Date: Fri Oct 26, 2001 07:00:28 PM US/Central Subject: iPod redux hey gang - below is a rant i just sent to a writer at The Register in regard to a story on Apple's new iPod. story link: http://www.theregister.co.uk/content/39/22463.html i figured some of you might find it interesting. personally, i think the iPod represents some darned good thinking about real-world issues of affordance and usability. ------------------------------------------------------------------------ i just read your artice in the Register, and while you hit some decent points regarding the iPod as a portable music device, you (and the rest of the tech press) have still missed the point dramatically, IMO. the iPod isn't an MP3 player. it's a network computing device. you remember network computing.. it was the trendy idea du jour back around 1997, when web browsers and Java were supposed to give us a universal, OS-independent applications platform. 'pure' network computing assumed that all a user's information would live in a central repository, and be delivered to workstation X via the network. while this was a nice idea in theory, such systems are susceptible to bandwidth limits, network failures, identity theft, and snooping by whoever owns the repository. an alternative idea was to give each user a 'card' that would hold all that user's information, and which would connect seamlessly to any computer in the network. at the time, that second idea had two fatal flaws, one technical, one social: you couldn't pack enough data into a small enough device, and even if you could, nobody would want to carry the darned thing around. it's *way* too easy for people to forget a device that will only be useful after they get to the office and connect it to a computer. Moore's Law has solved the first problem, and Apple, through some decidedly clever thinking, has solved the second. only the barest handful of dedicated geeks would ever buy and carry hot-swappable mini-drives, but millions of people buy and carry portable music devices. there's no physical difference.. one deck-of-cards-sized box of microelectronics is pretty much like another.. but there's a conceptual barrier that's very real. Apple has broken down that conceptual barrier by slapping an MP3 player onto its portable, hot-swappable data storage device. remember the central tenet of network computing: only one component actually makes a computer 'my computer': the hard drive. nor does the drive itself matter per se. all that matters is the data. you can even strip away all the data that comes as part of the default OS installation, leaving only the data that makes up my personal workspace. 'my computer' is a purely conceptual device, and with a little work, can be made independent of any physical device. Mac OSX collects all the data that makes up 'my computer'.. all my files, all my application preferences, all my global system settings.. into a single file tree. Apple's NetInfo utility tells a network how to find a particular user's home directory. if i put my home directory ('my computer') on an iPod, i can carry it to any machine in an OSX network, log in, and make that computer 'my computer'. presto.. network computing. and i even have something to listen to while i'm walking from one machine to the next. in addition, the iPod is a networking medium, along the lines of the old adage, "never underestimate the bandwidth of a station wagon full of tapes." even if you only devote 2.5 GB of space to data storage, the iPod gives good old sneakernet a bandwidth that compares favorably to 100Mb/s ethernet to a range of one mile (how many other ways can you put 2.5 gigs of data onto a machine a mile away in less than 30 minutes?). 'pure' network computing is centralized, but if the blooming of P2P services over the last few years shows anything, it's that distributed systems work in today's environment. the hallmark of P2P systems is their ability to carry data between any two computers at the edge of today's network, with little or no reliance on things in the center. iPod-sneakernet gives users a new transmission medium, thus freeing computers at the edge of the network from the centralized infrastructure (cables, routers, etc) we take for granted today. the iPod is far more than Yet Another MP3 Player. unfortunately, you can't convince people that something as mundane-sounding as 'a high-capacity portable storage device' is Insanely Great until you've shown them what it can do. so Apple threw in the MP3 player so people would buy the damned things. once someone does have one, though, the iPod turns MP3s into a gateway drug for network computing. that 1000 MP3s -v- 5GB data storage capacity is a textbook example of marginal utility balancing to anyone with even a smattering of basic economics.. if i already have 999 MP3s, would i be willing to give up one more in exchange for 5MB of data storage? how about 998 MP3s and 10MB of storage? how about 3GB of storage and 400 MP3s? the exact point of equilibrium doesn't matter. what matters is that users end up thinking about how much data storage they want, and what they can do with it. think outside the box. stop assuming that the statement 'this device plays MP3s' is logically equivalent to 'this device is just another MP3 player' and really *think* about what your life would be like if you had 5 gigs of data storage in your pocket. now take another look at an iPod, and tell us what you see.. mike stone ADDENDUM (with thanks to the moderators, who graciously asked me to throw in something web-related in order to bring this post in line with the CHI-WEB topic guidelines): what does the iPod have to do with the web? frankly, i have no idea. i think it will be interesting to find out, though. one of the web's great strengths is its ability to migrate to new environments. evolutionary biologists call it 'radiative dispersion'. species good at that tend to flourish. network computing will be a new environment. iPod-sneakernet networking will be a new environment. the web will work its way into both of those niches, because there's no good reason why it shouldn't. i think it's safe to expect some cross-pollination between today's web and the new environments. network computing will never be purely distributed, because there are some cases where a central repository is just plain useful. the web will be a natural candidate for providing access to such repositories. at the same time, i think that the web browser's days as a universal network interface may be numbered. Apple offers a service called iDisk, which mounts a WebDAV-shared directory as if it were a CD. the OSX programming environment contains OS-level support for loading information specified by URL.. and there's also a built-in XML parser. Apple is making it easier and easier for computers to share information via HTTP, which isn't quite the same as what we call 'the web' today. today's 'web' means 'web browser'.. a jack-of-all-trades (and master of none) UI shell. i think we're going to see a rise in specialized applications that make use of the same underlying protocols, but which aren't so closely tied to whatever mutant version of HTML is the de facto standard. i think that the-web-as-HTTP is more than ripe for specialization, because the limits of the-web-as-HTML are more than well known. i think we're going to see the same old pieces rearranged in ways that shatter our current assumptions of how 'the web' should behave, and that the resulting chaos will be good for everyone. i have absolutely no idea what the killer app of a networked-computing world will be, though. nobody expected the web, and nobody expected the weblog. if i were to make a rough guess, i'd say that an iPod-sneakernet offers an interesting niche for some sort of pony-express distributed data network that vaguely resembles Usenet, but deals in files too large for today's networks to handle. only time will tell, though. mike .