T2 queries on rules that no longer exist on the code

[menocsk27]

While testing for imldresden/nev#1 on the ancestors example, I opened nev with ancestorsOfBoth, removed
line 27 on Nemo Web, re-ran the program, and "focused" on ancestorsOfBoth in the (now old) nev tab, leading to this error.

1. Should T2s attempt to compute with non-existing rules at all?
2. Should we invalidate comms to nev tabs after changes in the code?

[monsterkrampe]

I think it is expected that after changing and rerunning the code on the Nemo side, things in Nev might not match the program anymore. I think this concerns all types of queries including the jump to rule functionality.
I think:

1. In the frontend (Nev or Nemo web) we have no way of telling if a rule still exists so we have to attempt. The validation should be done in the backend anyway.
2. This would be easy to implement I think but I think we would still want to allow minor changes in the code and then it is not easy to draw a line. So I would rather not disconnect the two tabs and live with error messages popping up if the program changes fundamentally.

So here, I think that showing an error is fine. However, this looks like a panic, which is not ok.
I think this needs to be addressed in the Nemo backend. I'm therefore moving the issue.