A Blank Check for Open Source

Today we announced a partnership with Drone.io to support all of the infrastructure for Drone Cloud - a hosted CI/CD service that provides free multi-architecture builds for open source projects.  

We’re excited by this project for a few reasons:

• It’s Drone!  We love Drone and use it ourselves.
• It solves a notable problem: building for architectures like 32-bit Arm, and Windows
• It’s designed in response to the open source community.

An Open Source Support Gut Check

There have been some good articles in the news recently (and plenty of talk over the last year or two), highlighting the tension that has arisen alongside the massive success of public clouds.  Much like other huge modern businesses, the public clouds are big consumers of open source, and there is a lot of hand wringing about the equity of the arrangement.  More importantly, open source companies and projects are wrestling with their future viability in the face of these trends.

I’m actually very optimistic about the long term viability of the open source movement (more on that in another post), but the short term view is murky.  Much like the internet itself, it’s easy to treat open source as a constant that will always be around - like the air around us, it’s hard to imagine it not being there when we need it.  

We know that is not true. Like an ecosystem, it needs to be tended and cared for. So, our question at Packet has always been: what does open source need, and what can we provide?

Always Bet on (Open) Software

When we started Packet, one of the main tenets of “why” the cloud needed a new provider, was that we saw this friction: public clouds pushing users further and further up the stack (VM’s, managed databases, serverless functions, translation as a service, etc) and software zooming down the other way (Docker, CoreOS, Kubernetes and a zillion others).  

We saw that tension as a natural order of things. We saw the main benefits of “abstraction as a service” as geared towards an I.T. buyer mentality. The developer buyer mentality, on the other hand, was embodied by open source, which was anxious to inject its opinion in ALL parts of the stack.  There was just one catch: it it had to be automated.

So what would software (and especially open software) creators need? You can see the answer in our product today: infrastructure. Raw, un-abstracted, diverse infrastructure.

In short, we bet our entire startup on the ideals and energy of open software. We bet on the GitHub generation being a bunch of opinionated developers. I hope we’re right!  ;)

The Blank Check Concept

We’ve been supporting open source in the way we know how (mainly free infrastructure) since we were able to get servers booting on our platform. Some of the most exciting examples for us have involved the CNCF Community Infrastructure Lab as well as individual donations to projects like Kernel.org, Wireguard and NixOS.

When a project approaches us, we try to say “yes”. A few build servers? Sure. Four global points of presence for making downloads faster for end users?  You got it. Access to a brand new Intel Scalable, or an Arm server for testing and optimization. We’re on it.

We never thought about it as a blank check, but looking back, we’ve never said no.  In fact, we’ve often pushed projects (who are run by volunteers and so naturally quite conscientious) to ask for more, knowing that they might need it.  Crazy!

Writing That Check to Drone

And so when Brad Rydzewski (founder of Drone.io) and Melissa Smolensky (a joint friend) reached out this Fall with the “crazy” idea of releasing Drone Cloud for free to the open source community, we immediately said yes.  

We were really excited about Brad’s reasons for rolling out Drone Cloud. When he spoke about all of the Drone (and other) users who are building for an exploding set of architectures and operating systems. Pairing Intel, AMD, Arm and (hopefully soon) other bare metal systems with new features in Drone meant that our two small companies could make a real impact on a lot of projects.

To be honest, it’s pretty cheap for us to support Drone Cloud at launch.  The real commitment comes with success. What if it grows? Like, a lot? We hope it does!

That’s why my answer to Brad: you have a blank check for the next two years. Go do the right, righteous, smart and timely thing.  Serve the community and we’ll back you up.