Forum user: Manuel Razzari http://ultimorender.com.ar/funkascript

This fix worked OK for me, even on 1.2.5.

Much appreciated!

Here's how.

$f("player", "flowplayer-3.2.5.swf", {
	
	/* On the common clip. */
	clip: {
		
		onMetaData: function(clip){
			var cuepointInterval = 5; // In seconds. 
			var cuepoints = [];
			for (var i=1, t=Math.floor(clip.duration / cuepointInterval)+1; i<t; i++){			
				cuepoints.push(i * cuepointInterval * 1000);
			}	
			
			clip.onCuepoint(cuepoints, function(clip, seconds){
				// You cuepoint handling code.
			});
		}
	}
});

Hello Tero, glad I could help.

I might write up my custom configuration if I have some time. No need for payment, I'm happy to be able to contribute something back to this project :)

By the way, I've just found something else... while testing isSupported() on WinXP IE6 fp 6.0.21, I got "Line 1590 - '0' is null or not an object".

I fixed this by adding version to the return statement on line 1566:

} catch(ee) {
	if (version[0] == 6) { return version; }
}

... I haven't tested this thoroughly, though it looks like it should work...

#1: Line 1398, if I understand correctly (midnight here ;) the flashembed's API getConf() method should return opts rather than conf (the later is undefined).

getConf: function() {
	return opts;	
},

#2: Line 534

if (html && !flashembed.isSupported([9, 0])) {

shoud be

if (html && !flashembed.isSupported(params.version)) {

Hmm ok. I believe I now understand the rationale for line 535's wrapper.innerHTML = ""; and take back my "Anyway" paragraph above. Since the standard use case for the root's innerHTML is a splash image in Flowplayer but alternate content (no-Flash) in flashembed, you are forcing flashembed into its most primitive behavior.

A quick workaround for this, for anyone who wishes to create an alternate fallback, for example to translate the "Flash version 9 or greater is required" message to another language, would be the following:

// Your regular FlowPlayer instance
$f("player", "flowplayer.swf", "my-video.flv"); 

// If the user doesn't have the FP required version
// and expressInstall is not an option...
if (!flashembed.isSupported($f().getFlashParams().version) && 	
     !flashembed.isSupported([6,65])){ 				
	// Add whatever custom fallback you wish to provide. 
	// In this case I use jQuery to add a class to the BODY.
	$('body').addClass('badFlash'); 
	// Prevent the player from loading
	$f().onBeforeLoad(function(){ 
		return false; 
	})
}

3.1.0.js, Line 533, notice the typo in wrappper.
So if the player version is not supported the script just breaks.

// do not use splash as alternate content for flashembed
if (html && !flashembed.isSupported([9, 0])) {
	wrappper.innerHTML = "";	
}

Anyway, I'm probably missing something, but I'm not sure why that code is there at all, since that should handled later by flashembed. It actually prevents the "fail #2.1" from working... and even if it worked... the #2.1 logic is empty! Maybe add an onError event there? or a CSS class like "unsupported-player" to the wrapper? In that way, users can customize (or localize!) the fallback message.

Also - maybe I just need a stronger coffee, but I've had a very hard time finding your SVN repo. Perhaps on the homepage you could use "Download the source code or browse the SVN repository." And/or on the Bugs forum you could add a "read me first" with those links...

That sounds great, glad I could help :)
And thanks to you for this awesome player.

Hello all,

This is a documented Flash issue.

This bug is often visible on .NET pages, which wrap the entire HTML inside a FORM tag.

Quick fix: search inside flowplayer-3.1.0.min.js for

}window.flashembed

Notice the curly bracket at the beggining?

Before that bracket, add the following:

if(document.all){window[o.id]=document.getElementById(o.id);}

If you're working with the non-minified flowplayer-3.1.0.js the fix is to add the following right at the end of the "Flash" function:

if(document.all){window[opts.id] = document.getElementById(opts.id);}

I'd like to be able to offer a nicer solution that doesn't require tampering the core file, but I guess this should be included as a bugfix anyway...

Hope this helps!