After two years, the extraordinary Radio Open Source, hosted by Chris Lydon and produced by Mary McGrath, is off off the air and off the ‘Net. It’s a sad day for this hybrid radio/podcast program’s producers and listeners. I wish I’d been wrong, but I predicted this two years ago in my essay, The Future of Public Radio.

Radio Open Source’s problem is just that: It’s a hybrid. It has the cost structure of a public-radio program — a $1 million/year budget — in a podcast-revenue marketplace. Even an infusion of $250,000 couldn’t save them. In the past twelve months, between IT Conversations, Social Innovation Conversations and our other channels, we produced about 400 programs at a total cost of less than $50,000. Radio Open Source produced half as many programs for 20x the cost. I don’t know Radio Open Source’s audience, but we reached nearly a half-million unique listeners and averaged more than 20,000 downloads per program.

To some extent, this is an apples-to-oranges comparison. Radio Open Source was a live show, recorded in expensive WGBH studios, and had a good sized paid staff. We have no real studios, no paid full-time staff, and nearly half of our programs come from conferences that pay their own recording costs. But that’s the point. Whether non-profit or otherwise, every business has to work within a marketplace. Radio Open Source’s content and production values were terrific but expensive, and there simply isn’t an online market sufficient to support it. As I explained two years ago, there isn’t a public-radio market to support it either.

And why didn’t Radio Open Source succeed as a public-radio program? (1) Scarcity. Radio suffers from that old-school model in which there are only so many hours in the day in which to air a program. (2) Peculiarity. For a station to pick up Radio Open Source (a four-show-per-week series), it had to cancel another daily show and fill the extra day with something. (3) Timing. If a station did pick up Radio Open Source, it most likely ran it at some odd hour of the day, whereas the podcast edition was always available on my schedule. (4) Esotericism. Let’s face it: Radio Open Source was aimed at a fairly elite audience, and broadcast needs numbers.

I hope that Radio Open Source can resurface as a podcast-only program in the fall, perhaps as a weekly rather than four times per week. But to do so they will have to scale back their expenses to match the realities of the podcast marketplace. They won’t be able to re-hire that great team of producers. They’ll have to scrape to get by, just like all podcasters except those who are supported by unrealistic venture capitalists. Record in a bedroom. Leverage part-time and volunteer staff. Focus on Chris’ brilliance, but Chris and Mary may have to do all the research and scheduling themselves. Even then, it’s not at all clear they’ll be able to support themselves from Radio Open Source as a podcast. Very few of us can. And there are very few funders in the non-profit sector who want to underwrite any podcasts, let alone those produced on a broadcast-style budget.

I’d love to have Radio Open Source join The Conversations Network in the fall, and I’ve said as much to Chris and Mary. I look forward to discussing that opportunity with them once they’ve recovered from the trauma of having to pack up, move out and lay off the staff. In any case I wish them well and thank them for producing a truly excellent and inspiring program for these past two years.

The buzz surrounding Facebook is extraordinary, and the rampup over the past 90 days reminds me of 1999. I think it’s more than accidental. Facebook have done a few things quite well. First, the design is very clean and the UI is intuitive. It’s very easy to find and do things, and there’s not a lot of crap on the screen to distract you or get in the way. Second, the performance is excellent. The site is as responsive as Gmail, which is now pretty much the benchmark for highly interactive sites. Third — and this is a huge one — is the Facebook API that allows anyone to build and run an application and make it appear as though it’s within Facebook. MySpace and LinkedIn, as examples, are scrambling to catch up on the API front, but it may well be too late for them. Facebook has the momentum.

