I’m thinking about this whole “missing blobs” problem after migrating between PDSes. Could a server process iterate over all the posts to check for missing blobs, and then import them automatically? What requires the user to manually detect this? The more automated, the better; for more users.

a little atproto bulk import hack: PDS instances have record operation rate-limits (for good reasons!) which can make bulk import of thousands of records all at one tricky. eg, "import your old insta posts".

I recently got super into the ATProto spec and how different services use PDS to store the data, I put together a quick client-only website that presents all those records in a single, consolidated feed of all my activity: at.diego.codes

a few atproto tools which made it easier to cryptographically verify individual records would make it matter less where records live and we could move away from a single-PDS world for use-cases which don't need universe-scale state (not saying go full Nostr, just thinking on a happy medium)

Hi protocol friends! I've been cooking up a Very Large PDS (vlpds) implementation capable of scaling up to billions of accounts and down to just one. It uses Object Storage as its only state, supports HA operation, live scaling up and down, is feature-complete with the reference PDS, and is FOSS.

a few atproto tools which made it easier to cryptographically verify individual records would make it matter less where records live and we could move away from a single-PDS world for use-cases which don't need universe-scale state (not saying go full Nostr, just thinking on a happy medium)