model results randomly feasible and infeasible

[disonvon]

What version of OR-Tools and what language are you using?
Version: 9.9.3963
Language: C++/Java/Python/C#
Python
Which solver are you using (e.g. CP-SAT, Routing Solver, GLOP, BOP, Gurobi)
CP-SAT
What operating system (Linux, Windows, ...) and version?
Mac Os 13.0 (22A380)
What did you do?
attached two model files are generated from two independent run with same input, but one is feasible and another is not. the two files are the same but the variable or the constraint order

What did you expect to see
the two results should be aligned and feasible
test_feasible.pd.txt
test_infeasible.pd.txt

[lperron]

Can you try with 9.10?

[disonvon]

9.10 also tried, perform the same behavior like 9.9

[lperron]

are you sure the models are equivalent ? you change switching one interval between no_overlaps

[lperron]

I do not believe this is a bug, Moving to discussions.

[disonvon]

it turns out that it was a bug at my side. Thanks for your time.