PDA

View Full Version : limitations of Openlaszlo (?)


MBachstein
12-21-2009, 10:04 AM
Hello,

I have the following grave problem.

When I add a new line to my source code and try to start the application in debug mode (debug=true) the flash application doesn't start any more! :confused:

The debug window does not pop up. There isn't any exception or error. The Openlaszlo-splash is just curling for 10 seconds and then the browser window turns black, but you can call the standard flash context menu.

With debug=false the same source code compiles and the application starts after about 10 seconds.

When I leave the new line out the application starts also in debug mode with the Openlaszlo debug window popping up.

My application is quite big (over 13000 lines of code) and contains a lot of components.

Now it seems to me that I cannot add any further functionality to my application, which is really frustrating. :(

Does anybody know whether there is a limitation in size of source code or memory or count of objects?

Or does the compiler even crash when the compilation takes longer than 10 seconds?

Or does the browser throw a timeout when there is no flash object within 10 seconds?

I don't know what to do. Can anybody of you give me a clue?

Kind Regards,
Marc

kmeixner
12-21-2009, 12:01 PM
Hello,

I have the following grave problem.

When I add a new line to my source code and try to start the application in debug mode (debug=true) the flash application doesn't start any more! :confused:

The debug window does not pop up. There isn't any exception or error. The Openlaszlo-splash is just curling for 10 seconds and then the browser window turns black, but you can call the standard flash context menu.

With debug=false the same source code compiles and the application starts after about 10 seconds.

When I leave the new line out the application starts also in debug mode with the Openlaszlo debug window popping up.

My application is quite big (over 13000 lines of code) and contains a lot of components.

Now it seems to me that I cannot add any further functionality to my application, which is really frustrating. :(

Does anybody know whether there is a limitation in size of source code or memory or count of objects?

Or does the compiler even crash when the compilation takes longer than 10 seconds?

Or does the browser throw a timeout when there is no flash object within 10 seconds?

I don't know what to do. Can anybody of you give me a clue?

Kind Regards,
Marc

Can you post the line that is causing the application to hang? Or do you mean it is a new line character that is causing it to crash?

My application is larger than yours and it works so that shouldn't be the problem. Of course I am still using OL 3.4.0 at the moment not sure if that makes a difference. Also, I think you would see some kind of error message is it was a timeout error.

Check for an un-escaped ampersand (&) returned in the XML to one of your datasets (if applicable), I think I had a similar problem as you one time it that was the reason.

spoco2
12-21-2009, 06:02 PM
There is a limitation in the flash player in terms of how long a chained series of calls can be. (so while the code may be separate, if the sequence of calls for a single operation where the initiator call waits on a return is long it still counts).

We ran into this exact type of issue (back on 3.4.x) except we got some cryptic message on our browser instead of nothing.

To avoid it break up your largest methods into smaller chunks and you should find the problem goes away. We found we had one method that called on a collection of others and in total it came to an awful lot of lines of code... broke that up a bit and haven't run into the problem since.

And our app is huge, let me tell you.

pugmaster
12-21-2009, 08:43 PM
You can start by telling us what version of OpenLaszlo you are running and for what runtime you are seeing this behavior.

Norman Klein
Author: Laszlo in Action

MBachstein
12-22-2009, 12:45 AM
Hello,

I am using OL-version 4.5.1.0 in the swf8 runtime.

The problem was indeed that one methode became too big (over 450 lines).

I splitted the method up into smaller chunks and now the application starts again with both debug=true and debug=false. :)

Thank you so much, spoco2!

Kind Regards,
Marc