Main

July 02, 2004

On Names and Innovation

As many if not most of the people reading this (generally neglected) essay site knows, Apple has just announced the latest version of their operating system: Mac OS X 10.4, codenamed Tiger. Tiger follows in the footsteps of its predecessors Jaguar (which signed a lot of checks) and Panther (the most successful of their OSs to date, which managed to cash most of the checks Jaguar signed). In what may be the best sign of Mac OS X's maturity, the announcement covered mostly a pack of features few if anyone asked for, and almost no core improvements to the OS. (A major exception to this is in Tiger's improved developer tools -- but that's something for developers and codemongers to be excited about, not end users).

Believe it or not, this is excellent news for Apple. When you reach the point where you start Microsofting (a verb I have just coined that means 'to add superfluous features in an effort to get people with a perfectly usable version of your software to buy a new version they don't really need) your OS, your OS is pretty rock solid. And as a hearty Panther user, I can attest to this. Panther works exceptionally well in all its core areas, and its own superfluous features, while spotty, can be ignored with impunity. (Anyone out there use Filevault? Anyone? I didn't think so.)

However, that leaves us with a glaring issue on Apple's part. They have to continue to appear innovative. They have to continue to seem not only cutting edge, but bleeding edge. Their stock in trade, and their core business -- the sale of computers -- relies entirely on the cachet of using Apples, and that relies on consistently being the kids with the coolest toys in town; the things that make you say "ooooooooooooo..." when you hear about them are what keeps Apple alive.

(This, by the way, is different from Microsoft's stock in trade. Microsoft makes its money on sales of its software and its Operating System. It needs to sell new versions of Windows in massive numbers to keep the dream alive. Apple doesn't sell OS X as anything more than a secondary revenue stream -- they need to keep the Macintosh at the top of the heap, so that they continue selling truckloads of computers. It's an entirely different business model, which is why Microsoft isn't really working to kill Apple off these days. Microsoft likes having Apple around, because Mac users buy Microsoft software.)

The problem with needing to innovate within the Operating System, however, is the complete and utter lack of need to innovate within the operating system. Once your OS is stable and fast, that's pretty much the ballgame, in terms of functionality. This is why the king of Operating Systems is Unix and its derivatives. The world runs on Unix, despite Unix's 'core innovations' being decades old, because Unix is rock solid and fast. Period.

Well, Darwin is pretty damn rock solid now, and the Aqua window and file manager running on top of it has reached a zenith in Panther. Panther is bloody easy to use, it's pretty damn solid, it's fast and it has all the features people expect in an Operating System. Which means innovation is damn hard for Apple now.

Well, if 'To Microsoft' means to add superfluous features to a product to get people to buy it, then the adjective 'Microsoftish' means 'to adopt the veneer of other peoples' innovations in an effort to improve your own product and eliminate competition.' And, following a clear trend started by Sherlock 3 (vs. Watson), Apple has wholly embraced Microsoftish coding practices this year, with the announcement of the most 'innovative' of their new features: Dashboard. A product that on the surface resembles a third party product called Konfabulator.

I say 'on the surface' because their engines are entirely different. Dashboard uses the webcore engine that Apple's developed for Safari, while Konfabulator uses a homegrown engine driven by Javascript. But, when you look at the proposed screenshot for Dashboard and the actual screenshots of Konfabulator, they appear functionally identical.

And... they call their subprograms by the same name: Widgets.

And therein lies the problem.

The idea of small programs that handle tiny tasks is as old as the Macintosh itself. "Desk Accessories" were a part of the original Macintosh, and were the clear spiritual predecessor of widgets. So, no one could legitimately complain that Apple 'stole' these subprograms from Konfabulator. They could argue that Apple is undercutting the innovation from third party developers (and have shown a habit of doing this, by the vastly clearer wholesale theft of Watson's functionality for Sherlock 3, the year after they honored Watson as one of the most innovative products of the year). They could argue that producing something with such a close look and feel was crass at best. But theft? Nah. Doesn't wash.

Unless, of course, Apple were stupid enough to steal Konfabulator's terminology. And 'Widget' is an idiosyncratic term at best -- I doubt I've used that term more than twelve times in my life, except when referring to Konfabulator. It's not an accepted term for a subprogram. American Heritage defines it as:

1. A small mechanical device or control; a gadget.
2. An unnamed or hypothetical manufactured article.
Software isn't hardware -- you might write a controlling program, but that doesn't make it a 'mechanical device.' And I'm sorry, but nothing in Dashboard constitutes a manufactured article. You can hold a manufactured article in your hand without it being enclosed by a computer screen.

So. Language evolves. That's fine and dandy. But when someone uses a term in a new way as an integral part of their software product, you don't get to take a product that does exactly the same thing, use the modified term in exactly the same way, and duck charges of intellectual theft. It doesn't matter when you started developing your product. It doesn't matter what engine you use to produce it. You can't take someone's terminology as your own and then claim you're not copying them.

Especially when you were already accused of doing it once.

This is a stupid move on Apple's part. Had they called Dashboard's components 'gauges' (which fits the dashboard model) or 'gadgets' or 'gizmos' or 'thingamajigs' or 'instruments' or anything like that, people would be sullen about Konfabulator but wouldn't have a real leg to stand on. If they'd called them 'desk accessories,' they could have both trumpeted the return of a classic Macintosh function and clearly demonstrated prior art. But they didn't. They called them 'widgets.'

And so branded Dashboard as a wholly stolen product, now and forever. And sent notice to third party developers that Sherlock 3 wasn't a fluke -- if you want to be sure Apple won't steal your ideas, better patent all of them or (more likely) better not write them for the Apple. Besides, there are more Windows users anyway, right?

And that, if it spreads far enough, kills Apple far more completely than failing to innovate ever would.

