I didn't have that error yet, but the message clearly says that it's a bug either in C2 or one of the plugins. If you want this to be addressed, open a bug report in the forum "Bugs", following the guidelines.

I refuse to accept that what Arima described or my own experience is a browser issue. Because, and I'm serious, if it was then browsers are not a reliable base for gaming - and that would render C2 useless.

What should the customers be told? "To start the game you have to be lucky that your browser has a good day and actually presents the start button. If not, check back later." ?

If it turns out that browsers interpret a standard like HTML5 just as they like and unreliable, one could only look at more insulated solutions.

No, instead I really hope it is a bug in C2 - because it could be traced by Ashley and corrected.

Sorry for my complains, but it really frightens me.

You don't need anything to pick a random instance, just that exact event you are using.

The problem here is that C2 seems to work the same way CC worked. That means if you create an instance it is only accessible in the same event, where it is created. Any other events can only access it one tick later. The instance is created at the end of the current tick, so when you tell C2 to choose a random instance, there are no instances to choose from.

If you give it at least one tick time, it works. For example, make the random pick event a subevent just like this:

+ (empty) -> System Wait 0.1 seconds

++ Pick a random ...

how long you have to wait is dependant on the framerate. With 60 fps one tick needs about 17 ms. If it also is a stable framerate, then waiting 0.02 seconds would be enough.

You could also store the current tickcount to a variable in the function, and then outside the function check if tickcount = variable + 1, and if so pick a random instance.

Thank you divaweb , honestly. And I just started working on r6 two days ago [<:o)]

parside you should pm me with the current status <img src="smileys/smiley2.gif" border="0" align="middle" />

Now that's interesting! I discovered the same just yesterday. A project has three drumloop samples, all playing the same pattern. I switch between them using seek to make sure the new sample starts where the previous just stopped. It worked fine. When exported, the first time running it, it didn't work. All samples started at the beginning. After closing and loading the page again it worked.

I started C2 again, previewed, and the problem was there. I then ended the preview, clicked on preview again - and the problem was gone.

I also experience this with Chrome (since IE doesn't work on my pc and Firefox doesn't fully support web audio). The only difference is that I could see this behavior in an exported project, too.

I wouldn't know how to provide any capx or other material that reliably reproduces this.

But I'm glad you posted this, because I really started to get crazy about this. Now I know I'm not alone.

newt

a&b_or_b&c.capx

<img src="smileys/smiley4.gif" border="0" align="middle">

Here is the corrected version. The mean thing is: it's done with r144.2 (I don't want to reinstall different versions while I'm developing a project)

There were a lot of issues, too many to explain them all. Most of them were already addressed in my last example. You might want to have a closer look at it and this new one. For example, you added an array to your project but the initial size was kept at (10,1,1). With already 10 cells filled with the value 0 you will never get any other frame than 0 ;)

Try to look very closely at the projects. All there is exists for a good reason <img src="smileys/smiley14.gif" border="0" align="middle">

Oh, and last: If all you want is having 3 instances that have different animation frames that change on some event, you don't need to destroy and recreate the instances. Just shuffling the frames would be sufficient then.

instancepickingNOconfusion.capx

Pretty sure most of that has little to do with the sdk, and absolutely positive none of it has anything to do with java.Shamelessly using this post to nudge newt - you have a message <img src="smileys/smiley4.gif" border="0" align="middle" />

I'm sorry, your capx' were done with r145, which is too buggy regarding audio for me to use. If you don't want to recreate them with r144.2 or r139, I propose to wait until a more reliable version is out. <img src="smileys/smiley6.gif" border="0" align="middle" />

I just read your post. I already did an example now, but it is more general. On the other hand, you might learn more from it. This example is pretty simple. There's one sprite instance. Whenever you click on any one instance a new instance is created. When you click on the button, three instances will be destroyed. You can mix creation/destroying as you like. The algorithm behind it should be quite obvious. If not please ask.

It just confuses me that you talk of sprites, instances and frames in a way that it all seems to be the same for you. Since those are terms for specific things in C2, I'm not quite sure exactly how the game mechanic will be.

random_unique_selection.capx

I don't know if Ashley already exposed those nodes to the C2's interface.Yes, the analyzer effect provides you with FFT and access to individual frequency bins.

That sounds complicated. <img src="smileys/smiley4.gif" border="0" align="middle" />

The array in the example can be filled with everything, not just a raising number. If you fill it with the UIDs of the sprites/instances in question, and then randomly choose an array index like in the example, it will return the UID.

Pick Sprite where Sprite.UID = array(r)

If I have some time tomorrow, I'll make an enhanced example.