Question

So updateDisplayList is for laying out, positioning and sizing a component’s children. It seems a waste to me, however, to have all that code running every time a render is called, even if no changes have been made to the relevant properties or child properties.

The way I get around this for performance’s sake (working on mobile) is to set a flag inside of my overridden updateDisplayList that lets a big part of my own sizing/positioning code run only the first time.

Is this a bad idea/ big mistake? I just can’t see the benefit to letting it run every time otherwise if I don’t expect anything that affects sizing/positioning to change.

Answer 1

It seems a waste to me, however, to have all that code running every
time a render is called, even if no changes have been made to the
relevant properties or child properties.

I agree completely.

The way I get around this for performance’s sake (working on mobile)
is to set a flag inside of my overridden updateDisplayList that lets a
big part of my own sizing/positioning code run only the first time.

What I’ve done–and is very common in the Flex Framework–is to have a property for every main item that needs to change. For example, our MX AutoCompleteComboBox has a property named “downArrowVisible” and if you set it to false it will hide the down arrow from display, while also resizing the textInput to spread across the full width of the component. If you set it to true; it will position items like the normal ComboBox.

So, whenever the property downArrowVisible is changed, we set a flag there, downArrowVisibleChanged; and use that property to determine whether updateDisplayList should change position of the elements. [It’s slightly more complicated than that, but that is the gist]. The default of downArrowVisibleChanged is true so that during the initial setup run the children components will be sized and positioned correctly.

This sounds very similar to what you’re doing; as I never have a flag property specific to updateDisplayList(). I make my flags specific to a certain piece of the layout, and only set those flags to ‘no longer need to be changed’ in updateDisplayList().

Is this a bad idea/ big mistake? I just can’t see the benefit to
letting it run every time otherwise if I don’t expect anything that
affects sizing/positioning to change.

I think it seems like a logical approach. You are correct, there is no need to having your layout code consistently running every time the component redraws.

If you will not be changing properties of your custom component at runtime and only want your updateDisplayList() routine to run once–and never again–then I think you can tie into the initialized property of the Flex Framework without adding your own Boolean value. IF the value is false, then the component hasn’t been through it’s initial creation process yet. If the value is true then it has.