Tuesday, January 08, 2008

The Datastick. Again.

Note: the datastick is a device I first encountered in Colorvision by Ron Cobb. While working on production design for Alien, Cobb assumed the crew of the Nostromo would carry around a device that was a combination flashlight, audio/video recorder, and computer. I incorporated such a device (along with standardized mass storage) into my science fiction setting. Reality has far exceeded any of this, but -- at least since the Newton -- nothing quite like the Datastick has yet emerged, even though it makes a huge amount of sense.

The 2008 MacWorld Keynote is fast approaching and of course there are plenty of predictions out there, along with John Siracusa's keynote bingo. So I thought I'd write down some of my thoughts -- more wishful thinking than predictions -- as to what Apple might have in store for us this year (if not on January 15th).

Everyone knows Apple will release a 3G iPhone sometime this year, so that's hardly worth mentioning. The pundits are pretty confident Apple will release an ultralight flash-based tablet and/or notebook, which I think may be wishful thinking. There's also a patent-filing-based rumor of a new dockable liquid-cooled MacBook, which sounds interesting but unless it's actually not quite what it seems is probably a little nuts. Who knows?

I think that the key factor behind every major Apple announcement since 2001 is convergence. The problem with AppleTV isn't so much what it does or doesn't do, so much as that it's another damn thing. At least a Sony PS3 can play DVDs and blu-ray disks (your library of blu-ray disks is huge, right?). Again, as with cell phones, the key is to reduce clutter, cables, and complexity.

The problem with the AppleTV is that it is missing obvious functionality, including (a) a DVD player, and (b) a DVR. Oh and once you add a DVD player, a DVD-R seems like an obvious next step. Each of these functions would add very little to the bottom-line price of the product, but enormously to its desirability. (You can currently buy brand-name DVD-Rs for well under $200. I have one, and boy do I wish I had one designed by Apple. Of course, Apple already sells an AppleTV with a DVD, and even with a DVD-R -- it's called a Mac Mini -- but it's kind of a major price hike and it lacks HDMI output.)

Here are two major convergence points Apple is poised to exploit, and I hope we'll see announcements accordingly.

The Truly Personal Computer: the Datastick

Here are several devices Apple's customers pretty much all own and wish could be converged: phone, notebook computer, desktop computer, iPod. Even if a customer only has a phone and laptop, they probably have a bunch of peripheral crap they do not carry with them, so that -- in effect -- their notebook becomes a desktop by way of a bunch of tangled cables and hubs.

Apple will (I hope) release a convergence device that replaces all these things, or at least absorbs them. Think of a notebook with phone circuitry and bluetooth support that can dock into an iMac-like display. Fundamentally, the overlap between phone, notebook, and desktop computer is so great that you're currently buying three devices and simply swapping data between them. And a lot of us have two or more desktop computers (one at home, the other at work, for instance).

Personal Server: the Hub

There's a huge demand for a modular gamers' Mac, but -- as I and others have noted -- anything too good in this category would probably kill -- or seriously undermine -- the Mac Pro market. There is a point at which killing the Mac Pro market would make a lot of sense, however, and this dovetails with Apple's overall strategy (since the second coming of Steve Jobs) and that is to make a lot of money from high volume, high margin, low cost products (like iPods) rather than a rather smaller amount of money from low volume, high margin, high cost products (like Mac Pros).

Consider this next time you're in Wal-Mart: to make a 5% margin on a $400 desktop computer, Dell has to ship a huge box to Wal-Mart and that box occupies a huge amount of shelf-space. In that same store, Apple is making 20% or more margins on iPod nanos that take up about as much space as a pack of 8 AA batteries, or a box of Zantac ($10 products sold on fairly low margins).

If Apple can converge a bunch of devices into a single, very compelling consumer device, sell it at a reasonable (slightly high) price, and make a solid profit, why that would be pretty amazing, no?

