Thanks Pete for your very quick response...
Does the link to the app succeed or fail for other people?
Now I'm in trouble! - I've tried running the app from the link in my
posting on 3 different machines, with a mix of Windows 7 and Vista,
Firefox, IE and Chrome - and cannot get it to fail during
initialisation like you did. I've never seen it do this during 6
months of development.
How do "we" begin to debug this?
I've tried looking at the failing code path to see if any smoke or
suspicious smells are visible - with no luck.
My best hunch for continued investigation is to look at the difference
in behaviour from potential run-time environment variations. (Is that
a good approach?)
It would be a great help if a couple of other people could try the
link to see if it works - and say something about your environment.
Does it only fail in one browser/OS/flash player version? (Firefox on
the Vista machine I'm typing on is using flash player 10,2,153,1 ).
Does it fail for you Pete repeatably?
The app doesn't explicitly load any extra, deployed resources at run
time (which I guess could introduce platform variation). Here's the
deployed manifest - created automatically from FB.
Volume in drive C is OS
Volume Serial Number is 08C5-6648
Directory of C:\tmp\manifest
03/08/2011 13:56 <DIR> .
03/08/2011 13:56 <DIR> ..
03/08/2011 13:56 151,912 charts_4.5.1.21328.swz
03/08/2011 13:56 325,307 framework_4.5.1.21328.swz
03/08/2011 13:56 <DIR> history
03/08/2011 13:56 322,038 mx_4.5.1.21328.swz
03/08/2011 12:05 657 playerProductInstall.swf
03/08/2011 13:56 5,906
rm3dee_flex_public_app_whitelabel.html
03/08/2011 13:56 270,822
rm3dee_flex_public_app_whitelabel.swf
03/08/2011 13:56 54,418 sparkskins_4.5.1.21328.swz
03/08/2011 13:56 466,785 spark_4.5.1.21328.swz
03/08/2011 12:05 25,574 swfobject.js
03/08/2011 13:56 186,394 textLayout_2.0.0.232.swz
10 File(s) 1,809,813 bytes
3 Dir(s) 150,596,280,320 bytes free
One other lead I could pursue is to look at errors or behavioural
differences arising from pulling data into the app from the back end
database. I can easily configure the app to not use database data, but
instead to use test data embedded in the app. I'll do that and repost
- to rule out that potential source of the problem.
Any help gratefully received - I'm in somewhat new territory here :-(
Pete
On Aug 2, 9:53 pm, Peter Kapelyan <[email protected]> wrote:
> Got this a couple times before it loads, and don't see anything : (
>
> Error: Unrecognised material bitmap name
> at ThreeD.ModelBuilder::ModelBuilder/_choose_material_from_string()
> at ThreeD.ModelBuilder::ModelBuilder/build()
> at Application::AppReInitialiser/re_initialise()
> at Application::AppBootStrap/_do_the_rest_of_the_initialisation()
> at Application::AppBootStrap/completeHandler()
> at flash.events::EventDispatcher/dispatchEventFunction()
> at flash.events::EventDispatcher/dispatchEvent()
> at flash.net::URLLoader/onComplete()
>
> -Pete
>
> On Tue, Aug 2, 2011 at 4:03 PM, Pete Howard (home account) <
>
>
>
> [email protected]> wrote:
> > Hello,
>
> > My web app (you can try it - see below) uses the popular idiom of
> > letting you pan and tilt using the hover camera by mouse dragging, but
> > when you do, after a certain amount of panning and tilting - errors
> > like the one in the subject line are stimulated. The behaviour is
> > consistent insofar that you can always stimulate it, but there seems
> > no obvious pattern about how soon it will kick in.
>
> > The Away3D version is 3.6.0 (svn tag)
>
> > The stack trace is:
>
> > RangeError: Error #1125: The index 85 is out of range 84.
> > at away3d.core.render::BasicRenderer/render()
> > at away3d.core.session::AbstractSession/render()
> > at away3d.core.session::AbstractSession/render()
> > at away3d.containers::View3D/render()
> > at ThreeD.Controller::ThreeDViewController/_handle_enter_frame()
>
> > The app is publicly reachable here:
>
> >http://rm3dee.appspot.com/static/whitelabel_app/rm3dee_flex_public_ap...
>
> > To stimulate the error - please use the slider to zoom in as far as
> > you go, and then twiddle the model using mouse dragging. (Continuous
> > dragging - no mouse button release). Nb. the zooming in is not a
> > necessary condition to stimulate the error - but it makes it show up
> > almost immediately.
>
> > The scene is built simply using <Plane>s using this code:
>
> > ...
> > ...
> > _parent_3d_object=new ObjectContainer3D();
> > this.view_3d.scene.addChild((_parent_3d_object as
> > Object3D));
>
> > for each (var plane:Plane in display_model.planes) {
> > _parent_3d_object.addChild(plane);
> > }
> > _parent_3d_object.rotationX=90;
> > _parent_3d_object.rotationY=0;
> > _parent_3d_object.rotationZ=180;
> > this.view_3d.render();
>
> > The pan/zoom handling uses this:
>
> > private function _handle_enter_frame(e:Event):void {
> > if (this._dragging == false) {
> > return;
> > }
> > const gain:Number=0.3;
> > _three_d_view_widget.camera.panAngle=gain *
>
> > (this._three_d_view_widget.stage.mouseX - this._drag_x) +
> > _last_pan_angle;
> > _three_d_view_widget.camera.tiltAngle=gain *
>
> > (this._three_d_view_widget.stage.mouseY - this._drag_y) +
> > _last_tilt_angle;
> > _three_d_view_widget.camera.hover();
> > this._three_d_view_widget.view_3d.render();
> > }
>
> > You will notice in the app that the rendering style changes to a fast/
> > cheap style as you initiate the dragging operation. In fact the entire
> > scene contents are replaced on this transition. Before the transition
> > we use ShadingColorMaterial with Renderer.CORRECT_Z_ORDER, and during
> > the dragging (when the problem shows up) we use WireColorMaterial with
> > Renderer.BASIC.
>
> > I'd really appreciate any support or guidance I get to overcome or
> > work around this problem - because my entire app pretty much hinges
> > (no pun intended) on it. Thanks in anticipation.
>
> > Pete Howard
>
> --
> ___________________
>
> Actionscript 3.0 Flash 3D Graphics Engine
>
> HTTP://AWAY3D.COM
