build 1030 onwards issue with Runtime device crash.

[bubblebobble]

this is something I have come across when switching from build 971 to 1031.

It was great news to see the current android.permissions changed.

My project only needs access to system.vibrate(), android build settings updated no worries, the vibrate is working correctly.

However now I get a runtime error and a crash on the device when the player loses the third life.

Now there is an error in the console but it doesnt affect the game in any way when built with 971.
Build it with 1030 and the runtime error appears.

Console error is an upValue issue.

I know its a little slack and not good practise to allow things to creep into the code.
What I would like to ask is: what else has changed for this issue to be caught and trigger the crash on the device?

I am sure I read a thread somewhere covering this issue but I am unable to find it.

Thanks [import]uid: 127675 topic_id: 37079 reply_id: 67079[/import]

[bjsorrentino]

Hi @bubblebobble,
It might be difficult to determine, since quite a lot has changed since 971. You mention an “upvalue” error report, but are there any more details like a specific variable name? Reference to a function or API call? Anything else to help narrow it down?

Thanks,
Brent [import]uid: 200026 topic_id: 37079 reply_id: 145310[/import]

[bubblebobble]

HI Brent,

super quick response.
I appreciate the issue is somewhat sketchy, so what I intend to do in the next few hours is selectively go through the daily builds in order to see which build it begins to fail on.
I will report back when I have my findings.

Like I say, I know it is slack coding that triggered the upvalue issue in the console but it didnt have any effect on the device whatsoever. I also appreciate I should fix the code, however I am interested to find the cause.

cheers [import]uid: 127675 topic_id: 37079 reply_id: 145313[/import]

[bubblebobble]

UPDATE

First I fixed the upvalue error and triple checked my build settings, all was perfect.
I then get a different runtime error (standard :?Runtime Error) once built with 1030. The game wouldn’t load whatsoever.

I figured I would roll back my builds and step forward to see when the error occurs as mentioned in the previous post.
I began with build 1005.1020,1021,1023,1024,1025,1026,1027,1028,1029; all with no errors or crashes.
Build 1030, 1031caused the error.
I then decided to jump to the latest build 1043, no errors and runs perfectly.
So error vanishes between 1031 and 1043.
I may go through selectively again and find what build it fixes itself just for my curiousness.

Maybe somewhere the error checking was bought forward and flags an error, instead of keeping the code live if its running; instead pulls up the error dialogue ?

[import]uid: 127675 topic_id: 37079 reply_id: 145388[/import]

[bjsorrentino]

Hi @bubblebobble,
It might be difficult to determine, since quite a lot has changed since 971. You mention an “upvalue” error report, but are there any more details like a specific variable name? Reference to a function or API call? Anything else to help narrow it down?

Thanks,
Brent [import]uid: 200026 topic_id: 37079 reply_id: 145310[/import]

[bubblebobble]

HI Brent,

super quick response.
I appreciate the issue is somewhat sketchy, so what I intend to do in the next few hours is selectively go through the daily builds in order to see which build it begins to fail on.
I will report back when I have my findings.

Like I say, I know it is slack coding that triggered the upvalue issue in the console but it didnt have any effect on the device whatsoever. I also appreciate I should fix the code, however I am interested to find the cause.

cheers [import]uid: 127675 topic_id: 37079 reply_id: 145313[/import]

[bubblebobble]

UPDATE

First I fixed the upvalue error and triple checked my build settings, all was perfect.
I then get a different runtime error (standard :?Runtime Error) once built with 1030. The game wouldn’t load whatsoever.

I figured I would roll back my builds and step forward to see when the error occurs as mentioned in the previous post.
I began with build 1005.1020,1021,1023,1024,1025,1026,1027,1028,1029; all with no errors or crashes.
Build 1030, 1031caused the error.
I then decided to jump to the latest build 1043, no errors and runs perfectly.
So error vanishes between 1031 and 1043.
I may go through selectively again and find what build it fixes itself just for my curiousness.

Maybe somewhere the error checking was bought forward and flags an error, instead of keeping the code live if its running; instead pulls up the error dialogue ?

[import]uid: 127675 topic_id: 37079 reply_id: 145388[/import]