Skip to content


Not enough to do.

In ancient times, pre-civilization included, was there such a concept as joblessness? Was there ever not enough to do in the world? I severely doubt it. So this notion of unemployment in its essence is a purely capitalist notion, in that providing for oneself is an intrinsic requirement outside of capitalism. Living off of savings or other forms of insurance is not a realistic option outside capitalist society, among the few lucky enough to gain capital accumulation. Without capital, in a capitalist society that has used the notion of private property to exclude all earlier methods of survival and self-sustenance (i.e. hunting and gathering is not permitted outside your own property lines), survival itself is impossible. A homesteader in history might have once had such a bountiful crop that his/her family didn’t need to tend it the next year, but I severely doubt they took the risk that their luck would hold so well for another year thereafter. The only people who could afford not to work were those subjugating others into providing for the group’s survival unassisted, but this “wealthy” class considered their acts of oppressing their fellow humans a sort of “tending” work. These wealthy tyrants have always regarded “tending to their subjects” as similar to a farmer tending to the plants of the field. The names have changed; but serfs, slaves, proletariat, and employees have all fit similar roles. I suppose even this “management” class wouldn’t ever consider themselves “jobless.”

So, can unemployment really ever be caused by a lack of things to do? Is the world that serene and settled? I would argue instead that joblessness is a state invented by capitalism, and fostered by its dependency on the profit motive as the only source of “worth.” Base survival is not a worthwhile endeavor in capitalist investment, however necessary it is to the survivor. In fact, if capitalists can make any necessity, including survival, more difficult for the “customer,” the better for their profits. “Captive markets” are the most wonderful thing in the world to a capitalist. It doesn’t really matter to them if their captives are starving or unhappy, so long as they keep on paying in some way.

Posted in Against Monopoly.

How do you beat any Operating System Monopoly?

To end Operating System (OS) monopolies in any hardware market, make the OS irrelevant to the operation between the hardware and target application. That is the only real purpose of any OS software. Everything else is taken care of by libraries and software, which can be developed using cross-platform API and SDK help.

The only OS that will matter on future PCs will be the embedded Hypervisor, the Virtual Machine host of the device. The guest OS of each target application will become much less important. “Fat” Operating Systems will merely seem like “application spaces” from there, similar to current AJAX-y  ”WebOS” environments, like on a Palm Pre. In fact, a cached network based guest OS could gradually become indistinguishable from a local storage OS.

Eventually, most new Operating Systems will be designed to depend completely on Hypervisor virtual devices, and forget about supporting a broad range of hardware. Hardware manufacturers will target drivers at a limited set of Hypervisor platforms, and wont have to deal with any other OS directly, unless optimal performance is a concern. Interactive media, and other high-performance applications, will tend to favor platforms that give them the most direct control, with the least overhead, and the least hardware support issues. High performance stand-alone applications, like interactive games, will become dependent on low-cost and low-overhead guest OS environments, which will drive proprietary and high-cost OS options out of that market segment.

This will create a great opportunity for the interactive games industry: turn any PC into an open game console platform. Allowing the game to completely control the OS environment it runs on opens up a wide range of anti-cheating tactics, that are more seamless than ever possible before. The game can come with its own custom OS, and load into nearly any thin Hypervisor host device, using hardware based encryption. Additionally, the Hypervisor could provide a set of virtualized drivers and emulators, to guarantee a minimum hardware performance for guest OS applications or games, even if they don’t support the hardware directly. Otherwise, minimum performance across configurations would be provided by open interface standards definitions, and by third-party standards testing services. Brand recognition will surround standards compliance instead of  any single proprietary brand.

So what do you think — do the links below reveal a trend?

Flash Hypervisors provide a Virtual Machine in BIOS.

Virtual Machines now have direct hardware  access. [PDF] (note pages 6 and 8 )

Data Parallel Virtual Machines for GPUs. [PDF]

Intel Larrabee reveals future GPU architecture.

