Fix disconnects breaking view updates for other connections

[joshua-spacetime]

Description of Changes

Fixes a bug in client disconnect logic that would mark a client's views as dropped(unsubscribed). However it was marking the identity's views as dropped, not the connection. So if an identity had multiple connections open, each subscribing to different views, and one of them disconnected, the subscriptions for the other connections would break. The observed behavior would be that they would stop receiving subscription updates. This could potentially lead to their client cache getting into a corrupted state.

Now, instead of dropping all of the views for a particular identity on disconnect, we drop only the views for that particular connection. And when I say drop, what I really mean is decrement. A view is not dropped completely unless it no longer had any subscribers.

API and ABI breaking changes

None

Expected complexity level and risk

2

Testing

Regression smoketest was added

[gmorenz]

That sounds like a good find, but I don't think it explains the issue I'm seeing in #4603. I have a different identity for each connection because I'm starting them with .with_token(None::<String>). And I just confirmed that

println!("{}", player1.conn.identity());
println!("{}", _player2.conn.identity());

prints two different things.

[aceysmith]

@joshua-spacetime Sorry to say, but I don't think this fix is working. I'm on 2.0.5 standalone and rust module (2.0.4 on TS SDK) and I still see this problem. If I connect with one user, then connect and subsequently disconnect with a second user, almost all of my subscriptions for the first user do not work and I see client errors like:

Updating a row that was not present in the cache. Table: mobs, RowId: 279

[AngelOnFira]

I believe I'm still seeing this as well with the Rust SDK at 2.0.5. Similar behaviour to @aceysmith, the second user connecting then disconnecting triggers a panic:

thread '<unnamed>' (1232367) panicked at /Users/forest/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/spacetimedb-sdk-2.0.5/src/client_cache.rs:220:13:
internal error: entered unreachable code: a delete update should correspond to an existing row in the table cache, but the row WithBsatn { bsatn: b">4\x1d#\xdf\r\xfb\xd2\x91\xe7\xe0XQ\xdb|\x8dK`\xecx\x85\xa9\xb7\x078\xd4f\xfaX\x86\0\xc2\n\0\0\0Pilot-343E\x01\0\0\0\0\0\0\0\xe8\x03\0\0\0\0\0\0\x01", row: Player { identity: Identity(c2008658fa66d43807b7a98578ec604b8d7cdb5158e0e791d2fb0ddf231d343e), name: "Pilot-343E", planet_id: 1, credits: 1000, online: true } } was not present
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

thread '<unnamed>' (1232959) panicked at /Users/forest/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/spacetimedb-sdk-2.0.5/src/client_cache.rs:220:13:
internal error: entered unreachable code: a delete update should correspond to an existing row in the table cache, but the row WithBsatn { bsatn: b">4\x1d#\xdf\r\xfb\xd2\x91\xe7\xe0XQ\xdb|\x8dK`\xecx\x85\xa9\xb7\x078\xd4f\xfaX\x86\0\xc2\n\0\0\0Pilot-343E\x01\0\0\0\0\0\0\0\xe8\x03\0\0\0\0\0\0\x01", row: Player { identity: Identity(c2008658fa66d43807b7a98578ec604b8d7cdb5158e0e791d2fb0ddf231d343e), name: "Pilot-343E", planet_id: 1, credits: 1000, online: true } } was not present

...

[joshua-spacetime]

Updated fix in #4648