You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think it would also be helpful to allow a subgroup of a group to be defined with a predicate that isn't assumed to take values in h-propositions. This would avoid us needing to strip_truncations and apply tr in various places, and would mean that we don't need a separate ab_cokernel_embedding, since ab_cokernel would already not need truncation.
But we also want the concept of subgroups where the predicate does land in h-props, since some statements are only true in that case. (E.g. equiv_path_subgroup and ishset_subgroup.)
However, I think that most statements won't need this extra condition. I'm not exactly sure how to structure things.
The text was updated successfully, but these errors were encountered:
As discussed in #2035:
I think it would also be helpful to allow a subgroup of a group to be defined with a predicate that isn't assumed to take values in h-propositions. This would avoid us needing to
strip_truncations
and applytr
in various places, and would mean that we don't need a separateab_cokernel_embedding
, sinceab_cokernel
would already not need truncation.But we also want the concept of subgroups where the predicate does land in h-props, since some statements are only true in that case. (E.g.
equiv_path_subgroup
andishset_subgroup
.)However, I think that most statements won't need this extra condition. I'm not exactly sure how to structure things.
The text was updated successfully, but these errors were encountered: