Monthly Community Update -- October 2022

Thanks for all the comments and feedback! A couple quick reactions:

  • First, some bad news: the October launch of improvements to our credentials and configuration management does not get us to global hosting on self-serve plans. I know how important that is to a lot of you, so I want to avoid painful misunderstandings here! That launch is one key step on a long roadmap of infrastructure improvements that will eventually get us there, but there’s enough work past that point that I’d be hesitant to give an expected delivery date. Sorry, I know that’s not what you were hoping to hear, but didn’t want there to be any confusion. As a reminder, for people who need hosting outside the US today and can afford our Dedicated configurations, you can contact sales to discuss a custom plan
  • Some better news: we hear you loud and clear re: investing in our ecosystem of Bubble developers, plugin builders, agencies, and everyone else in the community who contributes to making Bubble successful. As we discuss 2023 planning, Bubble’s ecosystem is one of the recurring themes, and we’ve been talking through what resources we’d like to put against it from a product and engineering standpoint. We’re also actively hiring two roles that will be full-time working with and advocating for the needs of developers and our community.
  • Re: downtime and breaking user apps, we definitely acknowledge we have room to up our game here. I’ve shared more in-depth thoughts on it here, but the short version is that we have a lot of room to improve but that improvement takes work and isn’t an overnight fix. Re: the specific issue with Operation timed out, this is on our radar, and something we expect to become less common once we release capacity auto-scaling. We also strongly encourage design patterns such as recursive custom workflows that keep each individual segment of data processing short so that there’s much less risk of approaching the five minute workflow limit. The suggestion in the thread about retrying on unexpected errors is an interesting one, but our take is this would likely just lead to repeated failures and consume a lot of app capacity: breaking work down into smaller chunks we think is a better workaround, and our long-term plans here involve automating the process of breaking things into chunks so that we can do it seamlessly behind the scenes without the app builder having to be aware of it.