Adobe's John Nack explains lack of PPC support in Soundbooth

Adobe and ppc? no no no!

"Do you think Apple gives a stuff about Power PC ? How about all G5 PCIe owners NEVER have been able to order a video card upgrade ! Contemptable behavior."

The only reason why Mac OS was able to move so quickly from PPC to Intel was because Apple secretly kept their x86 source tree up to date and wrote cross platform code.

Porting legacy / Classic applications such as Photoshop is very difficult. While it's certainly not as simple as clicking a button in xCode, it's significantly easier to design your app to support multiple architectures on new projects such as Sound Booth. Apple also includes several frameworks, such as Accelerate Framework, that help write cross-platform SSE / Velocity engine code on Mac OS X.

Hart Shafer, product manager for audio apps at Adobe, had more to say about this, and unlike John Nack he focused on the technical reasons for the decision as opposed to the philosophical ones:

http://blogs.adobe.com/hartshafer/2006/10/soundbooth_and.html

Chris Randall, who develops Audio Damage's plug-ins (which are cross-platform) chimed in on my own site about the issue and notes that cross-processor development is a LOT harder than people think:

http://createdigitalmusic.com/2006/10/31/adobe-defends-intel-only-mac-release-for-soundbooth/

Audio apps also happen to have a whole lot of the kind of bit operations that make cross-processor compatibility (and performance, incidentally) more challenging.

I think it's funny that people missed the *good* news, which is that the formerly Windows-only Adobe audio team have decided to develop for the Mac. And it's great that, unlike certain companies whose name begins with the letter "A", Adobe's developers are allowed to talk publicly on their blogs. They're reading what we're saying, and we can have a two-way conversation about their software. I think this could have a positive impact on Soundbooth, because it means they're getting feedback on what people want from the app feature-wise BEFORE it ships, instead of after. Adobe's not perfect, just like any developer, but I'm glad they keep these discussions open.

Why would any company in their right mind develop a new software product for a platform that is fading? It can't economically justifiable.

I think there is another possibility to consider that Adobe wouldn't necessary readily admit. Its possible that Soundbooth was a Win32-only development project until the announcement of the Intel switch. Here, Adobe would see the potential market increase and not fatally delay the development roadmap by adding OS X on Intel support.

I think that's total BS. How are they building it? In Xcode, you simply mark a checkbox for PPC and/or Intel. It's that simple.

I can understand Adobe's decision. If nothing else the Endian issue creates great challenges. The G5 series of chips actually dropped some of the functionality for dealing with little-Endian numbers that the G4 had and that caused significant delays in the release of VirtualPC for Mac when the G5 was released.

The advantage of Universal apps is that when Apple decides to release a PPC system again it will just work. An architecture isn't dead until there no code that runs on it. I remember when we had the funeral for x86, ooops don't look now.

This has nothing to do with Big/Little endianness but everything with those audio filtering algorithms being available of the shelf for the x86 architecture (CoolEdit, Audition). This would be a rather costly rewrite for PPC.

I assume that SoundBooth is not using CoCoa but their own inhouse framework since the same sourcecode needs to support both Windows and Mac OS X. This is a much bigger task to implement than supporting Big/Little Endian.

So much for the "little check-box"...