Hi. I read the sunsetting announcement of classic updates & expo publish and I feel frustrated. I’d like to apologise upfront for the candid feedback and tone of this post.
My team has been a heavy user of classic updates to quickly deploy fixes and features to our app. We’re very happy with the service because it works very well and it has no additional cost. This is why we are sad to see it sunsetting without a fair replacement, as EAS Update costs $5 per 1k monthly updaters.
Going straight to the point: I wish my team could keep using a expo publish-like workflow inexpensively after classic updates sunset on Feb 12, 2024. I would also like to avoid having to host custom-expo-updates-server (nextjs). Ideally, I’d like to host a plain simple AWS S3 bucket with web public access.
Has anyone done something like this? Considering the expo updates protocol, is this doable without significant changes to the existing client?
I imagine the expo team must have build some amazing new features on the EAS Update service, nevertheless it seems to me that the essence on the server side should be no more than a simple static web hosting (think AWS S3, robust and unexpensive - no CDN needed) as all the logic can live in the client. With this in mind, the EAS Update pricing feels quite higher than I expected.
Thank you.
PS: It shouldn’t matter for this discussion, but here’re some specific stats of our app: it has currently 60k monthtly users, but most of these are ephemeral new users that won’t become active. Actual active users are in the range of 10-15k.