Android tests - Printable Version +- Cuberite Forum (https://forum.cuberite.org) +-- Forum: Cuberite (https://forum.cuberite.org/forum-4.html) +--- Forum: Development (https://forum.cuberite.org/forum-13.html) +--- Thread: Android tests (/thread-530.html) |
RE: Android tests - NiLSPACE - 12-24-2013 Trying it now It works! Aww but it crashes when it starts generating a world.. Now debugging. RE: Android tests - tigerw - 12-24-2013 Why is it that nearly all crashes (SPARC, RasPi, this) happen when the world is generating? Can you get logs of why it is? Perhaps something within the world generator that is using a function/feature Android doesn't support. Or maybe you ran out of memory RE: Android tests - NiLSPACE - 12-24-2013 Weird.. it doesn't crash on a virtual android device.. RE: Android tests - FakeTruth - 12-24-2013 We once found a crash that was caused by unaligned memory. This was not a problem on Windows or Linux, but Android didn't like it :p RE: Android tests - NiLSPACE - 12-24-2013 Well I'l at least upload the .Apk so others can look at it as well: http://puu.sh/5WoAl.apk RE: Android tests - bearbin - 12-24-2013 STR, did you build from the command line? If so, how? RE: Android tests - FakeTruth - 12-24-2013 STR built the NDK library from commandline, not the APK, he uses Eclipse for that (just like me) RE: Android tests - NiLSPACE - 12-24-2013 No, but I'll look into it tomorrow. RE: Android tests - NiLSPACE - 12-24-2013 When I delete the world and let it regenerate It runs normal, but when it has to load or save the world it crashes. Thats the conclusion I have now. Yes the problem should be there because it doesn't crash when I use the forgetful scheme. RE: Android tests - xoft - 12-24-2013 if you use the forgetful schema and connect to the server, doesn't it crash immediately, too? Because the network protocol makes extensive use of possibly unaligned access. |