SebastianVerified account

I think the conversation is valuable to have, so I'm reposting with additional context and explanation. I also want to preface this that I'm not making absolute statements nor making any value judgements or criticising anyone.

The beginning original tweet: I want to release Rome with no prior announcement or early access

(What is Rome? A project I'm working on. See https://twitter.com/sebmck/status/1108407803545214977 … for context)

Why? There's a few reasons. The most important to me mentally, is controlling the hype. I enjoy publicly tweeting about what I'm working on but I'm implicitly setting expectations very high.

Release dates are stressful. They can sometimes result in projects being released when they aren't ready. Babel 6 was a good example. I released it at a conference. It wasn't bad, but I wish I had more time for polish.

With something like Rome, that wants to do so much, deliberately manufacturing hype feels disingenuous to the existing projects that it's meant to be "competing" with.

It takes a lot of hubris to say "I want to build a monotool that replaces the functionality of several projects that others have worked extremely hard on. Not only that, I want a big splashy release with lots of hype so I gain a lot of interest and adoption".

To me, I value organic adoption. Now that I've reached a particular position in the community, and people are familiar with the projects I've worked on, anything I release will implicitly get more attention than some others.

I don't want anyone using it because it has X person associated with it. I want it to be used because it's actually useful rather than some flashy new thing that people feel obligated to jump on.

Thinking about it more, there's probably not a lot I can do to prevent it. And doing an "impromptu" release could have the opposite effect. If I didn't want any expectations then I shouldn't have spoken publicly about it at all. That's the main reason I didn't prior.

I don't want early access. I'm afraid of building an ivory tower of privileged people. For Yarn we invited people at the beginning. It allowed them to influence the development process but we overrotated on their feedback.

It's possible to select people who are the most relevant, and I felt like we did. The feedback was valuable, but they weren't necessarily the ones who would actually be invested post-release.

So I spent time satisfying their constraints which would turn out to not be relevant. Again, this can be mitigated if you have a more cohesive community where everyone is mostly on the same page.

Another point is around inclusivity. I don't want anyone to feel left out. No matter how wide you cast the net, there's always going to be someone valuable you've left out. Performing value judgements on people is gross and not really something I want to have to do.

Especially since position in the community, which most use as a measure of relevancy, doesn't necessarily yield the most useful feedback.

I've never received private feedback that wouldn't have been more valuable to have publicly. Once your community reaches a certain size that can be extremely noisy though.

Also depends on the release position of the project. Rome would probably be released as an alpha or beta. Opportunity to change anything, which is typically the purpose of early access.

I feel as though I have the necessary experience and context to build a functional MVP. Any public or private feedback during that time would be unlikely to be unproductive.