On Names and Innovation

As many if not most of the people reading this (generally neglected) essay site knows, Apple has just announced the latest version of their operating system: Mac OS X 10.4, codenamed Tiger. Tiger follows in the footsteps of its predecessors Jaguar (which signed a lot of checks) and Panther (the most successful of their OSs to date, which managed to cash most of the checks Jaguar signed). In what may be the best sign of Mac OS X's maturity, the announcement covered mostly a pack of features few if anyone asked for, and almost no core improvements to the OS. (A major exception to this is in Tiger's improved developer tools -- but that's something for developers and codemongers to be excited about, not end users).

Believe it or not, this is excellent news for Apple. When you reach the point where you start Microsofting (a verb I have just coined that means 'to add superfluous features in an effort to get people with a perfectly usable version of your software to buy a new version they don't really need) your OS, your OS is pretty rock solid. And as a hearty Panther user, I can attest to this. Panther works exceptionally well in all its core areas, and its own superfluous features, while spotty, can be ignored with impunity. (Anyone out there use Filevault? Anyone? I didn't think so.)

However, that leaves us with a glaring issue on Apple's part. They have to continue to appear innovative. They have to continue to seem not only cutting edge, but bleeding edge. Their stock in trade, and their core business -- the sale of computers -- relies entirely on the cachet of using Apples, and that relies on consistently being the kids with the coolest toys in town; the things that make you say "ooooooooooooo..." when you hear about them are what keeps Apple alive.

(This, by the way, is different from Microsoft's stock in trade. Microsoft makes its money on sales of its software and its Operating System. It needs to sell new versions of Windows in massive numbers to keep the dream alive. Apple doesn't sell OS X as anything more than a secondary revenue stream -- they need to keep the Macintosh at the top of the heap, so that they continue selling truckloads of computers. It's an entirely different business model, which is why Microsoft isn't really working to kill Apple off these days. Microsoft likes having Apple around, because Mac users buy Microsoft software.)

The problem with needing to innovate within the Operating System, however, is the complete and utter lack of need to innovate within the operating system. Once your OS is stable and fast, that's pretty much the ballgame, in terms of functionality. This is why the king of Operating Systems is Unix and its derivatives. The world runs on Unix, despite Unix's 'core innovations' being decades old, because Unix is rock solid and fast. Period.

Well, Darwin is pretty damn rock solid now, and the Aqua window and file manager running on top of it has reached a zenith in Panther. Panther is bloody easy to use, it's pretty damn solid, it's fast and it has all the features people expect in an Operating System. Which means innovation is damn hard for Apple now.

Well, if 'To Microsoft' means to add superfluous features to a product to get people to buy it, then the adjective 'Microsoftish' means 'to adopt the veneer of other peoples' innovations in an effort to improve your own product and eliminate competition.' And, following a clear trend started by Sherlock 3 (vs. Watson), Apple has wholly embraced Microsoftish coding practices this year, with the announcement of the most 'innovative' of their new features: Dashboard. A product that on the surface resembles a third party product called Konfabulator.

I say 'on the surface' because their engines are entirely different. Dashboard uses the webcore engine that Apple's developed for Safari, while Konfabulator uses a homegrown engine driven by Javascript. But, when you look at the proposed screenshot for Dashboard and the actual screenshots of Konfabulator, they appear functionally identical.

And... they call their subprograms by the same name: Widgets.

And therein lies the problem.

The idea of small programs that handle tiny tasks is as old as the Macintosh itself. "Desk Accessories" were a part of the original Macintosh, and were the clear spiritual predecessor of widgets. So, no one could legitimately complain that Apple 'stole' these subprograms from Konfabulator. They could argue that Apple is undercutting the innovation from third party developers (and have shown a habit of doing this, by the vastly clearer wholesale theft of Watson's functionality for Sherlock 3, the year after they honored Watson as one of the most innovative products of the year). They could argue that producing something with such a close look and feel was crass at best. But theft? Nah. Doesn't wash.

Unless, of course, Apple were stupid enough to steal Konfabulator's terminology. And 'Widget' is an idiosyncratic term at best -- I doubt I've used that term more than twelve times in my life, except when referring to Konfabulator. It's not an accepted term for a subprogram. American Heritage defines it as:

1. A small mechanical device or control; a gadget.
2. An unnamed or hypothetical manufactured article.
Software isn't hardware -- you might write a controlling program, but that doesn't make it a 'mechanical device.' And I'm sorry, but nothing in Dashboard constitutes a manufactured article. You can hold a manufactured article in your hand without it being enclosed by a computer screen.

So. Language evolves. That's fine and dandy. But when someone uses a term in a new way as an integral part of their software product, you don't get to take a product that does exactly the same thing, use the modified term in exactly the same way, and duck charges of intellectual theft. It doesn't matter when you started developing your product. It doesn't matter what engine you use to produce it. You can't take someone's terminology as your own and then claim you're not copying them.

Especially when you were already accused of doing it once.

This is a stupid move on Apple's part. Had they called Dashboard's components 'gauges' (which fits the dashboard model) or 'gadgets' or 'gizmos' or 'thingamajigs' or 'instruments' or anything like that, people would be sullen about Konfabulator but wouldn't have a real leg to stand on. If they'd called them 'desk accessories,' they could have both trumpeted the return of a classic Macintosh function and clearly demonstrated prior art. But they didn't. They called them 'widgets.'

And so branded Dashboard as a wholly stolen product, now and forever. And sent notice to third party developers that Sherlock 3 wasn't a fluke -- if you want to be sure Apple won't steal your ideas, better patent all of them or (more likely) better not write them for the Apple. Besides, there are more Windows users anyway, right?

And that, if it spreads far enough, kills Apple far more completely than failing to innovate ever would.