Thomas Goyne

Is the order of the volumes the same? If it is then the names shouldn't matter. If it isn't then I think logic like what [vfuse](https://github.com/chilcote/vfuse/blob/master/pkgroot/usr/local/vfuse/bin/vfuse#L276) does would be needed,...

It's working fine for us, but we're using AutoDMG to build VM images rather than imaging physical machines.

Er, `on_change()` is absolutely not thread-safe in the slightest. It relies on the fact that it is always called on the single dedicated worker thread for it and explicitly calling...

Not blocking until notifications are ready would completely break using notifications for pretty much anything. That discussion is starting from a very strange place: "Calling Realm.Refresh from the main thread...

Determining why your `BindingContext::did_change()`-based approach did not work is very important, because something built on the logic for that that is exactly how I'd implement that function.

Oh, the obvious problem is that you need a check to see if you've actually made it up to the desired version yet. Conceptually you want something like: ``` bool...

I don't have an objection to a one-shot refresh notification which notifies only when the refresh is to the version which was latest at the time when the notification was...

The crashing line here is https://github.com/realm/realm-core/blob/master/src/realm/object-store/impl/list_notifier.cpp#L100 We check if the List is valid at the start of that function (`!m_list || !m_list->is_attached()`), and any sort of bug in the handover...

Oh, notification bugs related to unresolved links would be pretty unsurprising, and also would explain why it's only happening on one object (if that object is just the only one...

There isn't any obvious reason why that would cause problems.