| Thursday, December 12, 2002 |
00:13 - Longhorn Goes Maverick
http://slate.msn.com/?id=2075219
|
(top)  |
I've been mulling over this Slate article by Steven Johnson for a couple of days now. It's about OS X, Microsoft's "Longhorn" initiative, and the "media-based interface" metaphor that I've been talking about for a while now, here in this blog, and in earlier articles written before iPhoto and other iApps cemented the idea as the real direction that Apple's UI design philosophy is taking.
Put simply, the idea of what I'm calling the media-based interface is one in which every different kind of data-- each of which is inherently different from the rest, be it MP3s, photos, movie clips, documents, etc-- is handled by a specialized interface that allows the user to work with the data using the data's own intrinsic metaphors, rather than the artificial metaphors imposed on it by computers. For instance, and at the risk of sounding like a broken record ('scuse me-- sounding like, uh, an MP3 player with an obscure software bug that makes it skip backward and play the same ten seconds over and over again), instead of having to think about your music and photos as "files" and "folders", you think about your music as songs and artists and albums, and your photos as rolls and albums and direct visual thumbnail representations. It's my belief that Apple intends iTunes to be the standard interface for listening to digital music, rather than navigating through the filesystem and juggling folder paths and long filenames with ".mp3" at the end. You can still drill down and use that more austere and ill-suited metaphor to deal with the song files, but it's far more efficient and intuitive to work with the data through a software layer that gives you a different, more well-defined and natural set of grips to hold on to.
Now, what I didn't realize, though, was that according to Bill Gates, Microsoft knows about the media-based interface idea, but wants no part of it. In developing Longhorn, their new upcoming SQL-based filesystem, their design goal will be to dispense with all the different ways of handling different kinds of data. It's in the interest of simplicity, says Bill.
The ultimate goal is to prevent you from having to learn entire new programs to interact with your mail messages, your contacts, and your home movies—to ensure that each data type doesn't become the exclusive province of a specific application. (To take an example from the iApps, iPhoto is great at organizing your photos, but it's useless if you're trying to figure out which snapshot you e-mailed to your mother last week.) Think about searching for text strings in four different contexts: in a Word document, in your inbox, on the Web, and in your hard drive. There are four distinct search tools for those four tasks, each with its own interface, each "belonging" to a different application. But in each case, you're just searching for text. Why use a Swiss army knife when one blade will do? As Bill Gates put it on Charlie Rose last month, "Right now when you use Windows, the way that you step through your photos, the way you step through your music, the way you step through e-mail or files, they're all different. You have to learn different user interfaces, different search commands. ... The idea of Longhorn is to have one approach, one set of commands that work for everything, including all of those things. And so the number of concepts you have to learn is dramatically less."
Johnson approaches Longhorn from a skeptical angle, and frames it with analysis of OS X and the iApps, of which he is clearly a fan. The way he talks about it, he's unconvinced that this is a good move for Microsoft, and as it's described here, I'd have to agree. I don't doubt that there's more to Longhorn than we're seeing here, and I'd want to see a more finished prototype of it before I passed judgment. (I might be misunderstanding the whole thing; the examples Gates cites talk more about a unified text-search function than a genericized navigation metaphor.)
But from what I can see here, it seems almost as though Microsoft wants quite consciously to be seen doing something "un-Apple"-- making a decision on one of the most interesting philosophical design choices to come along in a decade, and taking the opposite tack to what Apple does. If Apple thinks people are going to want to manage each kind of data in a way that's native to that kind of data, Microsoft is betting that that will end up being too complex for users to want to deal with. They're gambling on what amounts to a reversion to DOS, then-- a more advanced iteration, yes, but something very similar to the DOS mentality: a unified and undifferentiated interface layer through which you can access every kind of data using the same filesystem tools. The contention is that the fewer steps the user has to memorize to access "data" of any kind, the easier it will be to handle specific data, from a common underlying access method.
What I think this decision lacks is an understanding of the fact that some steps toward a goal are more intuitive than others. It can indeed be more difficult for someone to understand three steps toward a goal if each of those steps makes him think about an unfamiliar, unnatural metaphor, than for him to internalize ten steps which make themselves obvious from the intrinsic nature of the data you're working with.
In Windows, to play MP3s, you navigate through folders, find the files you want to listen to (however you've chosen to organize them), and double-click to open them in your MP3 player, which immediately plays them. On the Mac, you first have to find and open iTunes; thereafter, you work with the music on the music's own terms, using the music's own intrinsic attributes, which are intuitively obvious within minutes of a user seeing the program for the first time.
The difference lies in the same distinction behind learning your times tables by rote, and figuring out in that flash of fifth-grade insight just what multiplication is.
If you have to think about files and folders, or whatever other interface layer Longhorn puts up in front of them, in order to get to your music or e-mail or photos... then at best it's going to be something learned by rote, a procedure that people will have to write down on yellow sticky notes and attach to their monitors so they don't forget exactly where and how to move the mouse each time they want to look at the pictures from summer vacation.
Whereas in a media-based interface, there is no unnatural metaphor to remember. The steps toward an organizational or operational goal require no memorization and no abstraction. The idea is that the only big step the user has to remember is the step that opens up the specific interface in question for each type of data. Hence the unmistakable "music" icon of iTunes. It draws the eye, and conveys the idea that "To listen to music, click on me. After that, controlling music will be intuitive, using the metaphors that are familiar to anybody who has CDs."
Now, I'm not advocating the media-based metaphor for everything; I suspect that in the future there will be a mix of objects presented in their native filesystem structure and objects viewed through a specialized interface, as there is today. Windows, for instance, groups its applications into a specialized interface-- the Start menu; whereas the Mac, while it has the Dock for shortcuts, primarily uses the filesystem itself for accessing applications, which are just objects you can freely move about the system. In that respect, Windows and the Mac have what seem to be the opposite roles from what you'd expect, given their present attitudes toward multimedia data. It's anybody's guess where these trends will lead those aspects of the respective platforms.
But more interesting to me is that if I'm reading Gates' intentions the way Johnson is, this represents the first major occasion in a long time that Microsoft has chosen to diverge from the Mac in UI philosophy. For the past two decades the two companies' platforms have been converging; Windows has been getting more and more Mac-like, and Apple's OS has been changing only incrementally, OS X notwithstanding. But now Apple is embracing the media-based metaphor, and Microsoft is actively committing to the idea that for everyday browsing of data, focusing on individual media types is the wrong way to go; they think people will benefit more from a unified multi-function browser that collapses the differences between data types and allows them all to be viewed in the same context. That's one of the benefits an SQL-based filesystem can bring. (It isn't a new idea, by the way; Linux users have been employing MySQL as a filesystem for years now.) Files can be tracked by unique IDs as well as having arbitrary amounts of associated meta-data which need not be displayed when all you're doing is browsing. From a technical standpoint, Longhorn would probably be a pretty neat idea.
But I'm not convinced. Experience has made me a believer in the media-based interface, and if nothing else, I find it fun to be able to work with my data in a way that doesn't require me to think about ill-suited software metaphors that fit the data about as well as Cinderella's glass slipper would fit on my size-12 pseudopods. I don't have to fight the data; I don't have to memorize metaphorical tricks or shift my brain's gears to adjust to the way the software thinks I should be treating my media files. I don't have to deal with wizards or think about the Web. I don't have to wonder why my favorite songs or last week's movie clips are being represented as little pieces of dog-eared paper stuffed into yellow poster-board folders. All I have to think about are those attributes that make sense in the context of the data itself.
It could well be that Longhorn will be just a substrate, allowing more filesystem flexibility while "media-based" apps like iTunes sit on top of it and filter the data for more intuitive presentation; though from Gates' comments, it appears that it won't be Microsoft providing that functionality. I suspect that Gates might honestly think Longhorn would be a benefit to consumers and a boon to usability; I think it would be a mistake. But I'm honestly excited to think that for the first time in a long while, there may in fact be a new fundamental philosophical differentiating point between Windows and the Mac, something new to base discussions of the truth-and-beauty of usability upon. If Longhorn makes Windows harder to use, and people start relying on third-party apps to provide basic data-browsing functionality, then the Mac will have had a nice new advantage handed to it on a silver platter-- after years of Windows becoming more and more "good enough" and nullifying the major reasons to use a Mac, suddenly a whole lot more people will be a whole lot more frustrated with their PCs and in need of a better solution. On the other hand, if Longhorn's metaphors turn out to make computing easier than it is today, even more so than the Mac would-- then it will have been the first insightful, original UI initiative Microsoft has come up with on their own since... well, since Microsoft Bob.
And I'd say it does stand a better chance of success than Bob ever did. (Well, duh.) But I still think Apple is closer to having the right idea.
|
|
