"No one said anything about CS3 being "not supported" on Snow Leopard. The plan, however, is not to take resources away from other efforts (e.g. porting Photoshop to Cocoa) in order to modify 2.5-year-old software in response to changes Apple makes in the OS foundation."
"Q. Will older versions of Adobe creative software—such as Adobe Creative Suite 3 or Macromedia® Studio 8 software—support Mac OS X Snow Leopard (v10.6)?
A. Older versions of Adobe creative software were not included in our testing efforts. While older Adobe and Macromedia applications may install and run on Mac OS X Snow Leopard (v10.6), they were designed, tested, and released to the public several years before this new operating system became available. You may therefore experience a variety of installation, stability, and reliability issues for which there is no resolution. Older versions of our creative software will not be updated to support Mac OS X Snow Leopard (v10.6)."
Yep, because it's not like Apple gave you a 9-year heads-up on having to port your app to Cocoa or anything. In fact, Adobe made no move towards porting until Apple took out binary-level support for 64-bit Carbon in Leopard.
These guys have no excuse to complain about taking resources away. They've been practicing shoddy development for years - consistently building on top of deprecated API.
Yet in the end it must be done, and the earlier you do it the cheaper it gets (both in terms of good/bad press and actual development cost). Instead of being proactive and investing in the port over several releases, they are now in the position of 11th-houring it after binary support for Carbon APIs have been removed.
This isn't some idle, idealistic call for good code - deprecated API will lose binary support at some point, it's entirely on Adobe's head that this has gotten to this point at all.
I've worked at companies similar to Adobe's model (e.g. professional tools, regular releases to keep the cash monies flowing), and the death march for "bullet-point-able" features is relentless. The constant mad scramble for new features to justify the release cycle, along with management's complete lack of willingness to invest in rudimentary code maintenance, results in a code base that simply becomes unwieldy over time.
Obviously I do not work for Adobe, but it's most probable that their current Cocoa porting efforts will end up far more expensive than if they had start the effort much earlier, and it certainly would present a lot less frustration to their customers now.
Seems like a win for everyone, all you need to do is drill into the heads of upper management that regular code maintenance is an absolute must to the long-term health of your business.
I don't think Adobe was lazy in terms of the Cocoa port. From what I can tell they thought they could pull a quick one on Apple like they constantly do to Microsoft - i.e. use their ginormous size and market influence to keep any deprecated API in binaries. It's a "we're too big for you to screw us" mentality, and I'm glad Apple stuck it to them. One of MS's most fundamental problems is that they can't remove API that's been deprecated for a decade for more - they won't just kill old software, they will kill lots of new releases also!
Not to mention, unless Adobe has been on a hiring binge or loading up on temps (never a good idea), the next release of Photoshop will be light on new features and heavy on invisible elements like the Cocoa port. So they threw caution into the wind in favor of more shippable, sellable features, and now they're stuck on an entire release cycle without.