I’ve created a few Facebook groups over the past few days:

• The Conversations Network
• IT Conversations
• Social Innovation Conversations
• PodCorps.org
• TeamITC

I’ve probably bought more noise-canceling headphones than most people, but just never been satisfied with the choices. Of course there are the Bose models, but everything that company makes is overpriced and overhyped. Their aviation headset (ie, used by pilots) beats the competitors, but those for passengers aren’t worth the $$. For two years I used a pair of inexpensive Sony MDR-NC6, which are a very good value, but at a much lower level on the price/performance curve. Last year in Frankfurt I picked up a Sennheiser PXC-150, which I’ve been using until now. But based on some recent reviews, I just bought an Audio-Technica ATH-ANC7
($199.99 from Amazon), finally a noise-canceling headset worth blogging about.

These are cup-style phones that fold fairly flat. They’re very comfortable. A single AAA battery slips into the left earcup. I first tried them with the noise canceling off and was very disappointed. They worked, but the level was low and there was a complete lack of high frequencies. (Some noise-canceling headsets don’t work at all with the noise-canceling feature turned off.) But flip on the active circuit, and the audio quality is excellent. I’d rate the audio as an A- and the noise canceling as a B+. At $200, I think they’re the best value at this end of the price curve.

Dave Winer is digging into the implications of services like Twitter and how they provide the infrastructure for extraordinary unintended consequences. Given Dave’s history of creating things that fundamentally change the way we do things (outliners, RSS, podcasting) I fully expect we’ll see something exciting from his work, even if we won’t know it’s exciting at first.

His recent post points out that Twitter is all about trivia. It fits into the new economy/ecology of the long tail, narrowcasting, tagging and as David Weinberger reminds us, everything is miscellaneous. I relate it further to crowdsourcing and our own efforts at PodCorps.org.

We’re getting away from thinking that success only comes from things that are important to everyone — mass audiences. We’re actually  getting away from narrowcasting, too. That was a one-way idea that broadcasters could offer targeting content to narrowly defined demographic groups. What we’ve learned is that those groups can and will create their own content and that it will be much more interesting and valuable to the members of those groups. I wouldn’t be surprised to learn that for people who use Facebook, Twitter, etc., they already receive far more aggregate value from (and spend more time interacting with) their social networks than with traditional media. Perhaps someone can identify the dates and events at which we turned that corner.

Interesting discussion on Artur Bergman’s O’Reilly Radar blog about the lack of SLAs for Amazon’s Web Services.

Mike Culver at Amazon has produced a screencast about how we use Amazon Web Services’ SQS as the backbone for the GigaVox Media infrastructure.

PodCorps.org has joined Ourmedia, Outhink Media and the Internet Archive to launch the Remix Politics channel. Our goal is to collect video and audio from political campaigns, speeches and other events and publish it under a Creative Commons license to support mashups and research.

Andrew Keen has stirred up a hornets’ nest with his new book, The Cult of the Amateur — How today’s Internet is killing our culture. To hear what all the fuss is about, listen to Andrew interviewed by Moira Gunn on our Tech Nation series.

I just upgraded Parallels Desktop for Mac to v3.0, and I love it. I was impressed with Parallels 2.x, but this is quite an improvement.

• I can run OS X, XP and Vista all at once, very smoothly. Haven’t tried Ubuntu, yet, but it should work.
• Coherence (as in 2.x) hides the guest OS window, so every application appears as though it’s running under OS X.
• I have complete drag-and-drop and clipboard transparency across all operating systems.
• I can transparently open files in any environment from any other.
• I can run CPU-intensive and (some) hardware-dependent applications like SoundForge. [I get audio input from Firewire devices like MOTU interfaces, but I can send audio output from Windows through the OS X audio system.]
• Memory and CPU loads appear to be reduced from 2.x.
• The new snapshot feature (with versioning) is handy and fast.
• Installing/upgrading was straightforward.

Decades ago, I used to write system software similar to what the folks at Parallels have developed. We didn’t have help from things like Intel Virtualization Technology in those days, but this is still tricky stuff. I think I’m qualified to say that this is a very well-executed product. Kudos.

Update: I should add that I’m on a hefty box: Two dual-core CPUs, 4GB RAM, 1.25TB disk on three drives.