Here's how I evaluate whether I'm going to use a given bit of OSS:

- Is the project important to me or can I replace it? If the latter, I'm more likely to allow failures of other criteria. If not, I need to be more strict. Bun is easy enough to replace if something were to happen to the project. Easy come, easy go.

- Are there any red flags in financing that could become problematic? Many VC funded OSS companies fail this test for me. What happens when they don't make it? What happens post IPO if they do? What happens when they get acquihired? Mostly that's up to share holders, not developers. Most VC funded companies actually don't make it and that's normal in the VC world. A few companies make it, everything else fails quickly. And there are a few examples of projects that have changed licenses under pressure of shareholders. That's why this is a red flag to me. I've used Redis and Elasticsearch, for example. And I switched away from Mongo before they changed the license. I used Terraform before they open sourced. All negative examples here.

Bun initially wasn't great on this. But the Anthropic acquisition has improved things a bit. It's still a risk. But it's unlikely they have any plans for Bun other than just keeping it alive by employing the main people working on it. Anthropic itself might still fail of course.

- Has the project been around for a long time. If so, it likely has a stable community and funding. There are no guarantees but the older the better. Bun is pretty newish still.

- Is the project stable and under active development? If it's stable because nobody makes changes anymore that's usually not a great sign. If it is stable despite a lot of active development, that's really positive. It means somebody competent is in charge. Bun seems pretty good on that front.

- Is the project otherwise structured right to be future proof. For me future proof is a combination of contributor community, commercial activity, and licensing. The more diverse the contributor community the better. If there are multiple companies sponsoring and making money of a project, that makes it less likely that a single one can hijack it for their own good. This is more common with permissively licensed software (but there are exceptions). Bun doesn't have much commercial activity around it and the regular contributor community is tiny. One person seems to be doing most of the work with only a handful of notable other contributors that are probably all Anthropic employed at this point. Out of these, the dependence on a single person looks the most problematic to me.

So, the overall score for bun is not perfect (there a few potential red flags) but I'm happy to risk using it because it's not that critical to me and easily replaceable.

My read of the whole Anthropic acquihire is that it is an improvement over the starting point which was a VC funded company that was probably going to fail otherwise. Otherwise, good tech and generally nice to use. I could see Anthropic going bad and this project surviving in one form or another. So, that doesn't have to be a show stopper.