In D15512#425772, @ashoat wrote:

I am skeptical of this approach. The core issue is that our FID extraction breaks when unexpected object fields are encountered. What happens if Farcaster introducing a new object field? Won’t the same issue occur? I think we need to fix the core issue instead of reverting this. Let me know if I’m missing something.

rebase before landing

I am skeptical of this approach. The core issue is that our FID extraction breaks when unexpected object fields are encountered. What happens if Farcaster introducing a new object field? Won’t the same issue occur? I think we need to fix the core issue instead of reverting this. Let me know if I’m missing something.

We need this for CORS requests

In D15486#425434, @ashoat wrote:

Going to land this as-is, but would appreciate a re-review from @kamil on Monday to make sure the changes address his concern

Going to land this as-is, but would appreciate a re-review from @kamil on Monday to make sure the changes address his concern

Good call! I asked Codex to address the concern, and it added an early exit to compaction.upload_files when the user_data is present but the user_keys is missing

Updated the description with a video of the whole experience.

In D15487#425421, @ashoat wrote:

Just confirming, pressing the sync button in Profile pops up the whole experience again, right? It's not just in the background?

Update the copy

Just confirming, pressing the sync button in Profile pops up the whole experience again, right? It's not just in the background?

Let's do this:

In update inbox and lightweight sync fetch the thread but without too many messages in the canceled state

This diff introduces an alert - @ashoat could you review the copy?

update