David vs. Goliath

At one point André Michelle, Kai-Philipp Schöllmann and I really thought that we had success after a long time of struggle. We had the audio hack back in the popforge days. It became more and more unstable. Then we used a Java server to have a stable output. We got rid of that one because we had the ultimate solution in our hands: Flash Player 10 with dynamic audio support.

Last week I was on holidays — no laptop, no internet, no mobile. Just wanted to get fresh eyes, relax a little bit — you name it. Yes, that’s right: I missed the launch of the Flash Player 10.

Today I went into the office and I heard about the news. The released version Flash Player 10 has a completly different behaviour when it comes to dynamic audio. We have been developing with Flash Player 10 for a while and it has been working like a charm.
But without telling anyone Adobe has changed the behaviour of the sound API. I can partially understand the changes which means I do not understand why it worked in the first place but the main point is that it worked. I have also another solution in my mind to keep it the same way as it was before.

We implemented several techniques to ensure a stable audio output. There are still parts in the Flash Player implemented native which we could not circumvent or would take a ridicolous amount of time to get around. Imagine if you set the scale on a big DisplayObjectContainer. The Flash Player itself will take a lot of time to perform this task. Enough time to block the audio stream. Therefore we were pre-calculating the needed audio information. Since the Flash Player or ActionScript is not multi-threaded it was the only option to send the data before performing such an expensive task. This is now impossible. The whole work we did was worthless and the application feels less stable.

Now there is no way for us to fix this issue with the latest changes. The reason why I am so disappointed is because nobody told us. It was probably a small change for the Adobe engineers but it has such a big impact nobody thought about that I ask myself how they deal with responsibility.

What can we do? We can wait. Get some feedback from the engineers and try to get the old version back. Of course we can arm up the MAKE SOME NOISE campaing but then what? Wait another year for Flash Player 11 and see how the API changes with the release? Without being notified?

Adobe, this was a real disappointment. I am not amused.

I am really happy that my work in the next couple of months will be focused on low-level API programming and server communication. I will not write that much ActionScript code — MIDI is about to make it into our tool and I have to write some nice stuff for you Mac folks without a proper Java VM ;)

Related Posts

7 Responses to “David vs. Goliath”


  1. 1 Mike Welsh

    Hi Joa,

    It’s very disappointing that these exciting new features have fallen short of our expectations. Including the audio synthesis, Andre mentioned the funky 3D math ( http://blog.andre-michelle.com/2008/what-is-wrong-with-the-matrix3dtransformvector/ ).

    Hopefully Adobe steps up and releases a quick update for these issues.

  2. 2 joa

    This is not the only issue. I did not include it in the post but did anyone ever tried using Vector3D.XAXIS, YAXIS or ZAXIS? You will get a nice null exception. This is only frustrating.

  3. 3 Johannes

    Yes! It is disappointing. We worked with the FP10 (alpha/beta) as well. My whole project is corrupt now!
    The Vector3D Class has now X_AXIS, Y_ASIX and Z_AXIS (i had to change that too). But, bloody hell, now the Matrix3D Class doesn’t work anymore :( i don’t know why.
    Feel with you! Keep on coding ;-)

  1. 1 Andre Michelle » Blog Archive » FP10 SoundAPI changes
  2. 2 Blog von Jens Franke » Das Arbeiten in einer Welt voller APIs
  3. 3 links for 2008-10-18 at suckmypixel
  4. 4 Hobnox - Audiotool, update | イナヅマtvログ

Leave a Reply






Close
E-mail It