amaesani
commented
11 years ago The very practical reason for this is that we simply reuse the neural network code also for the microcontroller.
On Tuesday, May 14, 2013, Olivier Wenger wrote:
Hi, Following comment #15https://github.com/amaesani/robogen-simulator/issues/15#issuecomment-17758426, I was looking how to split main function in several parts.
In my opinion, it would make more sense to do all the neural network processing closer or inside the Robot class. It may not be so easy to move this part of the code, this may appear latter. But for now I was willing to know if there was a good reason to perform this processing directly in the main loop.
ouly
— Reply to this email directly or view it on GitHubhttps://github.com/amaesani/robogen-simulator/issues/25 .
Hi, Following comment https://github.com/amaesani/robogen-simulator/issues/15#issuecomment-17758426, I was looking how to split main function in several parts.
In my opinion, it would make more sense to do all the neural network processing closer or inside the Robot class. It may not be so easy to move this part of the code, this may appear latter. But for now I was willing to know if there was a good reason to perform this processing directly in the main loop.
ouly