I saw the fruit of this effort a few years later, shortly after the publication of the first local manifesto by Hacker News. I met Peter van Hardenberg, one of Kleppmann’s co-authors, in a San Francisco cafe. Like Kleppmann, he was rebooting after a long journey in the cloud, first as part of the founding team at Heroku, which helped other startups launch their cloud services, and then as part of its acquirer, Salesforce. He wanted to show me an app called Pushpin, designed like a digital cork board.
Van Hardenberg pulled out a blank draft on his iPad. I loaded a replica of the same file onto my laptop. We started tinkering, adding images and text boxes to our own files, then allowed them to be merged. Sometimes it worked perfectly; other times, changes stopped loading or pixels moved with latency comparable to dial-up. Pushpin looked like a toy, the kind of app that a few bright-eyed Stanford students could code in the common room with visions of a seed round and then put away in embarrassment.
But van Hardenberg was far from embarrassed. According to him, the technical foundations were being laid for the first local versions of Slack, Discord, Google Docs and Photoshop. Better design applications, schedules and budgets. More complex programs too, if they could make Automerge much more efficient. There was the possibility of private end-to-end encryption for all these collaborative applications, since no servers would get in the way. CRDTs had technical limitations and many applications that the cloud would serve much better. But for him, the prototype looked like a revolution. There was no server between us. Yet it worked. Above all. We were two peers communicating, as the early Internet Masons intended.
Van Hardenberg’s vision was a little easier to see when we met again in St. Louis. The tech giants were in decline. Meta’s stock was at its lowest level in seven years. Twitter was in the middle of a hostile takeover from Elon Musk. Kleppmann spent a few hours each week as a technical advisor at Bluesky, spawned by Twitter as a decentralized experiment and now suddenly thrust into the spotlight, on the verge of becoming its competitor. Its “federated” design promised to give users the ability to leave servers and services that treated them poorly. Bluesky didn’t use CRDT, which would be far too slow to coordinate the feeds of millions of social media users, but the goal was similar: a better relationship with “someone else’s computer.” IT alternatives were in fashion again.
Among them, the CRDTs. Strange Loop was full of local presentations – a surprise for Kleppmann and van Hardenberg, who until recently tracked each project through Google alerts and word of mouth. CRDTs were also appearing around the world. Developers at THE Washington Post used them to create a tool to organize articles on the home page. People digging into the code that runs Apple’s Notes app have noticed CRDTs. Jupyter Notebooks, a popular data science app, restored its collaboration tools using CRDT after Google got rid of the cloud service it previously depended on.
Among the Strange Loop presenters was a Canadian developer named Brooklyn Zelenka, co-founder of a company called Fission. When she first read the local manifesto, she remembers: “I said to myself: that’s a great sentence. Before that, we had these awkward phrases, like “location independence” or “user-owned data.” » Zelenka had been interested in the ideas of Web3 – the nickname adopted by “decentralized” applications that use blockchain technology and cryptocurrency – but found its culture to be “aggressive”, which she attributes to the emphasis about money “so clearly, all the time”. It was good to get into the local business early. “Everything is at hand right now,” Zelenka told me.
