches and constant innovation, and we aim to foster discussion on its future.

Today, we have an exciting topic and a special guest, Mark Stone, who will shed light on open standards, licensing, and open-source experiences, particularly during his time at Microsoft. Mark is a seasoned open-source technologist with over 20 years of experience, making him a witness to the evolving landscape of open source in enterprise contexts.

Mark, the floor is yours, and we’re looking forward to your insights on Microsoft’s open-source journey and how it aligns with the document database community’s goals. Thank you for joining us today.

Mark Stone:

Thank you, Peter. I’m thrilled to be here. I’ve been involved in open source for over two decades, and today, I’ll be sharing my experiences, including my time at Microsoft.

I started in the late ’90s as an executive editor at O’Reilly, where we focused on open-source technology. I later worked at VA Linux Systems and DataStax on Cassandra-related open-source projects.

 When we look back to the early days of the PC ecosystem, a pivotal moment occurred in 1981 when IBM introduced the first IBM PC. They not only introduced new hardware but also published a technical reference manual, creating an open hardware standard. This decision led to the PC revolution, and the growth in PC sales from 1981 to 1995 was staggering, reaching nearly 50 million units annually.

hardware era

Peter Farkas: 

But why did IBM opt for open standards? The answer is quite straightforward: it’s better to lead an open standard than to own a proprietary one. IBM realized that their hardware sales were low-margin compared to high-margin services. By fostering a larger hardware market through open standards, they could sell more services, aiming to be the service leader in the PC market.

Mark Stone: 

Microsoft, on the other hand, recognized that two types of software made a PC valuable: an operating system and applications. By creating an open standard for operating systems with MS-DOS, they made application development more accessible and became the leader in PC operating systems. The lesson here is that being greedy and trying to own something entirely is not always the best strategy. Leading an ecosystem based on open standards and open source often yields more significant benefits.

Peter Farkas: 

That’s an essential insight, Mark. Open standards and open source encourage innovation and reduce market friction. By allowing innovation, imitation, and iteration, these ecosystems can thrive and expand. The collaborative nature of open source fosters continuous innovation.

 

outercurve

Mark Stone: 

Absolutely, Peter. By 2008, when I joined Microsoft, the company faced the challenge of returning to its roots in innovation, especially in the face of open source competition. Our mission in platform strategy was to lead Microsoft back to its innovative core.

Now, let’s explore how Microsoft tackled the challenge of open source by creating the OuterCurve Foundation, a non-for-profit foundation to which all relevant intellectual property could be transferred. This foundation would then release the source code as open source. The OuterCurve Foundation was established in 2011, operating independently from Microsoft.

To clarify the process, Microsoft would transfer all its intellectual property rights regarding a particular piece of software intended to be open source to the OuterCurve Foundation. This meant that Microsoft no longer held ownership of the software; the foundation did. The foundation could then exercise its intellectual property rights to release the software as open source. This approach addressed concerns from both the open source community and Microsoft.

growth curve

For the open source community, it ensured that Microsoft was not merely making something available as open source but was relinquishing its intellectual property rights. For Microsoft, it reduced the risk of downstream issues because any negative consequences would be tied to the foundation as the intellectual property owner.

Today, the OuterCurve Foundation is no longer in operation, but the work accomplished during that time was successful in fostering Microsoft’s participation in various open source projects. Microsoft’s embrace of open source has allowed it to return to the innovative dynamism of its early days.

Mark Stone: 

So, what can the document database community learn from this history? Let’s consider MongoDB as a cautionary tale. MongoDB’s success led to imitation by AWS and Azure, who offered their own MongoDB services within their cloud offerings. MongoDB faced a choice: embrace imitation as validation, growing the market and continuing to lead, or treat imitation as a competitive threat to be eliminated through legal means. We know the path MongoDB chose—altering its license to the server-side public license.

This change was justified as a response to large cloud vendors capturing value without contributing back to the community. However, it’s not clear whether this shift has truly benefited MongoDB’s business. Development without competition can lead to complacency, not innovation.

SQL open standard

Mark Stone: 

So, what should the document database community do to encourage innovation and maintain a truly open-source approach? I propose learning from the Microsoft lesson and seeking an independent document database standard that can mediate between powerful organizations and eager community participants.

An example of this approach can be seen in the world of Apache Cassandra and the Stargate open-source project. Cassandra, launched by Facebook, is used by a range of enterprises, from Netflix to FedEx, highlighting the potential for open standards and collaboration.

Stargate is an API gateway that works with Cassandra, simplifying database interactions for applications by leveraging familiar APIs. Recently we launched a JSON API, turning Cassandra into a Document Database. Rather than look at MongoDB code, we relied on two related open source projects: 

  • mQuery, as a standard for how to query a Document Database;
  • Mongoose, the popular object data mapper, as a reference implementation for how a client library can conform to mQuery. 
MongoDB alternatives

Mark Stone:

By treating mQuery as our document open standard, we have built a Document Database out of Cassandra

There are two advantages to this approach:

  • First, we don’t have to concern ourselves with the MongoDB license, since we never leverage any MongoDB code.
  • Second, we have used Cassandra in a novel way to create a new approach to Document Database design.

This innovation benefits two open source communities. On the one hand, we are expanding the options for JSON-oriented developers when it comes to databases. On the other hand, we are expanding the use cases for which Cassandra is applicable.

Starting with FerretDB
blank

Peter Farkas: 

Thank you for sharing these insights, Mark. It’s clear that embracing open standards and open source while fostering collaboration is essential for long-term innovation and success in the document database community.

Mark Stone: 

Absolutely, Peter. Collaboration and open standards are keys to sustained innovation and growth. Let’s continue exploring how these principles can shape the future of document databases.

Peter Farkas: 

Indeed, Mark. Our journey through this discussion has been enlightening, and we look forward to more insights on the path ahead for document databases.

Mark Stone: 

The essence of open source is that code once released into the world can take on a life of its own. It can be forked, modified, and adopted by different communities. While MongoDB has attempted to assert control over its technology with the SSPL license, the open-source landscape is vast and adaptable.

Additionally, the development of open standards, like the mQuery specification mentioned earlier, can play a significant role. Standardization provides a common foundation for multiple projects and can lead to interoperability and innovation across various implementations.

Ultimately, the success of MongoDB’s approach depends on how the broader open-source community responds. It’s possible that MongoDB may reevaluate its licensing strategy in the future based on community feedback and market dynamics.

So, while MongoDB’s decision is a significant development, it doesn’t necessarily dictate the final outcome for the technology or the community. Open source has a way of finding creative solutions and adapting to new challenges.

Peter Farkas: 

Mark, it’s fascinating to hear your thoughts on the evolution of document databases and open source. Ten or so years ago, there was this notion that document databases might replace relational databases. What’s your take on that?

Mark Stone: 

Well, Peter, it’s essential to consider the use cases. For financial institutions needing strong transactional integrity, relational databases are still the way to go. But there are scenarios where strong transactions aren’t critical. For instance, Facebook built Cassandra to handle massive scale, where eventual consistency was acceptable. So, it’s about finding the right technology for specific scenarios.

Peter Farkas: 

So, you’re not a believer in a one-size-fits-all database?

Mark Stone: 

Not at all. It’s unrealistic to think one database can handle every use case. We’re in a field where there’s always more to explore and innovate.

Peter Farkas: 

Shifting gears, you’ve got more than 20 years of experience in open source. Any book recommendations for someone looking to learn about open source?

Mark Stone: 

I’d recommend “Open Sources” and “Open Sources 2.0,” both published by O’Reilly. They feature thought leaders sharing insights on open source’s importance and how it works. Bruce Perens’ essay on the Debian social contract is a gem.

blank

Peter Farkas: 

Open source licenses can be confusing. What’s your preferred license?

Mark Stone: 

In a nutshell, licenses fall into two categories: permissive (like BSD, MIT, and Apache) and non-permissive (like GPL). I lean towards permissive licenses. The Apache Public License (APL) is a great choice. It’s permissive but encourages attribution.

Peter Farkas: 

We have more generic open source questions, but we’ll tackle those on our database Slack. Now, looking ahead, what do you foresee for the document database community in the next three years?

Mark Stone:

Ideally, in three years, we’ll have an agreed-upon document database open standard. And alongside that, a vibrant community with diverse approaches. The goal is not to replace all databases but to specialize and innovate in different directions.

Peter Farkas: 

Thank you, Mark, for your insights today. Your perspective on document databases and open source is invaluable.

Mark Stone: 

My pleasure, Peter. I’m excited to see where the document database community is headed. 

 

0 Comments

Related posts

Subscribe to Updates

Privacy Policy