No one who talks about this problem posts code to demonstrates the issue. I did testing that shows EA's can backtest very much faster on 610 than they did on 509, that should tell you your problem is code specific. How would you debug a code specific problem without ever seeing the code ? Write a thousand EA's until you eventually hit on something that replicates the issue by chance ?
some more data:
- My EA, .ex4 obtained with 509 version: 188 kbytes
- Same EA, but compiled with 600 version: 721 kbytes (why this huge difference?!?!?)
PERFORMANCE OF EA SIMULATION
- EA simulation with v600-compiled EA: many times slower (> 10) than simulation of v509-compiled EA
- Simulation of v509-compiled EA in v600 environment is still fast (at least as simulation done in v509 environment).
I used the same test setup for 509 and 600: non-visual mode, same .set file.
History: downloaded from my broker.