Month: December 2004

<![CDATA[It's great that Nick Denton of Gawker Media is featured by the Wall Street Journal as one of the 15 people to watch in 2005, because of his burgeoning blog publishing business.
But, here’s a hint for the Journal: The media evolution did not begin nor will it end in New York.
I also hope that Nick doesn’t suffer the disfigurement of his fellow honoree, Viktor Yushchenko. If Jason invites you out for dinner, Nick, don’t eat the soup….]]>

<![CDATA[I know we're all living in a broadband age, but the claims of Orb Networks sound highly improbable to me. Tony Perkins of Always On is raving about this “personal media portal” that just launched a $79.99-a-year service that promises to deliver the media stored on your home PC (Windows only) to any compatible (Windows, Windows CE and Windows SmartPhone) device over any network. It sounds really impressive, but there’s the catch. Seeing is believing in the media market and, well, I’ll get to that.
The secret sauce here is a combination of a server-in-the-sky that mirrors the home PC data and retransmits it to a target device and a neat sounding term for optimizing streams, transcoding. Transcoding is typically defined as conversion of content from one encoding scheme to another; in PC architecture, Transmeta’s Crusoe chips transcode Intel-specific code into a Crusoe’s assembly language and back again to operate a Windows OS on their non-Intel/AMD architecture.
But because the Orb system is so heavily reliant on Windows Media Player and DRM, it is certain that the encoding format is not changing much, rather the stream is probably being optimized to the detected network througput and screen geometry. For example, if I know that I am going to deliver a stream to device with a 2-inch by 2-inch screen over a 56Kbps network connection, I can optimize the image by removing a lot of detail from the background and probably get a pretty clean image over the network.
Can I deliver 30 frames per second, which is what television offers? Doubtful. Now, if I know I have a Wi-Fi connection to a mobile device then I can probably do pretty well. But over a typical GPRS connection, no way. I have a Sprint PCS phone, the Samsung MM-A700, which promises TV quality clips and it is, well, for shit. It takes a long time to buffer and looks like a sea of interference punctuated by dim images of human form.
Orb also claims to allow users to watch live television streamed from their home PC and that is simply an incredible proposition. If it is true, meaning that the picture is consistently clear and better than 24 frames per second, it is miraculous, but I suspect this is a demo that will not live up to the promise when subjected to real-world conditions. (Also, which is the right channel to sell this? Computer retail it ain’t. Wireless providers? Doubtful? Direct and through television? Expensive for a company with around $4 million in capital.)
This transcoding technology is not a unique idea, but it is the first time it is being applied to the personal content delivery problem. It also opens a significant business opportunity with the media industry. Imagine that transcoding features are linked to DRM to prevent files that do not carry a particular permission from being delivered or at the best quality. This would be a compelling solution for media companies.
And if you have a secure connection to people and know what they like in media you can sell them other stuff. So explained the founders of Orb to Tony Perkins:

“For example, your PC may see that you like Beyonce because you play her music all the time, so your PC may automatically request Beyonce related stuff (tickets to concerts, screensavers, etc.) via the Orb Network and serve it up to you,” Dr. Julia says.

This is remarkably similar to the reasoning the compelled Time Warner’s interactive television trial in Orlando, Fla. When I attended the launch in 1995, they were talking about selling lots of stuff in conjunction with programming, like Sienfeld baseball jackets and caps with the logos of favorite bands. I wasn’t convinced this was a rational business then and I am not now. Rather, they need to open the channel to programming that provides deep information about a product or simply give away premium content, like Major League Baseball in order to generate commerce opportunities, like ticket and jersey sales, if that’s where Orb really wants to take this.
Orb’s not far off base on the potential for commerce, but it’s thinking too small. Why does television succeed? Because it provides entertainment, yes, but also a truckload of advertising (I am speaking of financial success, not about aesthetics or social value).
The system must work, though, for any of this to make sense. While network bandwidth has improved dramatically on wired and wireless networks, there are many layers of the connection between a PC running Orb and a mobile device accessing media on that PC. Virtually the whole of this network environment is out of the control of Orb. The company cannot make a quality of service guarantee, yet it will be the primary recipient of any blame for poor performance, even if it is the result of poor network performance.
Orb calls its system distributed computing, but since it relies on its own servers to act as an intermediary, this is not a distributed computing system and more a store-and-forward architecture. For instance, if you take a photo with your camera phone, it is uploaded to Orb’s servers, then downloaded to your computer at home. Not distributed.
Orb is many years too early to offer streaming video from a home server over any connection. It’s facing a launch that could be embarrassingly like that of the Apple Newton, which overpromised and under-delivered because processor and memory technology were not up to the challenge. Wireless networks, and some wired networks, are not able to deliver the throughput necessary to keep Orb’s elaborate promises.
Finally, the price seems problematic to me. how often does it occur to someone that they need a picture, song or video stored at home, one they didn’t carry along on their laptop or a portable digital music player? Orb needs to gain a customer base to turn up a media distribution channel and that requires consumers to fork over $9.99 a month or $79.99 a year, plus $3.99 a month or $29.99 a year for additional users. A family of four using Orb would spend $170 a year simply to access media stored on a single PC in their home.
How much will people really pay if the functionality is widely available from different providers or for free through an ISP? Perkins and Orb claim they are the only players in this market, but there are three or four more I know about that take different but functionally, from the user’s perspective, similar approaches to the problem of connection people and media.
The VPN functionality in Orb needs to be free if it is going to become a channel for content in the medium term; media companies will provide a share of revenue to a content distributor that protects their copyrights—they do it today with cable carriers, CD pressing and DRM companies, to name just a few examples—so there is hope for a non-subscription business model or, at least, a substantially less expensive service. If Orb can manage its server capacity and network costs, it could fly. But not with a starting price of $79.99 a year, because the typical person doesn’t currently want to access media, other than pictures, stored at home. And there are plenty of alternatives for carrying media that don’t cost anything or very little incrementally to device purchases people are already making, like iPods, Windows PDAs, Treos, Archos video jukeboxes and so forth.]]>