Here are a few devices Apple could converge into an xMac that would sell enough units at a high enough margin to justify gutting the Mac Pro market:

  • Windows PC

  • DOS PC

  • Macintosh

  • DVR / AppleTV

  • Console Gaming Device(s)

  • DVD Player / Recorder

  • MediaCenter / Digital Hub

If I were Apple I'd try something like this: build a big Mac Mini or a small Mac Pro with very strong onboard video as a BTO option, Cablecard support, HDMI, etc., a DVD-R, lots of RAM, fast hard disk, slots. This device has a standard wireless games controller as an option -- ideally, it's a shameless ripoff of the PS2 controller, and/or possibly the Wii controller.

This hypothetical product offers TiVo-like simplicity for timeshifting TV shows (but you need to buy a .Mac subscription -- all of a sudden, .Mac looks incredibly compelling since it's cheaper than a TiVo subscription and does so much more) but with the added bonus that .mac will let you watch stuff on your Mac DVR from your hotel room when on the road (and allow you to modify your season passes, etc.), you can effortlessly sync programs to your iPod, iPhone, iDatastick, etc., and you can (eventually -- when the lawyers have done their dirty deeds) burn your favorite shows to DVD.

This new product also comes with a version of WINE optimized to play Windows games. It could easily be bundled with a couple of games that are known to work (e.g. The Sims). (The WINE component is already announced, and the necessary hooks for Leopard to support this have been revealed to already be in place.) Of course, you still have BootCamp for full Windows (and Linux) compatibility.

For bonus points, Apple could develop and include a PS1 and PS2 emulator (we know that Connectix knocked together a PS1 emulator in a few weeks and survived Sony's lawsuit). If they were really clever, they could scale the graphics resolution so that PS1/PS2 games actually run at full resolution (something Sony could have done pretty easily with the PS2 and PS3, but chose not to for obvious, if stupid, reasons). Note that Apple doesn't have to do this in spite of Sony -- they could cross-license OS X to Sony for consumer electronics devices and get PS1/PS2 (and more?) support for OS X in return.

For more bonus points, Apple could include FreeDOS in a Window to run your old DOS games. In any event, the Open Source community could easily produce something that bundled VirtualBox and FreeDOS into a legacy DOS games platform along the lines of MAME.


My original datastick concept missed one key technology -- ubiquitous networking. The problem with the original datastick is that if you lose it you're seriously screwed. But if the datastick is really just a local point-of-presence for your data store (which, ideally but not yet practically, is redundantly stored in the "Cloud") then that problem (and several others) go away. In this case, the xMac is your "base station" where your main data (and, unfortunately, your backups) reside, while your MacBook/iPhone/iTablet ... the thing I'm calling the Datastick ... is essentially your portable client.

You do want two computing devices (Macs) and not a dock. You do not want your base station to have to access your main storage hub via the net -- not just yet, anyway. And if you only have one computer and it docks when you're home, then what's going to talk to your 2TB of local storage when you're on the road? What's going to record Scrubs for you when you're on the road, and what's going to convert your near-DVD-quality video library to low-bandwidth streaming video on the fly so you can watch it in your hotel room?

Similarly, it's great to be able to grab photos or video off a camera and do rough edits on location. Wouldn't it be even greater if you could get all your data onto your main storage system from your hotel room?

So, each device is amazingly compelling on its own. One replaces pretty much every device you need to carry around with you and recharge, as well as your office computer, while the other keeps all your data in one place, backs it up, and lets you access it from anywhere. And all the groundwork is in place.

Let's see what we see on January 15th.

The Wire

I used to pay for HBO but I found its scheduling infuriating and ended up never watching much on it (even The Sopranos despite it being, at the time, one of my favorite shows). So, I come to The Wire very late.

My favorite books are either fiction with a lot of informative content, or non-fiction written in a great narrative style. So, some of my favorite books include Shelby Foote's The Civil War: A Narrative and David Simon's Homicide: A Year on the Killing Streets. The former is a work of art, but the latter is simply fascinating, especially if you've ever been a lover of crime fiction.

Only on HBO?

