But you haven't matched the experience
The hierarchy of needs for a website might look like
and we have accessibility (many unblocked platforms) and usability (decently fast), but we don't have any of the others. We don't have functionality - nothing's easy to use and deploy to yet. We don't have scalability - it's hard to support multiple apps. And we definitely don't have invisiblity - we're a while from the platform disappearing into the background.
The annoying thing is we're almost there
| Name | Usable hostname | Speed | Good free tier |
|---|---|---|---|
| GCP VM | ⚠️ | ✅ | ❌ |
| Fly container | ✅ | ✅ | ⚠️ |
| GCP CR Serverless | ⚠️ | ⚠️ | ✅ |
| Fastly Serverless | ⚠️ | ✅ | ✅ |
| Cloudflare Workers | ❌ (blocked) | ✅ | ✅ |
| Vercel Serverless | ❌ (blocked) | ✅ | ❌ |
| Deno Serverless | ✅ | ❌ | ✅ |
| Azure Functions Containers | ⚠️ | ❌ | ✅ |
| Val (Serverless) | ✅ | ❌ | ✅ |
but we haven't matched the experience quite yet.
So at this point there are 5 workarounds we can deploy:
- Monoliths. Combine services and apps into one, so you only need to deploy it once. If a free tier only lets you deploy one app or gives you terrible cold starts, a monolith would easily offset them. (Needs to be combined with another domain provider to reseparate.)
- Use another static host. Then you can stop caring about how the compute provider charges for bandwidth or how they name their subdomains.
- Use a free domain. This means playing a lot of feeble games, but it could work; not sure.
- Use a paid domain. If you can pay for it and get it unblocked, you now own a part of the internet where you don't need to worry about the free default subdomains. You can even get back to Cloudflare.
- Get off the free tier. This one is really just for Fly, but you could start separating apps and disabling scale to 0 and tank the cost.
At this point, I know what to do, and I think I'm ready to match the experience.