Greg Morenz

You can still install this by cloning the repo and running `cargo install --path . --locked`.

> Could you add a new data type, something like a VectorList and a new node that takes that vector list as input and adds all the vectors together? Hmm,...

> The 'new node' dropdown is not clipped to the graph's ui contents. I'm not sure if this might violate assumptions in egui, but I prefer the menu not clipping...

#30 makes this a bit easier once it's merged. You can now lookup the nodes outputs via `graph.outgoing(output_id)` to check if they are connected to anything. Which is more efficient...

Hmm... it's obviously not succeeding, I'll try and figure out why. Is it a good idea to be running the normal runtime code on Android in the first place though?...

With [this line](https://github.com/tomaka/android-rs-glue/blob/master/cargo-apk/injected-glue/lib.rs#L272) being commented out, `g_mainthread_boxed` never becomes `Some(_)`, so `is_app_thread_terminated` always returns as if it was a new process. That line was commented out in [this](https://github.com/tomaka/android-rs-glue/commit/07d0be2d6b694bf3628265209bbf4c91092504db) comment, I'm...

> The real code won't work if you don't call the normal runtime code. > For example that's why the glue uses raw pthreads instead of Rust threads. Doesn't the...

There should be a way to hide it (semi-)permanently, probably for the rest of the session. It's good to know when something goes wrong, but if you have code that...

What I'm trying to use it for right now is to prove code correct that in turn proves that certain turing machines don't halt, as part of [this](https://bbchallenge.org). This comes...

Thanks for the replies :) > With a feature like this we would also need to think through handling of unwind paths, which Creusot currently ignores by compiling with panic=abort....