Skip to content

Instantly share code, notes, and snippets.

@schalkventer
Last active October 25, 2025 10:44
Show Gist options
  • Select an option

  • Save schalkventer/dba069d2e13926382743f017db6474d1 to your computer and use it in GitHub Desktop.

Select an option

Save schalkventer/dba069d2e13926382743f017db6474d1 to your computer and use it in GitHub Desktop.

⛅ Cloud Codependence

Rediscovering the Simplicity of Local-First Software

On 20 October 2025, Amazon Web Services (AWS) suffered a massive international outage. As to be expected, this disrupted a range of products built on the AWS cloud infrastructure, including Amazon itself. Within a few seconds a large portion of the internet imploded: from Disney+, Roblox, OpenAI and (even on local shores) Standard Bank.

For most of these products a cloud-based service, able to broker communication between devices scattered across vast ranges of fibre optic and copper cables is simply the cost of doing business.

Yet, during all this ensuing chaos, something else happened:

A smaller, yet equally significant category of products (those with no obvious need for an unbroken internet connection) also came to a complete standstill. Users swarmed social media with examples of corrupted data, or even real-world damages caused by this loss of control/access to their own data/devices. The most egregious example came from a company that saw several of their $2000 flagship products overheat due the electric heating control requiring an active AWS connection.

The twist: The company sells beds (Eight Sleep).

For many of us in the tech industry, this is an odd regression. We grew up in an era of floppy disks, CD-ROMs, portable drives, and flash storage, where we had complete ownership of our data. Furthermore, most computation on your data would be done directly on devices you physically owned. The internet, when used, was a bridge to someone else’s data, not a requirement to access or control our own.

Over the last decade, the industry rapidly moved away from this model. The convenience of cloud-based infrastructure and the allure of subscription-based services has largely shifted both computation and storage of data we previously owned directly to third-party remote locations. While this new paradigm unlocked immense opportunity for products that require large amounts of concurrent and scaleable network traffic, the pendulum might have swung too far.

Fortunately, due to a range of new developments within the world of Software, not only are we able to rediscover the simplicity of this age old approach of the cloud as optional background syncing layer, but we are witnesing the opening on brand new possibilities where you can combine the best of both paradigms in interesting new ways.

We will explore the modern state of the local-first software; examples of local-first products in the wild; common discovered patterns/methodologies, as well as an increasing range of tools purpose build to assist developers in a local-first era (even some conceived on South Africans shores).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment