The Long Tail

Friday Fanboy returns: RC electric planes

[Note: Friday Fanboy was a series I started a couple years ago to gush about something I use and like once a week. I can't remember why I stopped, but it was fun so I'll sporadically resume. File it all under "off topic"]

When I was an early teenager and all the other guys were playing sports, dating girls and otherwise being cool, I was into flying radio controlled model airplanes. Actually, I wasn't even cool enough for that. I was into building model airplanes. I was way too much of a wuss to actually attempt to fly them (the inevitable crash and destruction of months of work seemed like a poor bargain, which was a rare moment of sanity for the teenage me).

This was the late 1970s, and the big innovation in RC modeling at the time was small microchip-driven radio and servo components and heat-shrink plastic film for wing covering. I still remember building my first radio control transmitter and receiver, which was perhaps the closest I'll ever get to Make-level competence with a soldering iron and circuit tester.

Fast forward thirty years, and now I have kids of my own, which is all the excuse I need to pick up RC planes again (it's for the kids, allright?). In the intervening years there's been a revolution in the sport, thanks to spin-off technologies from the computer industry. The high-end radios are digital and use the same spread-spectrum techniques as wi-fi and cellphones, which gives them long range and freedom from interference. But the real revolution is that over the past five years, electric planes have become as powerful and ubiquitous as the noisy and fiddly glow-plug gas engines were in my youth, thanks to two major innovations:

• The brushless motor. In traditional electric motors, a brush transfers electricity to a rotating electromagnet, which spins within a ring of permanent magnets. In a brushless motor, the electromagnets are in the outer ring and don't move, and instead the core spinning shaft contains the permanent magnets. An intelligent electronic controller does the necessary current switching in the electromagnetic coils, with no contact required, which makes them far more efficient and quiet than traditional electric motors. Brushless motors motors first showed up in CD/DVD drives, and have now migrated in more powerful form to everything from hybrid cars to model planes.
• Better batteries. Thanks to the insatiable demand for higher power density for consumer electronics, we've moved from Nickel-cadmium (NiCad) batteries to Nickel metal hydride (NiMh) to Lithium Ion Polymer (LiPo) in our cell phones, MP3 players and laptops. This has carried over to model airplanes, and in combination with brushless motors electrics can now equal the performance of a gas-powered model with far less noise and hassle.

To get a sense of what the combination of those two enable, check out this awesome demonstration (thanks for the link, Francisco). The power-to-weight ratio of this kind of flat-body foam aircraft is about 2-1, which is a real tribute to the potential of electrics these days.  

That guy has serious skills. For the rest of us, they way to start is with something a lot less demanding: ready-to-fly planes with brushed motors and NiMh batteries. I tested a few such planes for Wired's gadget section last year, and I've since tried a few others in my quest to find a way back into the sport that works for kids and adults alike.

The first thing to keep in mind as you enter this sport is that it's hard. The combination of wind, the disorienting effect of flying from the ground rather than the cockpit, and the sharp turning required to keep a little plane in eyesight makes crashing frequent, expensive and, with the wrong planes, demoralizing. You can literally invest hundreds of dollars into planes that you can't keep in the air for 30 seconds, and the crashes can be catastrophic. 

That said, here's what I recommend (you can see that I've mostly gravitated to the ready-to-fly Hobbyzone line, which is based on a v-tail, carbon-fiber body spar and foam wing design that I think strikes the best balance of performance and crash-tolerance):

Best First Airplane: Hands down, the Hobbyzone Firebird Commander 2. Do not be enticed by the cool-looking and cheap RC planes, such as the Air Hogs line, that you can find in toy stores and Radio Shacks. They're underpowered and can't handle any wind at all.  Likewise, don't start with a three-channel (elevator, rudder and throttle) airplane, because that's too much to handle for beginners. The Firebird Commander is a lovely balance of stability and performance, with enough power to handle wind (which is very hard to avoid in the real world) and gain altitude quickly, while still being as simple as possible (just throttle and rudder). Tip: don't even try to fly any proper RC plane in anything less than a football field-sized space, and ideally something even larger. Trees are your enemy.

Best Second Airplane: Several candidates. After you're comfortable with the basics of launching, circling and landing, you should advance to three channels (typically rudder, elevator and throttle, but sometimes aileron, elevator and throttle). In that class, we're currently flying the Hobbyzone Firebird Freedom (its mid-wing design handles wind well, but can be a bit twitchy, especially on launch), the Hobbyzone Aerobird Swift (which has ailerons rather than a rudder, which give more control but requires a little coordinated elevator work to achieve the full effect. Nevertheless it's my favorite in this class). We're also flying a Multiplex Easy Star all-foam plane (shown), which is easy to repair (superglue fixes everything). Although its performance is just average it does seem to handle abuse well.

Best Third Airplane: I'm still looking for one. The ideal candidate would be stable and powerful enough to allow us to start learning proper stunts, from loops to rolls, while still being hand-launched and very crash-tolerant. The flying-wing style planes look very fun, and I've seen them often at the park, but are they just too hard to keep in the air? Advice, please.

 

UPDATE: Chester offers some real world experience with this kind of plane:

 

 

 

 

 

 

 

So instead, I think I'll go with this cool Spitfire instead:

New Lego Mindstorms!

Readers of this blog will know that I'm a big fan of Lego, both the products and the company's Long Tailish way of tapping its consumers to design new products and share their own creations.

So I was really excited to hear that after five years, Lego is coming out with a new version of Mindstorms, the geeks' favorite programmable robot construction set. The new version, which was announced this evening at CES, is called Mindstorms NXT. It has a 32-bit processor, proper servo motors, new sensors (including color vision and hearing) and bluetooth connectivity so it can be controlled by a cellphone.

But what's really cool about the new Mindstorms is that they're a collaboration between Lego's engineers and some of the fans who run the very active Lego community projects built around open-source versions of its software. Lego used one of the fan-designed sensors and asked other fans to help design some of the software used to program the Mindstorms.

Someday business schools will use this as a case study in applying Web 2.0 peer-production techniques to traditional product innovation. In the meantime, you can read all about it in the cover story from next month's Wired, which we've released early as a sneak preview.

Xbox 360 Media Center Update

After my enthusiastic post about the Xbox 360 as a Media Center Extender, I belatedly discovered that updating the software for the 360 had disabled my older extenders, including the original Xboxes and a Linksys hardware extender.

I hunted around online for help and didn't find any, so I spent a fruitless hour and half on the phone with Microsoft tech support (being bounced from Xbox support to Xbox 360 support to Windows support, each time eventually escalating to managers who weren't able to help). Finally, at wits-end, I asked Charlie Owen, a Microsoft project manager who runs a great Media Center blog, if he had any suggestions.

Charlie put me in touch with the team that had worked on the Extender port. After a few days of running diagnostic tests, we discovered that it was due to a version conflict between some of the earlier Extender software I'd been running on the original Xboxes and the new Extender manager that you download and run on the Media Center PC as part of installing the 360. It's now solved and I'm happy.

Two lessons from this:

1) If anyone else is having trouble getting older extenders to work once you've got an Xbox 360 on the network, do this:

Make sure you have the latest Extender software. That's 1.01 for the Xbox, and to use the below process you need the DVD version that was sent out earlier this year. Delete any files from the previous version by going to the Xbox dashboard, selecting "memory" and deleting the "Media Center Extender" entry. On the 360, disconnect the Media Center (it's on the media tab). Uninstall the extender software from the Media Center PC, and reboot.

Then download and install the new PC extender software. Put the 1.01 extender disk in the original Xbox and go through through the usual 8-digit code entry to associate it.  On the 360, use the media tab options to do the same. This should clear your system and ensure that everything's working with the latest versions of the software.

2) The Microsoft team (Rob Lehew, the MCX project manager, and his colleagues) were totally great and quickly got me to the solution by diagnosing packet traces and otherwise walking me though some process-of-elimination steps. Obviously I'm not the average customer and they don't usually have project managers doing tech support. But because the team has a number of active bloggers who are accessible and willing to respond to users, it's much easier for anyone to find answers quickly from people who know the most about the product.

This is a great example of how company blogs can improve consumer relations by putting a human face on the development team. That's helpful in problem solving, as in my case, but it's even more useful in passing on tips and tricks from the pros and inviting suggestions from users on future development.

The old model was mostly to use newsgroups and forums for this, and that still has its place for really specific tech support. But I find blogs far easier to navigate and read, and you can subscribe to them in a way that you can't with newsgroups. Obviously not all developers want to take on the email and comment burden that comes with having a blog, but it only takes a few to really improve the customer relationship. Hats off to Charlie and the rest of the Media Center team for the fine role model.

"The Movies": machinima for the rest of us

    I'm finishing the Long Tail Producers chapter and spending more time researching machinima as an example of what can happen when the tools of film production are really democratized. The promise of machinima is to put a Pixar in every home, just like GarageBand and its kin put a recording studio on every desktop. We're not quite there yet, but it's getting closer.

    The problem with standard machinima based on first-person-shooter game engines is that you're painting with an incredibly limited palette, both in terms of the characters and settings and in what you can do with them. There's a reason why all the Halo 2 movies are about space marines fighting aliens or each other. The Sims 2 was the first effort to make a game based on a more realistic setting machinima-friendly, and the result was The Strangerhood. But even that's hard to control, and prepping for a scene can take hours.

    Which is why my kids and I were really looking forward to The Movies, a new game/machinima studio from Peter Molyneux, the legendary games designer of Black & White fame. We've now had a few weeks to play with it, and it's really quite wonderful.

    At its core, it's a SimStudio game, where you have to build a Hollywood studio, operate it, make movies and the rest of the usual Sim-style resource management stuff. That's fun enough, but the really interesting bit is the "Sandbox" mode, where you can make your own movies, pretty much from scratch.

    You choose from a wide variety of sets, costumes and, best of all, pre-animated micro-scenes ("Walk in with ax", "Discover body"; "Slip on banana peel") many of which can be tweaked with more or less tension, humor, or whatever makes sense in context. You pick pre-created actors or create your own with the usual facial modeling tools. Then you string all these together in a storyboard with some gentle plot and pacing assistance, and when you're done "filming" you can type in your dialog as subtitles or record a voice-over in a sophisticated post-production studio.

    We're having a great time with this, mostly trying to replicate the kids' favorite The Movies movie, Attack of the Space Chickens (but minus the kissing part). The only hitch is that you've got to play the main studio sim game from the 1920s to the current era to unlock all the costumes and sets and that's way too much trouble for us. So far we're in the mid 1930s, and have unlocked a Gorilla and Robot costume to go with the usual array of outfits less amusing to small children. I have no idea when the Chicken costume will appear, although I hope it's soon for domestic harmony's sake.

    I wish Lionhead would release a stand-alone Sandbox version that has all the assets available from the start. Perhaps there's some cheat code that does this, but the one I found doesn't do the trick. Anybody know a better way?

    Meanwhile, we're not the only ones to see The Movies as a machinima breakthrough. Loads of people are uploading their own films, ranging from light comedies to politics. Speaking of which, check out Clive Thompson's post on the French machinimists who used The Movies to create a commentary on the recent riots. As Clive notes, "The raw DIY feel of this project is more punk rock than anything anyone's done with music in about 20 years." I agree: there's something really interesting going on here.

Machinima 202: Losing the targeting reticle

Two months ago I posted on the fantastic Codex Series, a machinima movie made on Halo 2 (which subsequently won a bunch of awards). It was my kids' favorite movie for a while, and after the director, Alexander Winn, kindly sent us a CD of the soundtrack, that became their favorite album, too. But the great thing about machinima is that the act of watching it is an inspiration to make some yourself.

So my eight-year-old and I got set up to recreate one of the scenes from the Codex. I was stunned by how intuitively he got the somewhat complicated mechanics of machinima. For instance, for a scene with N actors, you need N+1 players, because one has to be the camera (the six-year-old came in handy; when she got bored, she was replaced by my foot). Also, you need to choreograph the scene carefully--it's so easy to move quickly in Halo 2 that if you don't practice all the moves and camera placement one or more of the actors will find themselves bouncing off-screen.

The trickiest part, however, was turning off the targeting reticle, that circle at the center of the screen where your gun is aiming. Proper machinima shouldn't look exactly like a videogame, and turning off the heads-up-display elements is a big part of achieving that. We Googled a bit and couldn't find the answer, so we asked the master, Alexander Winn himself, for advice.

Here's what he told us. (Note: the following is for hardcore machinimists only, and I'm actually just posting it so the next people who Google with the same question will find it.):

Ah yes, the reticle.  Here's the explanation, but keep in mind that it sounds a lot harder than it is...

Set the starting weapons to whatever you want them to be, but make sure that the "Weapons on map" settings include Plasma Pistols and that the gametype is either Assault, Oddball or Capture the Flag.  With the cameraman, drop one gun for a Plasma Pistol, and then dual wield with the other slot.  (So, for example, you have a plasma pistol in one slot and in the other you're dual wielding two SMGs). 

Walk over another plasma pistol and hold X, (NOT Y, X!!!), and you'll switch the plasma pistol to your right hand.  Drop the gun in your left, and you'll have a plasma pistol in each hand. Once you've done that, walk over to the special object, (bomb, ball or flag), and charge up the plasma pistol.  While it's charged, try to pick up the object.  You should drop both, leaving you with only one plasma pistol.  Do the same thing again, and you'll be weaponless.

NOTE:  Being weaponless in the Halo 2 engine is "unnatural", meaning that you'll pick up a gun if you walk over one and you'll have to do the whole process again.  So be careful where you step!

So there you have it. As it happens, my son and I didn't get much further in our Halo 2 movie anyway, because we found an even easier way to do machinima. But that will be the subject of the next post...

First take: XBox 360 Media Center extender

Thanks to my cool day job, I got an early Xbox 360 and spent some time today running it through its paces. I'll leave the reviews of the release-day games to others (although I found Amped 3, a top-notch snowboarding game,  to be delightfully off-beat and stylish, with loads of 8-bit and retro design touches to counterbalance the slick graphics of the main game). I was looking at it more as a digital home enabler.

Given my interest in the Media Center PC as a Long Tail video platform, I was particularly interested in how well it worked as a Media Center "extender", serving as the link between one of the TVs scattered around our house and the Windows Media Center PC running in the study that serves as our media library and DVR.

The simple answer is very well indeed. We'd been using the original Xbox as an extender, but it was always a somewhat clunky solution. For starters, the extender software required a disk to be loaded each time, so if there happened to be a game disk in the machine already, switching them was an extra step. The Xbox also had to be turned on and off manually, an artifact of the fact that it's basically just a regular PC in a closed box.

The Xbox 360, on the other hand, has the extender software built-in, and the remote control included with the "fully-loaded" package has a dedicated green Media Center button that both turns on the Xbox 360 and takes it directly to the Media Center interface. The remote can also turn the machine off, which is handy.

Performance, unsurprisingly, is significantly quicker. It boots up faster, connects to the Media Center a bit sooner, and the on-screen interface is more responsive and crisp. It's not all perfect: although the Xbox 360 supports HDTV-resolution games, it doesn't yet support streaming HDTV video [correction: it does stream high-def video, but the Media Center DVR can only record  HDTV from over-the-air broadcasts (is someone out there actually watching those?) right now. See this for more. The following will no doubt extend that to recording HDTV from digital cable, which is how most people get it.]. But Microsoft's recent deal to include digital cable card support in the Media Center may bring that soon.

Interestingly, you don't need to have a PC running Windows Media Center edition to get some of this functionality. The Xbox 360 has its own music player (complete with trippy visualizations from the famed Jeff Minter) that can stream music and photos from any Windows XP PC.

I suspect that the release of the Xbox 360 is going to be one of two breakthrough events that take the Media Center concept mainstream. The 360 is a mass-market device (the original Xbox sold 22m units worldwide, and the Xbox 360 will presumably do better than that) that is built from the ground up to distribute digital content around the house. Having a Media Center extender built into a hot videogame console will go a long way to legitimizing that concept.

The second breakthrough event will be the release of Microsoft's next version of Windows, Vista, which will come with the Media Center technology as a default in the home version (although they better fix the interface before they release it). Already about half of new PCs are sold with the Media Center version of Windows (although most don't come with TV tuners, so they're probably not used as DVRs). Analysts project that 5m such Media Centers were sold this year, a figure that will double next year. Once Vista comes out, perhaps in late 2006, that could rise severalfold.

Between these two forces--the inclusion of Media Center software in most new PCs and the spread of tens of millions of Media Center extenders in the form of videogame consoles--it's not hard to see the Media Center becoming the leading DVR/streaming standard in a few years. Its rise is also helped by the fact that it's both a relatively open platform on which other companies can create software and services, and it supports more standard media formats than the closed-box DVRs of TiVo or DirectTV or the proprietary technology of cable company set-top boxes.

I never thought I'd say this, but by the standards in this industry Microsoft is actually looking relatively innovative (Apple is playing catch-up with Front Row, but until it comes up with its own version of the extender concept to distribute content easily to TVs around the house, it won't have broad appeal).

What's important about the Media Center is that it takes the DVR concept and extends it to all forms of content, whether broadcast or downloaded from the Web. By having a broadband-connected PC at its core, it's by nature a full-featured connected device that can keep up with the pace of innovation in digital media online. If the Xbox 360 and the new content marketplaces of its associated Xbox Live service continue to take off, we really could have the beginnings of a Long Tail platform that could challenge broadcast TV.

Friday Fanboy: the Zen Micro + Rhapsody To Go

I really wanted an iPod Nano, I did. It's a triumph of industrial design, a delight to hold, and the sweet deal Apple got from Samsung for the 4GB SRAM chips is a tribute to the power of monopsony (don't tell the South Korean FTC). In terms of hardware alone, I think it's pretty clearly the best portable music player on the market right now.

But instead I bought a 6GB Creative Zen Micro. Why? Three words: Rhapsody To Go.

For all the genius engineering of the iPod Nano, its functionality is largely determined by the iTunes software and music service, which really leaves me cold. It may be the most popular music service, but it is far from the most fully-featured. Compared to Rhapsody, iTunes has bare-bones artist descriptions, few editorial reviews or much other useful metadata, an unhelpfully broad single-level genre hierarchy and lame recommendations. Plus the over-restrictive DRM in its AAC files makes even Microsoft's WMA look good.

On top of that, I prefer Rhapsody's all-you-can-eat subscription model to iTunes' $0.99/track downloads. For $8.25 a month (on the yearly plan), I can listen to all the music I want from the 1m+ tracks the service offers. As a result I'm far more inclined to explore new bands and genres, something that's now fun, easy and risk-free. I'm even starting to feel a little cool again.

So the main reason for not choosing a Nano is that I didn't want to leave Rhapsody and the ability to stream unlimited music. But the reason for choosing the Zen Micro in particular had to do with a newish feature that I was keen to try, extending that all-you-can-eat subscription model to a portable music player.

A regular Rhapsody subscription allows you to store any tracks, even those you don't own, on your PC so you can access them even when you're not online, such on your laptop on a plane (as long as you continue to subscribe to the service). But the Rhapsody To Go option, at another $5 a month, allows you to copy them to a portable device so you can carry them anywhere.
Right now there aren't many MP3 players that support this, but the Zen Micro is clearly the best of those that do.

Charlie Owen has posted on his own research and logic in choosing the Zen Micro over the iPod Nano. Among the factors that influenced him was the cost per capacity, which he summarized with this:

• 2GB iPod Nano = $99.50 per GB
• 4GB iPod Nano = $62.25 per GB
• 6GB Creative Zen Micro = $33.17 per GB

My own reasons included the Micro's small size, intuitive controls and interface, and USB 2.0 data transfer and recharging, which meant fast loading and no extra charger.

So now I've got a cool black Zen Micro. It's loaded up with 2,000 cool (and legal) tracks that I didn't have to pay for. (That's $2,000 worth of iTunes music.) Maybe next month it will be a different 2,000 tracks. And another 2,000 the month after that. All for an extra $60 a year. Cool.

Friday Fanboy: Lego's Long Tail

I'm often asked for Long Tail examples outside of entertainment, especially those that apply to physical goods, not just digital bits sent down a broadband connection. There are quite a few (aside from eBay, the obvious one), but perhaps my favorite is Lego.

If you just know Lego from kids' birthday parties and the display shelves of a toy store, you've only seen half of the company. The other half is the Lego that caters to enthusiasts, ranging from kids who want more than the stock kits to adults who have turned to bricks as the ultimate prototyper's toolkit.

It all starts with Lego's mail-order business, which began as a traditional shop-at-home catalog and is now increasing organized around its website. In a typical toy store, Lego may have a few dozen products. On its online store, it has nearly 1,000, ranging from bags of roof tiles to a $300 Deathstar (shown). If you want to see how different the online market is from the traditional retail market for Lego, check out their topsellers list. Only a few of those products are even available in stores, and most of those are inexpensive items added to other purchases to bring them over $50 and thus qualify for free shipping.

It's worth pausing here and considering the Long Tail implications of this. At least 90% of Lego's products are not available in traditional retail. They're only available in the catalogs and online, where the economics of inventory and distribution are far friendlier to niche products. Overall, those non-retail parts of the business represent 10-15% of Lego's annual $1.1 billion in sales. But the margins on these products are higher than the kits sold through Toys R Us, thanks to not having to share the revenues with the retailer. And because the virtual store can carry products for all Lego fans, from kids to adult enthusiasts, and not just the sweet spot of nine-year-old boys, the range of prices can be a lot greater online, from $1 bricks to the aforementioned $300 Star Wars kit.

The next level of Lego obsession is joining one its two clubs. One gets you the monthly Lego magazine and catalogs. The other is the online club, where all the games and other cool stuff is. Basic membership for both is free, but if your kids are really into Lego you might want to consider upgrading to the Brickmaster level, which brings a bigger magazine with a lot of DIY projects (like MAKE for bricks), five exclusive kits that show up at your door, and a ticket to Legoland.

After that, it's time to start getting serious about your own creations. Lego has a long history of offering tools online to encourage model trading and other collaborative peer production. In 2000, its "My Own Creation" project led to a contest for the best user-created model. The winner was  blacksmith shop that Lego licensed from its creator and offered for a while as a commercial kit. After that, it offered Lego Mosaic, which allowed users to upload images that were converted into 2D Lego brick patterns, downloadable by all.

Earlier this year, Lego launched its most ambitious peer-production effort of all, Lego Factory. The idea is that you download software that allows you to design and build virtual creations, then upload them to the company. A week or so later, you get a kit with the necessary parts delivered in a box with an image of your creation on the front. What's especially cool is that others can buy your kit, too, and there's a nice selection of user-created models, such as this truck, available for purchase. More than 77,000 models have been designed this way, and some of the best of them are also being released as official Lego products (Lego pays the creators a royalty).

However, all is not what it could be in Factory land. Mass customization is cool, but when you have 7,000 possible parts in 75 possible colors (that's more than a half-million possibilities), the fulfillment challenge of offering users full freedom quickly becomes overwhelming. So Lego limits choice in two ways. First, each model can only be built from a single set brick palette, such as car parts. Second, those parts come in pre-packaged bags of a fixed number of bricks, so you'll likely get more than you needed. If you're not careful, a simple vehicle that might cost less than $10 in retail can turn out to cost nearly $100 in Lego Factory simply because it uses those bags of parts inefficiently.

Fortunately, there's a hack-around. Lego enthusiasts compiled a database of what bags were in which palettes and created software that helped builders use those bags efficiently, avoiding having to buy an expensive bag of parts for a single brick. And to its credit, Lego encouraged this. But that's still too hard and limiting for most people (including me), so Lego is now considering how to improve the experience, starting with easier-to-use design software.

I asked Michael McNally, Lego's senior brand-relations manager, whether Lego saw parallels in any other company's of approach to catering to niche market segments and encouraging peer production. Interestingly, he gave Apple's iTunes as an analogy. iTunes lets you download individual songs, not just albums. Although you can't upload your own music to iTunes, there are plenty of companies that can do that for you. And can make your own playlists and share them with other users, which is a bit like a custom Lego creation from standard parts. "What iTunes does for music, Lego Factory is doing for people who like to build," McNally says.

Friday Fanboy: The Codex Series

Last weekend the six- and eight-year-olds came back from a friend's house raving about a new movie they'd seen, The Codex Series. Never heard of it? I hadn't either. But I looked it up and found the most extraordinary thing.

The Codex Series is a machinima film, made with networked Xboxes running Halo 2. As in all machinima, the human actors are essentially puppeteers, controlling the videogame characters to act out scenes in the game's multiplayer environments while the actors do the voiceovers. All the visuals are generated by the game itself, so the role of the filmmaker is simply to write a script, act it out, and edit the results. The technique is the same as the earlier Red vs. Blue, the Beckett-like surrealist comedy of bored marines stranded on an alien planet, which has been a favorite of ours at Wired for several years now (read Clive Thompson's excellent NYT Mag piece about it here.)

The end result of all these projects is super low-cost computer graphics animation, like a basement Pixar. Anyone can make a CG film this way and every year hundreds of talented amateurs do. Some of them are amazing.

The Codex Series is the best machinima I've ever seen. It has a stirring plot (an alternative history of the battle between humans and the Covenant in the Halo universe), edge-of-your-seat pacing, distinctive characters, and a pulse-quickening soundtrack. The kids watched it all weekend, and now beg to see it again. I'm on my third viewing myself, and I must say that Episode 18 is a stunner, a masterful interweaving of simultaneous rally-the-troops speeches by commanders about to battle. It builds to a crescendo that will leave you breathless.

The series has 20 episodes, and then, unlike Red vs. Blue, it stops. The creators, who are all in their teens, have now graduated from high school and scattered to the winds. Plus the film's runaway success has left some ruinous bandwidth bills for the millions of downloads:

Q:  Are you going to make a second season?

A:  Most of the Codex Crew went off to college right after finishing Episode 20.  While making The Codex this summer we met every day at noon and worked until at least eight o'clock every night, sometimes going as late as 3:30 the next morning! Obviously we don't have that kind of time in college, especially since we have been scattered into three different time zones. The Codex has also been a very expensive project for us, and now that we're in college we can't really afford expensive hobbies.

If you're not sufficiently impressed yet, consider this: My children's favorite film was not made by Disney, but by a dozen Dallas teenagers playing a videogame in one of their parents' basement. By Hollywood standards, the film cost essentially nothing to make and is free to download. It's had 13m viewers so far. There is now a DVD and a soundtrack CD.

If that isn't Long Tail, I don't know what is.

