Oil, Real Estate and the Computer Industry

We've all been reading the headlines about energy prices, and the real estate bubble. They're both global phenomena. Most consumers are accustomed to thinking about the impact of high oil prices on industry - airlines are punished by high fuel prices, so they make less money and hike ticket prices. Retail gas prices go up, so we tend toward more energy efficient cars, etc. But few folks, until recently, have thought about their impact on the computer industry.

$70/barrel for oil, given that a very substantial portion of our energy is generated by burning fossil fuels, implies seriously expensive electricity. It also implies a financial drag on companies with energy inefficient datacenter facilities - a drag that gets more noticeable by the day. I was with a search company a few weeks ago that told me electricity is their single biggest operating expense. Power matters.

Datacenter rents in cities across the globe are skyrocketing, too - but who in their right mind would build a datacenter in center city London (one of the world's most expensive places to do so)? Any financial institution whose competitive advantage is influenced by proximity to exchanges, or the speed of light (just go talk to a proprietary trading shop - yes, latency matters). Talk to a CIO, you'll hear them say (as I do, almost daily), "my biggest problem is that I'm out of space and power, and I can't exhaust the heat from my existing facilities - I'm finding myself allocating more floorspace to aisles than computers!"

We've had EnergyStar for desktops for quite a while - we're beginning to work with legislators to create similar standards for the datacenter. Which will expose the companies that have been focused on your living room, instead of your datacenter, for what they are. Environmentally irresponsible.

Because at $70/barrel and $10,000 per square meter/yr, believe me, innovation matters - and not just in theatrics around price of acquisition, but practical operation - all in, all up. Hardware, software, power, real estate, management talent - all in.

Just ask the fine folks at the University of Buffalo, who chose to build a grid based on parts from a company that deliberately took the word "Computer" out of their name. (btw, Marc's analysis is pretty stunning - I encourage you to read it, and to watch our September 12 launch event to see our solution to their pain - help is on the way, folks!).

Is it a coincidence Dell's located nearby to so many energy companies?

Hey, I'm just asking.

Posted on 02:51PM Aug 29, 2005 |

Share this post  | | | | | |

DReaM, the Open Media Commons and the Future of IP

You've probably seen a bunch of the coverage from what we announced on Sunday evening around an open digital rights commons at the Progress and Freedom Foundation's Aspen Summit. I thought it'd be useful to put it all into context, and try to clarify some of the misimpressions. (Original video and audio here.)

Let me start with what's still viewed, by some, as controversial. I believe in intellectual property. Since Sunday, I've gotten a round of communications from those that believe the whole concept of IP is suspect, and digital rights management of any form is immoral (one writer used "evil"). I clearly don't share that view - you'll recall, Sun Microsystems is basically an intellectual property fountain - pour dollars in the top, some of the brightest minds go to work turning it into innovations, that are then translated into value by our customers and channels (and manufacturers).

With networks that connect everyone to everyone, the transmissability, and the value, of intellectual property is obviously on the rise - as is the need for IP owners to want to control their destiny. By IP owners, I mean everyone from an amateur photographer with a valuable snapshot or movie, to a studio owner with a blockbuster new film; from a software company delivering an operating system, to a fab owner with a new process innovation. It's all IP - bits on the internet - and it's all got value to someone (and it's all readily distributed over the internet).

Now, any number of challenges in the marketplace (from Supreme Court decisions to teenagers in movie theaters) suggest the need for an alternative to the systems and technologies being proposed to help intellectual property owners monetize their investments - what exists is ineffective, inefficient, or threatens to slow the Participation Age. So our view is that it's time for a change. That DRM is useful, but not in the draconian or dictatorial forms that have been discussed. DRM, to us, is simply a security model for intellectual property. And just as accessing your bank account should be protected by a security layer, so should accessing your data - as just another form of IP.

Our view is that an effective solution to the challenge of ensuring IP owners can manage their own security and access controls - and compensation, if that's what they seek - must make a few basic assumptions. Let's outline them:

There is no one busines model for all intellectual property.
With Sun giving away Solaris and its source to the world, we clearly believe the open source model creates the most attractive long term business model for us. But there are some software companies that don't share that view - they're entitled, it's their business to run. Some media companies have started giving content away - others charge prior to download. Others offer monthly or yearly subscriptions, or offer all you can eat season passes. Some want you to pay everytime you consume (personally, I think that model will yield a very small viewership, but that may be ok if that's what the owner desires - a hedge fund manager giving hints and tips to other hedge fund managers may be worth $50,000 for a single view). The point is there is no one model - and any DRM solution that presupposes one model is bankrupt at the outset.

Thus, we believe an effective DRM solution must be capable of adapting to a broad and ever broadening spectrum of ideas and business models. There needs to be a neutral language that expresses a broad spectrum of choices made by IP owners - and at some point, there will be a need to implement such rights expressions in software. We see what we're doing in the Open Media Commons as being a possible source of such software. We also see the need for the owners of IP - whether it's the name "Electronic Frontier Foundation," or a speech that sends chills down your spine - to be exclusively in charge of charting their own destiny.

There must be no royalty or patent risk.
I was with a large telecommunications customer recently, who told me how excited he was to be partnering with a movie studio to deliver first run movies on handsets - until his general counsel received notice from a patent authority claiming they had tripped over patents associated with their encoding and playback mechanisms. The licensing authority had asked for $1 per movie stream, or a percentage of their annual revenue. Which would effectively shutter the opportunity. In our view, the market wants the choice of not having to submit to such trolls and tollgates.

And therefore, we believe an effective DRM solution must be available to the world without royalty or patent risk. In order to ensure continued interoperability and freedom to reimplement, we also believe the solution should be available in open source form, under a files based license that in no way threatens to interfere with original ownership.

There must be no hard linkage to any one device.
I was with the Chief Executive of a music company recently, who told me how thrilled he was to have a growing percentage of his revenues being derived from digital distribution. But there was one caveat - 95% of the digital distribution came through one vendor's product and service (guess which), the owner of which had let him know his royalty stream was being radically reduced, unilaterally, in a new contract. No negotiation.

And therefore, we believe an effective DRM solution must be available to IP owners on all devices - not be hidebound to one.

There must be no hard linkage to any one media format.
Another studio head let me know how adopting Windows Media only allowed him to compete for customers on the Windows platform, which with Vista coming up, was looking like Microsoft's exclusive platform to monetize - he wanted a neutral DRM, and an open encoding standard. He very much wanted a "write once, run anywhere" platform, similar to the web standards popularized by products like Firefox, Java and other network standards. We obviously agree.

So for the record, I want to reiterate a few statements from above. I believe IP owners should be in charge of how they monetize their IP. You have every right to choose the license under which you'll make your IP available. You have every right to monetize that asset as you see fit - and to control its distribution, if that's what you seek. I support the 4th amendment.

Do I believe the government has a role to play in selecting a technology? No (and it wasn't my idea to get them involved in the issue, you'll recall). Do I believe they have a role to play in progressing the debate, and in driving a neutral standard for interoperability? As they've done for electricity, for water distribution and for every other utility that's preceded computing? Yes, I do. And as we continue to drive DReaM and the Open Media Commons, as we've done with the Liberty Alliance, we expect they'll want to get involved. We're looking forward to it.

Posted on 10:39PM Aug 26, 2005 | Comments[3]

Share this post  | | | | | |

The Analog Hole in your Datacenter

I recently met with the Chief Technology Officer of a big media company about something called the "analog hole." The analog hole, as its called by that industry, is the vulnerability of a digital asset to piracy when it passes into the analog world - a digital camera can be used to take a picture of a copyrighted picture, for example. Or a digital camera can record a movie in a theater. He wanted to talk to me about supporting legislation that would mandate proprietary technology to detect invisible watermarks on content to stop its duplication or redistribution - no watermark, no playback (or distribution). From his vantage point, this would ensure he and his peers in the industry would be fairly compensated for their content, and pirates could be kept at bay. Legislation would mandate such a technology be included in all "computers and computing devices" with analog inputs or outputs (try to think of a device that escapes).

Rather than provide a response in the room, I turned a question back to him. First, the network you're supposing will deliver a movie to a theater or a camera to a file server is the same network I'm presuming will run throughout your datacenter. On the internet, it's tough to distinguish a feature length movie from a data warehouse application (bits is bits) - so would your datacenter folks support the tech industry certifying content behind your firewalls with a digital watermark? In running business systems?

