Bug: Crash when attempting mastodon fetch against new instance, because actors haven't been populated

[lmorchard]

Moved my account to a new instance - from hackers.town to masto.hackers.town.

I'd ingested an export from hackers.town, which ensured that I had entries in the actors database.

However, I just tried a fetch against masto.hackers.town to start incrementally backing up my toots. That crashed.

Turns out that since I never ingested an export from the new instance, the actor for my account is not already present in the database.

Should add some code for the fetch process that does a get-or-fetch for actors as needed

[lmorchard]

Just as a followup, I requested an archive from my new instance and imported it. That fixed the crash, so I think my hunch is correct: the actor wasn't populated in the DB until after the import

[skpy]

Fiddling with fossilizer and wondering if I hit the same problem? I self-host a GoToSocial instance, which does not offer an export a la Mastodon. When I try to fossilizer mastodon link I get an error:

./fossilizer mastodon -i dungeoncrawler.world link
[2026-01-07T02:38:07Z ERROR fossilizer::mastodon::instance] Failed to register app

thread 'main' (22924145) panicked at src/cli/mastodon/link.rs:24:56:
called `Option::unwrap()` on a `None` value
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

I obviously have an empty database with nothing imported.