Hacker News

With the NATS fiasco, Synadia never achieved any of their stated goals, and ended up in a worse place than before. Why would they do this?

Synadia ended up relinquishing their trademarks (which they had already promised to transfer), they gave up taking the project away from CNCF, they gave up their plan to relicense under a BSL license, they alienated a lot of customers, and they got a lot of bad press.

The whole thing seemed to lack foresight and planning. The strategy — to blame CNCF for NATS not thriving as a community — was contradictory, since their "rescue plan" involved making it closed source. But I suspect the real reason for failure was that there wasn't really any hope to get the trademarks back; that would have required a legal battle they couldn't afford to pay for.

I'm a big fan of NATS as a technology, and I like Synadia as a company. They're doing something a little different than the DataBricks of the world. I completely understand their need to survive financially, and I understand if this pressure has increased after getting VC funding. (They raised $25m a year ago, so I'm guessing the screws started tightening.)

However, many other companies have been able to make billions without resorting to a BSL. Unlike Elasticsearch (which also created similar drama and changed their license, then went back to open source after a short while), nobody is threatening Synadia by offering a competitor to their cloud version of NATS.

I can only speculate, but I suspect Synadia's existential crisis is self-inflicted. If you look at their product offering, there's simply little reason for any company to pay for a support contract; they made NATS incredibly easy to self-host, and what you get (support and a nice dashboard UI) doesn't stand out as particularly valuable.

As someone who used a lot of open source, I'm a little conflicted by this. I want Synadia to succeed as a company. But I think it comes down to this: Either you fully invest in open source and manage to build your business around it, even if that means the business will be a little smaller and never turn into the next Microsoft. Or you don't make any pretenses, and you double down on BSLs and commercial licensing.

There isn't a middle ground where you pretend you possess the spirit of open source (which Synadia's communications kept professing) but must also turn to a license hostile to it.

> I can only speculate, but I suspect Synadia's existential crisis is self-inflicted. If you look at their product offering, there's simply little reason for any company to pay for a support contract; they made NATS incredibly easy to self-host, and what you get (support and a nice dashboard UI) doesn't stand out as particularly valuable.

This is just a really sad incentive in place nowadays. If you make something that’s easy to self-host, and open source it, you might end up with a popular project that’s just not financially sustainable. So you see people either overengineering their architecture to the point it’s painful to host, reaching out for a more restrictive licensing, or both (looking at you, Sentry... not judging though).

Of course, there’s the open core model as well. There are also mischiefs that pretend their product is open core and enable some features behind a license key, but everything is open source in the repo. (A YC startup, nonetheless! Not calling names because I don’t want to jinx it, but if it was intentional – you have my deepest respect. And I promise I’ll pay you plenty, once I have some money to sustain myself.)

So... there might be a spirit of open source somewhere in these “commercial open source startups”, but it’s usually in those that don’t play pretend open source. And in there, money still comes first, of course.