Friday Fanboy: The Media Center PC

I'll no doubt be savaged for praising a Microsoft product, but my favorite technology of the week is our Windows Media Center PC. We resisted a DVR for years because we hardly watch any television, but as the kids got older and more numerous we found ourselves putting them in front of the TV more often to have a few moments to ourselves. Random cartoons are just awful, so we finally had a good reason to get a DVR: to limit the kids to just the TV that's worth watching.

As it happened, my wife needed a new PC. We realized that by picking one with Media Center 2005 pre-installed we could get a DVR essentially for free (or, to be precise, $150 extra to add a dual-tuner TV card and a remote control to the configuration). Anti-DRM zealots who think we should have home-brewed our own open-source DVR instead are invited to make this case to my wife.

The cool thing about the latest (2005) version of Windows Media Center is that it can run in the background of a regular PC. So as far as my wife is concerned, it's just her standard workhorse PC, sitting in the study. But while she's doing email and such, it's recording TV on a 400 gig hard drive or streaming it to TVs around the house. Since we already had the cable TV line coming to the study for the cable modem, installing the PC was simply a matter of a Radio Shack adapter that split the coax into three--one to the cable modem and one to each of the PC's twin tuners.

The other cool thing about Media Center 2005 is the way it works with "extenders", which are set-top boxes that you place next to each TV you want connected to the DVR. This allows the Media Center to act as a central media server for the whole house, giving any TV simultaneous access to the same recorded content. We already had an Xbox on one TV, so turning that into an extender just required inserting a disk. We bought a wireless Linksys extender for the other TV.

For us, Media Centers have five big advantages over traditional DVRs, including TiVo.

• No monthly fees.
• Centralized storage means that all TVs around the house have instant access to the same content.
• Unlimited storage capacity.
• Can stream all the other media on your PC to any TV, including music and home videos.
• By DVR standards, it's a relatively open platform (certainly compared to the DVRs offered by your cable company), and there are an increasing number of plug-ins that expand its features.

There are, to be fair, a few disadvantages, too. It's a Windows PC, so you have to restart it once in a while. If you do that while the kids are watching television upstairs, they're going to yelp. And because it's a PC that's running all sorts of other software, there is the risk you'll install something funky that messes things up. I put in a new sound card and the DVR stopped recording TV sound until I undid most of the changes. Finally, the Linksys extender, which is the only hardware extender available and feels a bit V1.0, freezes every now and then and has to be restarted.

As with our Roomba, there's a parenting strategy angle, too. The kids, having grown up with computers, are desperate to control their screen experience. So in exchange for the giving the two oldest (6 and 8) the right to use the remote control, they are required to skip all ads. And, amazingly, they do (along with replaying the fart jokes in the show, skipping the parts with boring singing and otherwise being kids).

It appears that, after a slow start, Media Center PCs are finally taking off and now account for 43% of desktop PCs sold in retail (although less than a third of them have TV tuners). Microsoft is promoting it by eliminating the price premium over the regular Windows XP home edition. As a result, you can now find Media Center PCs for less than $900. It's also going to be built into the default home version of Windows Vista, the next major Microsoft OS. And it's at the core of Intel's new Digital Home strategy.   Analysts expect Media Center PCs to reach US sales of more than 20m a year by 2007, and some are rather hyperbolically predicting that it will be "next year's iPod".

Finally, there's a strong Long Tail angle to the Media Center. It is, at its core, a platform for unlimited-choice TV. It connects the Internet to the TV screens around your house via a simple, TiVo-like interface. Right now, most of the video content comes over the broadcast network, is cached on your PC, and then streamed over your home network. But that content can just as easily come from anywhere on the net, and independent video marketplaces such as Brightcove and Akimbo are planning to release their services as Media Center plug-ins to deliver just that.

The broadcast era has peaked and the on-demand era is beginning. The cable and telephone companies are betting this means more video-on-demand over their proprietary networks, which may indeed be the case for a while. But I'm betting that the Web model--with the link between content and the pipes that delivers it severed once and for all--will ultimately prevail. And a Media Center PC is the closest thing to that you can find today. It's a glimpse of the future.