Skip to main content
10 events
when toggle format what by license comment
Mar 24, 2022 at 20:22 vote accept prochet
Mar 24, 2022 at 17:14 history edited LSpice CC BY-SA 4.0
Capitalise title
Mar 24, 2022 at 13:46 answer added Jason Starr timeline score: 6
Mar 21, 2022 at 9:30 comment added Jason Starr Okay, I think I have a counterexample where the base is an open affine over a “classical” Enriques surface (as opposed to “singular” or “supersingular”, all of which are smooth Enriques surfaces). I will try to write more details soon.
Mar 21, 2022 at 7:27 history edited prochet CC BY-SA 4.0
added 7 characters in body
Mar 20, 2022 at 21:35 comment added prochet I'm really over an affine base. Several things are much different whe the base scheme is projective, there is much less torsors, but I'm curious about your counterexample.
Mar 20, 2022 at 11:31 comment added Jason Starr It seems easier to find counterexamples if you allow a projective base scheme rather than the affine scheme $\text{Spec}\ A$. Are you interested in counterexamples over a projective base scheme?
Mar 19, 2022 at 20:30 comment added prochet But why such a torsor would admit a maximal torus after a finite surjective (or finite flat) map?
Mar 19, 2022 at 11:20 comment added Jason Starr For definiteness, consider torsors for a split, simple algebraic group $G$ of type $E_8$. Each of these gives rise to an inner form of $G$. Now consider the case that $A$ is the coordinate ring of a simply connected surface over an algebraically closed field $k$. By Serre's "Conjecture II" (solved by a large number of contributors in this case), every $G$-torsor is Zariski locally trivial (this also uses the Grothendieck-Serre Conjecture, which is also solved in this case). So you are asking whether every $G$-torsor is globally trivial. I believe that is false.
Mar 19, 2022 at 9:19 history asked prochet CC BY-SA 4.0