You can also use the subtract blendmode. BitmapData.draw supports using a
blendmode, and subtract does exactly what you want.
--Brian
On 8/21/07, Bojil Vassilev [EMAIL PROTECTED] wrote:
Thank you both for the replies. There is some urgent stuff at work right
now, but I'll test more later and
While what you say is true, it's not really a 'good' reason. Quicktime has
been around for far longer than flash, and it allows seeking to any time.
Seeking to a non-keyframe is much slower than seeking to a keyframe (since
you must decompress the intermediate frames), but it can be done. If
As you have discovered, changing class hierarchy and -keep don't go very
well together. Classes need to be initialized in the proper order, or the
prototype chain isn't set up correctly. When you use the keep flag, mtasc
keeps all existing classes (and preserves their order), and then adds all
How are you embedding your flash? I remember reading about a bug fix for
swfobject which addressed crashes when unloading flash movies that used
rtmp. It may be something that is only an issue in certain versions of
browsers/flash players.
--Brian
On 6/12/07, Johannes Nel [EMAIL PROTECTED]
I may be wrong, but I don't think you can do it just with a color matrix,
but you could try drawing the solid color you want to tint with, and use the
greyscale image with the hardlight blendmode on top. I think that should
give you the desired effect.
check out
did you try filling with 0? You're setting the background color to an
opaque color, and then drawing the transparent bitmap on top of it.
var m_cBitmap = new BitmapData( target_mc._width, target_mc._height,
true, 0x0 );
if you don't specify the color, it might be using 0xff00 (black with
Someone already mentioned what it is - the for in loop vs the straight for
loop.
for (var i in array) ends up pushing all the keys of the array on the stack,
and then popping those off.
for (var i =0; i array.length; i++) doesn't.
The test attaches 500 listeners to each dispatcher, so when you
Yes, I've noticed that too. Classes are only loaded once, so if swf A uses
class Foo, and then loads swf B into a lockroot, if swf B also has a class
Foo, it's going to be using the class defined in swf A (with the wrong
root). Note: it's not just the interpretation of _root. class Foo from
Look at mx.events.LowLevelEvents.as
The way it works, as soon as you add an eventlistener for one of the low
level events, it will replace onPress, onRelease, etc with new methods that
dispatch an event.
Code that's in your configuration folder is only looked at once when
starting flash, so if
Each DataGridColumn supports a label function, and a cell renderer. For
each row, add the same object multiple times (as many as you have columns),
and set the columns up to have different label functions. Or use a custom
cellrenderer.
--Brian
On 11/15/06, Mark Hawley [EMAIL PROTECTED] wrote:
A trap I often fall into when debugging flash is missing the obvious. Take
for example,
If I put movieclip._y = 60 , it works fine
If I put movieclip._y = textField._height , it doesn't work fine
and then
trace(textField._height) prints 60, yes.
then my question would be - did you
If you're using the Flash classes for playback (like the FLVPlayback class),
it actually looks at the suffix of the loaded url to determine whether it's
an FLV or xml (playlist?). You can hack it to work by loading
video1.flv?id=26-09-2006type=.flv
I haven't actually dug through the classes to
The bitrate is only a target. If you use single-pass encoding, the codec
will attempt to match the bitrate, but if it encounters a segment that uses
more bits, it'll hope that it can make it up later. But if the keyframe
interval is too low or resolution is too high, it may never be able to,
13 matches
Mail list logo
