Thanks for the help Alex, I was unable to figure out why this subcomponent draws correctly in one instance and not on the other. The properties of the drawn instance (except for the initialized/invalidate) look comparable to the drawn one, both have a nestLevel of 5.
I have tentatively gotten around this issue by removing the undrawn subcomponent and re-adding it on the CREATION_COMPLETE handler for the parent component. Thanks, - Ivo ----- Original Message ---- From: Alex Harui <[EMAIL PROTECTED]> To: [email protected] Sent: Friday, May 30, 2008 9:20:57 PM Subject: RE: [flexcoders] what prevents a FlexEvent.CREATION_COMPLETE? Nothing really. All components are created with $visible=false until creationComplete. This actually implies that the components haven’t been addChild’d to the parent or some parent higher up the tree is not on the display list. If nestLevel isn’t greater than 2 or so, then that tree of objects hasn’t been properly parented. ________________________________ From:[EMAIL PROTECTED] ups.com [mailto: [EMAIL PROTECTED] ups.com ] On Behalf Of ivo Sent: Friday, May 30, 2008 3:56 PM To: [EMAIL PROTECTED] ups.com Subject: Re: [flexcoders] what prevents a FlexEvent.CREATION_ COMPLETE? Thanks again for your reply Alex. Seems that in the instances that are not drawn neither the commitProperties( ) nor measure() functions are called. Reviewing the properties of the drawn and not-drawn instances on the creationComplete event of the parent shows them to be basically equivalent, but for a few. The not-drawn instance has the following properties: $visible false initialized false invalidateDisplayLi stFlag true invalidatePropertie sFlag true invalidateSizeFlag true The drawn instances has those same properties showing the opposite value. What does the $visible property indicate? Thanks, - Ivo ----- Original Message ---- From: Alex Harui <[EMAIL PROTECTED] com> To: [EMAIL PROTECTED] ups.com Sent: Friday, May 30, 2008 10:17:49 AM Subject: RE: [flexcoders] what prevents a FlexEvent.CREATION_ COMPLETE? The order is commitProperties, measure, then updateDisplayList. If code in measure() ends up calling invalidatePropertie s, you go back to commitProperties and might never get over to updatedisplayList ________________________________ From:[EMAIL PROTECTED] ups.com [mailto: [EMAIL PROTECTED] ups.com ] On Behalf Of ivo Sent: Friday, May 30, 2008 9:34 AM To: [EMAIL PROTECTED] ups.com Subject: Re: [flexcoders] what prevents a FlexEvent.CREATION_ COMPLETE? Thanks for the reply Alex. One additional question, what would prevent updateDisplayList from being called after a component is added to the display list? Further tracing shows that in the correctly drawn instance the updateDisplayList of the child components is called. In the one where the child subcomponents are not drawn it is not called. If I iterate over the children of the parent component it does list the subcomponents are being present. Thanks, - Ivo ----- Original Message ---- From: Alex Harui <[EMAIL PROTECTED] com> To: [EMAIL PROTECTED] ups.com Sent: Thursday, May 29, 2008 2:56:00 PM Subject: RE: [flexcoders] what prevents a FlexEvent.CREATION_ COMPLETE? If the child never finished validation because a call to updateDisplayList invalidates size or properties, you won’t get creation_complete ________________________________ From:[EMAIL PROTECTED] ups.com [mailto: [EMAIL PROTECTED] ups.com ] On Behalf Of ivo Sent: Thursday, May 29, 2008 11:42 AM To: FlexCoders Subject: [flexcoders] what prevents a FlexEvent.CREATION_ COMPLETE? Hello listers, I have a custom component I use in two locations in my app. In one instance it appears with all subcomponents laid out correctly but in the other it appears with an inner subcomponent blank, it is not drawn. This only happens on the web Flash Player tho, the same component under AIR shows both instances drawn correctly. Tracing I find that under AIR I can handle the FlexEvent.CREATION_ COMPLETE for the inner subcomponent for both instances. In the web version only one of them (the one that appears complete) triggers the event handler, the other one never gets triggered. I imagine this would cause the empty component symptom since this is dispatched once drawing is complete. The CREATION_COMPLETE event for the parent is always triggered tho. Thanks, - Ivo