Much has been said of The Wire being (a) the greatest TV show ever made, and (b) something that could only be done (in the US, anyway) by HBO. It's worth noting that Homicide: Life on the Streets (based on Simon's book) was actually very similar in both theme and quality to The Wire lacking only in the bad language and explicit sex department. (And even on network TV it managed to have a bisexual detective who had sex in a coffin.) The big difference (format-wise) between Homicide and The Wire was that Homicide largely retained episodic foreground plots (in addition to large arcs) as a method for drawing viewers along, but that was more a consequence of the evolution of TV as a medium (TV viewers have been taught to follow arc plots, slowly, over the last 25-odd years) and subject matter (homicides vs. narcotics, at least initially) than network limitation.

Both Homicide and The Wire required viewers to pay attention. A single chance remark in one episode might have major ramifications a season later.

Even with strong competition from Homicide, I would agree that The Wire is the best TV show I have ever seen. It seems to me that the chief "advantage" of being an HBO show (aside from HBO's lack of creative interference and willingness to green-light a show with mediocre ratings) is bad language and explicit sex, neither of which make The Wire a better show than Homicide. The Wire's big advantages are that it simply deals with broader subject matter and has taken the "training wheels" of episodic foreground plot out of the format, allowing more time to deal with minutiae. Heroes, on NBC, does the same thing, but just happens to be infantile (if, briefly, enjoyable).

The Cops. The Crooks. The Big Rich.

As an aside, it's always depressing to me that -- at least for drama -- the crime-driven story dominates television so thoroughly, and that the best TV shows ever made have almost all been cop/lawyer shows:

  • The Wire

  • The Sopranos

  • Homicide: Life on the Streets

  • Murder One (first season)

  • Hill Street Blues

There are a few other shows that might contend for greatness that don't quite fit the crime drama mould ... but they're generally just very good Soap Operas (e.g. thirtysomething or ER). I guess there's always the new Battlestar Galactica.

Why Watch the Wire?

It's very rare for writers (and The Wire is fundamentally a literary work) to grapple with how the modern world really works. Stories tend to center on characters, generally a small number of characters, and it follows that the actual way in which things happen has to be compressed into something that is apparent to those characters. A typical episode of Law & Order, for example, focuses on perhaps seven people (the two cops, the two prosecutors, the DA, the defendant, and the defendant's lawyer) and has them do everything of significance in the case, from interviewing key witnesses to figuring out obscure logical flaws in the defendant's alibi.

Almost anything in the modern world is done by an army of characters who often don't know or have anything directly to do one another. The Devil is in the details, and the details are often minor characters who perform vital tasks and are composited, merged, or ignored in the interest of brevity, clarity, and pace. Oh, and don't forget we need to make the heroes look good.

Anyone who has seen the movie South Pacific but who has not read Michener's book will probably be amazed to discover that the book was a grimly detailed procedural account of the amphibious invasion of a fictional Pacific island during WWII, told by attempting to perform a "vertical slice" of the action. Reading the book, we see how decisions made at the highest levels of command lead, for example, to decisions as to how many and what type of bandages to provision, and, ultimately, to whether individual soldiers live or die on the beach.

Similarly, in Richard Powers's (in my opinion) finest novel, Gain, the author performs stunning tour-de-force -- I'd call it the written equivalent of a "frozen pan" (as in "The Matrix") or the cgi zoom-out at the beginning of the movie "Contact" -- when, towards the end of the story, someone in a hospital room takes out a disposable camera and he suddenly shifts and describes how that camera was manufactured -- from the way the paper for the box is chosen so as to print the special ink that all products of that brand use to the camera being discovered, forgotten, days later by an orderly and thrown in a waste basket.

The Wire is, pretty much, nothing less ambitious than an attempt to depict such an instantaneous cross-section of a modern US city in decline. I imagine that watching it all at once (which I will do when the final season ships on DVD) will be quite something, although if you wait until you can see all five at once, it may be shattering. It's not a pretty picture, but it's a picture very much worth looking at.