I don't think I have a very good understanding of your "loop". You seem to have some sort of recursion that will cause your call stack to grow and grow. it looks as though 1 calls 2 which calls 1 which calls 2 which calls 1 which calls 2 which calls....
At some point anything you do should cause a stack overflow. This normally would create an error without a crash, but it might be that some operations assume space and will crash. If you don't intend the recursion, you might be able to break out the start-from-begining essentials of one command to be called by both commands. I'm just guessing here; I don't have a good understanding of what you are doing. Since this runs fine on W10, I have a feeling that my guess is goofy. Dar > On Jun 2, 2016, at 8:38 AM, Tiemo Hollmann TB <toolb...@kestner.de> wrote: > > Hello, > > OS X 10.11.5, LC 8.0, IDE. Same program runs fine on Windows 10 > > I have a pretty nasty issue, where I am right now a bit clueless, in which > direction I have to search. I have set the dontuseQT of player "myPlayer" > to false because I am using old QT videos, which can't be played with AVF. > > In a *videoPlay* handler I *set* the filename of player "myPlayer" to > tVideoFile, which works fine in all situations and the video is played > afterwords fine. > > Now I have a pretty deep structure of handlers, as a kind of a loop with the > modal userdialog and playing a video. The handler with the sheet/modal stack > waits for some user interactions, calls the *videoPlay* handler and calls > again the first handler, so that it starts from beginning. In the > *videoPlay* handler randomly this error occurs, when setting the filename of > the player object. > > > > A simplified structure looks like this: > > command1: > > *modal* stack myDialog > > command2 > > > > command 2: > > *set* the filename of player -> error > > command1 > > *modal*stack myDialog > > > > > > I have already tried to set a *wait" with messages before the *set* the > filename, but it still crashes randomly. > > As far as I can see, there are two main differences between the handler, > where the issue occurs and where not. 1. The deeper structure of handlers > and 2. The modal stack in between this handler "loop". > > > > My feeling is that the engine choks somewhere in the "loop", but with > different handler calls with/without *send* / *dispatch* I would probably > mess up the "loop" structure. > > > > Any idea, where to start? > > Thanks > > Tiemo > > > > > > > > > > _______________________________________________ > use-livecode mailing list > use-livecode@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your subscription > preferences: > http://lists.runrev.com/mailman/listinfo/use-livecode _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
