Small improvements for TicTacToeEnv (#692)

* More explicit logging in TicTacToeEnv

* Fix type instability in TicTacToeEnv

Author: jonathan-laurent

src/ReinforcementLearningEnvironments/src/environments/examples/TicTacToeEnv.jl

@@ -58,9 +58,9 @@ RLBase.legal_action_space(env::TicTacToeEnv, p) =
 
 function RLBase.legal_action_space_mask(env::TicTacToeEnv, p)
     if is_win(env, CROSS) || is_win(env, NOUGHT)
-        zeros(false, 9)
+        falses(9)
     else
-        vec(view(env.board, :, :, 1))
+        vec(env.board[:, :, 1])
     end
 end
 
@@ -135,7 +135,7 @@ end
 
 function get_tic_tac_toe_state_info()
     if isempty(TIC_TAC_TOE_STATE_INFO)
-        @info "initializing state info..."
+        @info "initializing tictactoe state info cache..."
         t = @elapsed begin
             n = 1
             root = TicTacToeEnv()
@@ -160,7 +160,7 @@ function get_tic_tac_toe_state_info()
                 end
             end
         end
-        @info "finished initializing state info in $t seconds"
+        @info "finished initializing tictactoe state info cache in $t seconds"
     end
     TIC_TAC_TOE_STATE_INFO
 end