From: Joachim Breitner <breitner@kit.edu>
Hi,
the declaration
inductive foo :: "int ⇒ bool" where "P x ⟹ foo x" -- "note the free P!"
makes the system loop during "Proving the simplification rules ...".
Such an inductive predicate is probably not very useful, but it maybe it
is still possible to make "inductive" a mit more robust in this respect?
Thanks,
Joachim
signature.asc
From: Florian Haftmann <florian.haftmann@informatik.tu-muenchen.de>
Hi Joachmim,
thanks for your report. According to your description, the computation
of the inductive.simps seems to be faulty. Is anybody out there who can
investigate this?
All the best,
Florian
signature.asc
Last updated: Feb 01 2025 at 20:19 UTC