Is flag `-type-in-type` required for this repository? about sethits

[m-lindgren]

The problem is that the PR calls bool_rec : ∏ (P : bool → Set), P true → P false → ∏ b : bool, P b and coprod_rec, which also use the Set-eliminator. I suggested some changes to the PR that I tried successfully compiling SetHIT with, so no need -type-in-type.

[m-lindgren]

So I was wrong, -type-in-type is currently needed, but for a completely different reason:

        coqc SetHITs/code/examples/groups.{glob,vo} (exit 1)
File "./SetHITs/code/examples/groups.v", line 118, characters 12-20:
Error:
The term "group_ax" has type "UU" while it is expected to have type
"Type" (universe inconsistency: Cannot enforce UU.u0 <=
hit_signature.u0 because hit_signature.u0 < UU.u0).

        coqc SetHITs/code/examples/rings.{glob,vo} (exit 1)
File "./SetHITs/code/examples/rings.v", line 192, characters 12-19:
Error:
The term "ring_ax" has type "UU" while it is expected to have type
"Type" (universe inconsistency: Cannot enforce UU.u0 <=
hit_signature.u0 because hit_signature.u0 < UU.u0).

...

I believe these comes from this recent commit with innocent looking changes like:

  (** Labels of group axioms *)
-Inductive group_ax :=
+Inductive group_ax : UU :=
 | assoc : group_ax

I assume this is because UniMath has Definition UU := Type. instead of having UU be a Notation (or just using Type).

[m-lindgren]

Closing this following #14. -type-in-type is not needed for this repository. The original issue was that the PR UniMath/UniMath#1467 inadvertently used Set-valued eliminators which caused a universe inconsistency when building this satellite.