An estimable gentleman and former college roommate of mine recently upgraded from a 2008 MacBook Pro running Snow Leopard to a new model model running Lion. He and I spent a good chunk of our time at Middlebury playing StarCraft with our friends. Although I've not played much since, he had stuck with it until he tried to launch it on his new computer, only to find that it - along with other classic games like Diablo - was no longer compatible with OS X.
StarCraft was made available for Mac OS 8 and 9 in 1999, and continued to be supported through OS X. It was made for PowerPC computers, and when the Intel switch happened, Blizzard relied on Rosetta for continued compatibility. Now Rosetta is gone, and Blizzard has announced it will not be supporting StarCraft, Diablo, and other games on Lion or any future version of OS X. My friend was understandably miffed - as an end-user, he just wants his stuff to work. His plight got me thinking: whose responsibility is it to ensure that classic software remains compatible with current and future operating systems, the OS maker or the software maker?
I think there are legitimate arguments to be made for both sides. In many cases, software is the major incentive for purchasing a computer. That's not to say an OS needs to be compatible with every piece of software ever written for every version of it, but it is to say that there is a particular group of applications that are either held in such high esteem or are so seemingly indispensable to the experience of using the computer that they must be supported. The canonical example is Office. Anyone who thinks their computer (and for the sake of convenience, I'll use "computer" to refer to desktops or laptops, not mobile devices like the iPhone or iPad, which should be considered computers but are outside of the scope of this particular post on legacy software because they are less than five years old) is going to sell to the masses without being able to run a version of Office is living in a fantasyland. StarCraft doesn't exactly occupy the same space in public consciousness as Office, but the fact remains that it is one of the classic computer games of all time, and its longevity is a testament to how users connected with it. If the goal of a home computer is to best serve its users' needs, and if those users decide that StarCraft is an indispensable part of their computer use, then one could make the argument that it is the responsibility of the OS maker - in this case, Apple - to make sure that the software just works.
The counterargument is that it is the OS maker's responsibility above all to move its platform forward for the benefit of its users, and that holding back development to cater to a few applications ultimately hurts more than not having them. By moving forward, the OS maker can reveal to users that applications they thought were indispensable actually don't matter that much. A great example is how Apple moved away from Internet Explorer in 2003. At that point, with no Firefox and no Chrome, IE was the web browser, and to try to sell a system without it was crazy. But Apple moved on, built a compelling product in Safari to break IE's hold over Mac users, and fired the first shot of the current generation of browser wars. Now you can't find a Mac user who would touch IE, even if he could. The changes caused people to reevaluate what they needed in a product, and now demand for IE on Macs is nonexistent. Should Apple have worked overtime to make sure future version of OS X were IE compatible? Doesn't appear that way. So why should it bow to anyone else?
Those are broad arguments. In this case specifically, I lay the blame at Blizzard's feet for a couple of reasons. First, the Intel switch was announced seven years ago. That's seven years Blizzard had to create a version of StarCraft that could run natively on Intel Macs, and it didn't do it. Instead it bet on Rosetta being a part of OS X for as long as people wanted to play StarCraft - a silly bet, given Apple's history of eliminating stopgap technologies. Now Rosetta's gone, and having missed the chance to eliminate this problem years ago, Blizzard is just moving on. It is within Blizzard's power to correct this, but it won't.
Second, a company like Blizzard, mighty though it may be in its own sphere, simply can't sit at the table with Apple as an equal. All it would take is a little bit of self-awareness to see that this is an asymmetric relationship, and to understand that in Apple's world, you either play by its rules or you don't play at all. Apple's rules state that you keep up with its roadmap (or at least within seven years of it) if you want access to its extremely lucrative user base, and if you're not willing to make that happen, then you're going to deal with the consequences. Apple is dealing from a position of strength right now, Blizzard isn't, and they need to adjust their strategy accordingly.
That's of little comfort to my friend, I'm sure, but with OS X moving to an annual release cycle, legacy issues are something we're almost certainly going to have to deal with going forward.