Zulip Chat Archive
Stream: general
Topic: Bug in NNG?
Martin Dvořák (Oct 10 2024 at 07:25):
I was notified about a (probably?) bug in NNG where the game says "unknown identifier add_left_cancel
" instead of the message "tactic apply
failed, failed to unify" which would be more helpful.
add_left_cancel.png
Martin Dvořák (Oct 10 2024 at 07:27):
In the best case, the game would tell the user that they attempted to use a lemma for the opposite implication, but this is not technically viable probably. However, saying "unknown identifier" when the game displays add_left_cancel
as available is very confusing.
Kevin Buzzard (Oct 10 2024 at 07:27):
But it does unify
Martin Dvořák (Oct 10 2024 at 07:28):
Ah yeah!
Kevin Buzzard (Oct 10 2024 at 07:28):
Of course it's still worth reporting :-)
Martin Dvořák (Oct 10 2024 at 07:29):
In this case, the tactic should probably emit an auxilliary goal to specify n
.
Martin Dvořák (Oct 10 2024 at 07:29):
I am not sure what the intended behavior of NNG is in such a case.
Jon Eugster (Oct 10 2024 at 07:48):
I created
https://github.com/leanprover-community/NNG4/issues/78
for when I'm back working -- or if somebody else has the time. It might be a problem with either the opened namespaces or the imports of that level, but should be quick to fix.
Last updated: May 02 2025 at 03:31 UTC