You can run Mac OSX on almost any Intel PC.

DirectX support is being provided cross-platform via WINE and other projects.

Cadega provides software shaders for real-time rendering in CPU.

Linux can be run on a consumer Sony PlayStation3.

Linux can also be run on a XBOX360.

Linux is one of the only reasons to still run an old XBOX.

The PC Gaming Alliance was practically DOA.

Alex St. John: “Consoles as We Know Them are Gone.”

Alex St. John: “Vista Blows.”

Unreal creator Tim Sweeney: “PCs are good for anything, just not games.”

Tim Sweeney, Part 2: “DirectX 10 is the last relevant graphics API.”

Posted in Against Monopoly, Games, Operating Systems, Virtual Machines. Tagged with , , , .

Is DRM ever worth the cost? Probably not.

This is a long and unwanted answer to a too-simple question posted on LinkedIn about DRM.

OK, feedback please. Is an average anti-piracy protection for a Nintendo DS title of over 70 days safe from hacks (and counting) a good result?

It takes two big questions to answer whether Digital Rights Management (DRM) is worthwhile or not. I am skeptical that any DRM method can provide a Return on Investment (ROI), regardless of time span between release and hack.

  1. How much more did it cost to implement, test, and ship the DRM protection? This is the Total Cost of Ownership (TCO).
  2. How much more money did you make off the product for it not being “hacked”, for however long that protection lasted? This is the Return on Investment (ROI).

ROI-TCO = value

Always remember that anyone who can extract profits from the TCO side of this DRM value equation will attempt to inflate the ROI side, often in ridiculous and unrealistic ways.

Both of these are difficult questions, and I doubt anyone has sufficient data to be completely sure of either answer. The TCO of DRM tends to be under-reported in standard accounting, because complications of DRM can introduce production and testing problems at many levels, far beyond the cost of integration alone.

The ROI of DRM tends to be far over-estimated, because it’s doubtful that “pirates” would ever become paying customers if “free” (in exchange for time, equipment, and know-how) access to media wasn’t an option for them. It also hasn’t been established what percentage of “pirates” use protection-stripped copies as a try-before-buy system, or even as a backup for DRM encumbered purchases already made. Common answers to item 2 also tend to under-estimate the number of lost sales to consumers who are resistant to buying DRM encumbered materials, as the DRM tends to cause more problems for legitimate users than for the pirates who know how to strip it or work around it. Any DRM or license restriction that strips resale value also limits first-sale value estimates by knowledgeable consumers. In many cases (EA’s Spore comes to mind), the ROI from sales lost and gained due to DRM alone could be a negative value total.

A study in Norway actually found a positive correlation between music “piracy” and a higher average rate of legitimate music purchases, which supports a try-before-buy theory of digital piracy.

Depending on how effective your demo distribution methods are, assuming any “piracy” is actually an attempt by consumers to try-before-buy is more valid than assuming each copy is a “lost sale” (which has never been proven, in any study, of any media or market context).

For-profit piracy actually comes under counterfeiting and trademark law, so law enforcement in different governments might have methods of estimating losses, based on seizure rates of related counterfeit materials. It may be worthwhile to search for records of federal seizures of counterfeit media in your target markets. It might be better to just avoid markets where the counterfeit rate exceeds the legitimate distribution rate (i.e. China), except where any sale (however small) is better than no sale at all (i.e. markets that don’t impose any additional costs for localization or distribution).

In any security implementation, the ultimate measure of success is when the ROI > TCO, where the ROI can be roughly estimated as projected loss without protection minus net loss with protection. Only real revenues can be compared with any accuracy, so these ROI estimates are never accurate. Not every copy is a lost sale, so not every copy can be counted as a projected loss. In the case of DRM methods, ROI < TCO in most probable scenarios. This is especially true on proprietary hardware formats like the Nintendo DS, where the cost per copy is much higher than standards based digital media, so only the most dedicated copiers even make the attempt.

Posted in DRM, Security. Tagged with , , .