You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm guessing what happens is, when prepare tries to connect to the listening socket owned by the other Windows user, it fails, so it believes the 2nd instance is actually the 1st instance.
But then when register tries to open the socket and listen for itself, it fails because they're in the same namespace.
I thought of a few possible patches, but it's not critical for our project so I haven't started on any:
Move the bind out of the worker thread and change the expect into returning an error before the worker thread is started. This means you can't run the app in 2 Windows sessions at once, but at least it won't panic in the worker thread
Add the user's ID into the socket's ID so that you can have 2 separate sockets, but only one per user.
The text was updated successfully, but these errors were encountered:
Saw this over on the Firezone client: firezone/firezone#2808
The 2nd app instance panics here:
tauri-plugin-deep-link/src/windows.rs
Line 59 in 0c28b75
I'm guessing what happens is, when
prepare
tries to connect to the listening socket owned by the other Windows user, it fails, so it believes the 2nd instance is actually the 1st instance.But then when
register
tries to open the socket and listen for itself, it fails because they're in the same namespace.I thought of a few possible patches, but it's not critical for our project so I haven't started on any:
expect
into returning an error before the worker thread is started. This means you can't run the app in 2 Windows sessions at once, but at least it won't panic in the worker threadThe text was updated successfully, but these errors were encountered: