Open core can work, but you really have to find very strong product market fit on the proprietary side--ideally with features that discriminate between users who are relatively happy to pay and users who are not. (There's a reason "SSO tax" is so common.)

And you really have to believe in open source and have the discipline to keep investing in it, otherwise the temptation is ever present to throw more and more effort and resources into the proprietary parts.

Yeah, I mean Deno’s success is predicated on enterprises moving production apps from NodeJS to Deno. Node is extremely established and entrenched, and migrating the goddamn runtime of a large production server is not usually an easy project. If I have a 5-10 year old Node project, it might work well on Deno, but almost no one has the time to champion a migration when it just doesn’t have that many benefits.

Yes, it comes with batteries included, but a big node project already has setups handling things like testing, linting, formatting, and dependencies. Moving to Deno for any of those might actually be easy, but migrations in the JS ecosystem never end up being easy, so people who could sway the company to change tools don’t have the appetite to tell leadership about migration projects with minimal upside and unknown duration. And under a startup with an unknown future.

NodeJS succeeded at undermining existing server toolchains, because web devs were easily sold on writing JS for their servers, so tons of successful startups built with Node, and when Node got pretty well established, it was easier to adopt for greenfield projects in the enterprise.

Deno is Node, but better. So it’s not giving a whole market of devs access to a tool that is way easier to write for. It’s marginally easier to manage and you could maybe drop some other toolchain dependencies. But again, those toolchain things are automated/hidden away from developers directly… like they don’t care we use eslint, they just care CI catches problems before they hit prod and that the linter throws an error early in the process. It’s already easy for them to run locally. So it’s not like Deno lint changes anything about the dev user experience, it just changes what DevOps/platform teams have to manage.

People want open source software, but they do not want the compromises that come with funding it. When people try and fail then you get shitty blog posts like this one. It's sheer entitlement. I think the days of building open source tooling and expecting to be able to commercialise it are now completely gone.

"Become integral enough to the toolchain at OpenAI or Anthropic that they buy you" seems like the new one. Normally I dislike startups built with the intention to be acquihired instead of being a sustainable business, but with open source devtools maybe that's not the worst thing. I'm pretty confident that neither bun nor uv will stop existing anytime soon, and the makers got paydays out of it.

> Support maybe?

LLMs seem likely to kill or at least vastly weaken the support model.