Thinking as a blogger, whose rights are we seeking to protect? In the Participation Age, individuals are as likely to create the news as consume it. Maybe moreso. We saw this in London, where individuals were media outlets. (Paul Graham has a few interesting thoughts.) I, like millions of others, now produce, and distribute, movies from my phone (yes, the quality leaves something to be desired, but more due to a lack of artistic prowess than image fidelity).

So any attention, scrutiny or governance surrounding digital media would have to comprehend not only the teenager in the movie theater scenario, but the reciprocal extension of those models to the digital assets and platforms produced by companies like Sun Microsystems, Inc. (whose software distribution models are moving toward free and freely copyable), as well as those of "Schwartz Productions, Inc." (and the blogosphere, more generally - where economic interests are as varied as network end points).

On the former question, related to DRM in the datacenter, he said he'd run it up the flagpole with his IT folks and get back to me.

After a few days, I got a response. He'd spoken with their CIO, who dismissed the relevance of my proposal to manage all digital assets under the same scheme. "You'd have to start by proving I've stolen something."

Hm. Sure wouldn't want a double standard.

Posted on 12:25PM Aug 20, 2005 | Comments[4]

Share this post  | | | | | |

From the Bizarre to the Bazaar

How's this for bizarre. A friend called on Friday saying her sister on the east coast saw me mentioned on David Letterman. Right, surely they were confusing me with the more media relevant Jonathan Schwartz. I didn't think much of it (and besides, I'm supposed to be taking a vacation this week, and I didn't pack a search engine).

So this morning, don't ask why, I zipped over to CBS.com, and sure enough, there I am (see #8). Is this my 15 minutes of fame? I've yet to see the video, but if they used the photo only my mother could love, I'll need to have a new one done before Oprah calls.

On a more mundane note, I was in Portland last week, hosted by Tim O'Reilly and crew at OSCon. I was a bit early for my Q&A with Nathan Torkington, so I had a chance to listen in on a few of the speakers before me.

Andrew Morton (of Linux kernel fame) gave a low energy speech in which he said "Operating systems are commodities, now we're moving to commoditize the rest of the stack." He made those statements as if to suggest "commodity" meant "no longer of value." I don't think Andrew understands commodity markets. Here are a few: oil, gas, financial services, telecommunications and electricity. Commodity markets are the biggest and highest value markets in the world - because they represent products and services for which there's global, perpetual demand. They tend to be markets won or lost based on research and development (check out the largest R&D spenders in the world, they're almost all serving commodity markets). Computing may be a commodity - computers, and their operating systems, are most certainly not.

So when Andrew took a couple of obligatory potshots at Sun, saying "they should just give up on Solaris," I thought to myself, he and I (and 2 million licensees) clearly see the market differently. A point amplified when I met with his employer after my speech to engage them (and their sponsors, many of whom support Solaris already) in partnering to drive the adoption of Open Solaris. Stay tuned, good things are underway to grow the marketplace, and tear down the walls. And OSDL may play a role.

After Andrew, Jeremy Zawodny gave a very compelling speech on the usage of open source at Yahoo! What was interesting to me, beyond the listing of all the open source projects used at Yahoo!, was the dichotomy in their business - between the usage of open source software, vs. the creation of proprietary products and services, like FlickR and Yahoo! Messenger. For which the code is obviously not available.

Don't get me wrong, I absolutely love FlickR (frankly, I'd love a conversion service to help me ease my family's migration from other photo services) - but it once again pointed up the value of diversity in the ecosystem. There is no one hammer, or license, for all nails. Just as there's no one software product (or kernel, Andrew).

The interview with Nat has yet to be posted - and although we were nowhere near as entertaining as Letterman on Friday, it's worth a listen. Like the start of my week off, the questions ranged from the mundane (what's Andy Bechtolsheim like? Answer: Tall.), to the bazaar [sic] (why didn't you GPL Solaris? Answers: too many of our customers - especially those embedding Solaris - preferred a files based license; and we didn't own all the IP within Solaris to predetermine a singular license for every contributor).

An interesting start to a week off.

Posted on 10:10PM Aug 07, 2005 |

Share this post  | | | | | |