Sign In Register

How can we help you today?

Start a new topic

Corona SDK performance

I am starting to integrate Gamesparks into my corona sdk game. However as I've started to implement parts of my game I wanted to make sure that the Gamesparks works under all connection conditions, or that I can communicate to my users that the connection is poor. So I downloaded the Network Link Conditioner for iOS in order to test my game under a range of network conditions. My game works fine under almost all conditions, however if the network is bad enough the app just freezes and becomes unresponsive. I also tested this within the gamesparks corona sdk example and noticed the same thing. https://bitbucket.org/gamesparks/gamesparks-corona Is there something I can do here? Or is this something you at gamesparks would have to fix within the plugin?


Hello Jarrod,
you could first check the internet availability at the very beginning of your app, there is plenty of code on web that shows how to implement it. However, the problem may affect it again.

Giuseppe

 

Hello Jarrod,
Thank you for the feedback, we have started to look at this issue.

Giuseppe

 

Hello Jarrod,
When exactly does your application become unresponsive? When it opens the socket connection or during sending/receiving requests/responses?

Thanks,
Giuseppe

 

This is the only line it prints before it appears to freeze. Right now it will eventually finish, but it appears to freeze a few times before it does. 


13:57:04.614   *** GameSparks SDK v1.1.2 connecting to wss://preview-293711ZXWjA9.ws.gamesparks.net/ws/device/293711ZXWjA9 


I'll be honest I'm new to working with network functionality, so I'm not sure what step the problems start. 

Hello Jarrod,
It seems it gets stuck at the starting of the connection. In this case unfortunately there is no fix for this at the moment, because of poor implementation of socket on Lua. You could try any example that uses the socket.open method and should have sometimes the same behaviour.

 

Giuseppe

No worries, its just good to know it isn't my implementation. Is there any sort of recommended workaround? A check? If not are there plans to update the implementation. 

Login to post a comment