Pluralism and progress

Kim Cameron makes an excellent repost to Doc Searl’s comment that “two ways to do identity is one too many.”

Kim writes: “The problem is that - at the same time - one way to do identity is too few.  And this is what explains why the creation of a universal system of identity is one of the greatest challenges blocking the evolution of technology and the virtual world.”

Kim goes on to cite the idea of pluralism as a law relating to identity systems, but it is more widely applicable than that. One way is always too few ways to do anything. Evolution demands diversity, survival all the more so. Doc goes into this at length, adding:

First, politics hasn’t spoiled identity yet, because identity has hardly started as a topic, much less as a useful service for anybody on the Net. For all the effort that’s gone into identity, it’s still newer than RSS was when it still stood for Rich Site Summary (or whatever it was) rather than the Really Simple Syndication all of us bloggers employ today. And while the Liberty Alliance its participants deserve credit for all their work around federation, that’s still mostly BigCo stuff that benefits us as individuals mostly on the back ends of our relationships with various companies. We need something that works for us, as individuals, in a simple and obvious way.

Second, identity won’t happen as a service unless it comes up from the grass roots, from independent developers, the users who support them, and the big guys who follow indie developers and users into the marketplace. (Think about how the big publishers have deployed RSS, for example.)

I argue that politics is precisely what is saving identity from BigCo solutions.

In a situation where BigCo solutions are eschewed either for their failure to convince users or users’ concern about ceding power to a centralized entity, it is grassroots dissent that prevents a victory by those entities. It would be a good idea to recognize that pluralism is politics and embrace the debate as essentially good. I know what Doc means by “politics,” but that’s not politics, it’s market power, which is unrepresentative when existing influence is used to prevent competitors from introducing alternative options into technical standards or products.

If we get past damning politics by recognizing that the structure of the debate is what defines its value—an open exchange is far more egalitarian than a closed, industry-led one—we can tap the human passion for participation to reinvigorate standards setting and much more.

I will now remove my Semantics Police hat.

Now, from my identity thinking cap: An identity infrastructure needs a fundamental schema, but it must be extensible, so that every system introduced can communicate basic information while adding functionality. The debate need not revolve about what is included, rather it is best conducted in terms of what should be excluded in order to reduce that fundamental schema to the absolute essentials. The result will be a system in which we move between identity infrastructures, becoming more or less identifiable in different contexts; the idea that some of us will be identified one way and others another—that a competition between infrastructures will result in a winner—is misguided.

Creative and apart, but part of the world

I like this. Hugh McLeod does a nice job of explaining quickly and entertainingly how hard it is to be creative. I particularly agree with this:

“Don’t try to stand out from the crowd; avoid crowds altogether.”

Avoid labels and joining, too, while living in community. Folks have asked what my email handle, godsdog, means. It was originally “coyote@well.com” and somewhere along the way when I moved Netcom, the handle “coyote” was taken so I chose “godsdog,” the Hopi nickname for coyote. There’s a trickster connotation, which is what most people think of when I explain this, but the real reason is that a coyote is a shy dog that stays at the periphery watching everything. It’s not that I am frankly anti-social, but I’ve always agreed with Groucho Marx’s maxim: “I don’t care to belong to a club that accepts people like me as members.”

Two more Herring postings on tech business in 2005

For your consideration at The Red Herring:

  • Apple’s one-percent solutionApple Computer’s market share is still meager in contrast to those of Dell, Hewlett-Packard, and others, but the company has reinvented its computer business on the shoulders of the little iPod. During 2005, Apple clawed back slightly more than 1 percent of U.S. PC market share and is projected to finish the year at 4.9 percent of PCs sold.
  • Microsoft’s very bad yearMicrosoft, the largest and most feared software developer, was stopped in its tracks this year by the hobbits, elves, and free people of the IT world. While it isn’t in the intensive care unit, the Redmond, Washington-based behemoth has scars from the battle and is looking ahead to 2006 as a turnaround year as important as its legendary turn to address the World Wide Web in 1995.

My brief decapitation to cap a decade of savings!

The site was temporarily inaccessible for about 20 hours, as I switched to a new DSL service. For the first time since 1996, I don’t have a synchronous T-1 level connection to the Internet from my home, but I’ve also seen a vast savings over that time that, in these tight economic times, justifies surrendering some upstream throughput. However, I am paying 93 percent less or connectivity than back then.

I moved from a $310-a-month 1.5Mbps/1.5Mbps DSL connection to a 3.0Mbps/768Kbps DSL connection that costs only $110 a month. For about $2,400 a year in savings, I can live with the slower upstream connection and by offloading some fileserving to a hosted site I have, provide the same or better site performance for you.

Back when I got my first T-1, in summer 1996, the circuit cost $1,700 a month, or $20,400 a year. It paid off in spades, allowing me to experiment with Web sites while offering my writing and research customers very fast turnaround on work, because I was not sitting and waiting for a dial-up connection to load pages. Where another writer might have to wait five or ten seconds for a page to load—longer with a lot of graphics—I was getting pages in a second. Over the course of 1,000 pages of surfing, I might pick up 100 minutes of extra time that my competition lost to waiting for information. That allowed me to write more and, thus, earn more.

The fast connection, for instance, allowed me to take on producing a daily news site, ZDY2K, for Ziff-Davis, from home. When I joined ON24, it allowed me to work remotely a day or two a week, keeping my family in the Northwest where we wanted to be rather than picking up and going back to San Francisco. Only when I joined an industry that valued face-time over delivery of intelligence—investment banking—did my remoteness become an issue (along with the general decline in the business anyway, so it wasn’t a singular problem) and I moved on, back to living over the connection rather than flying to SFO every Monday and home every Thursday, which I’d been doing many of the years since 1993.

Early on, I learned that synchronous connections were powerful because they offered you a voice equal to or larger than a corporation employing hundreds. So why abandon synchrony now? Simple, with the distributed computing facilitated by the Net, I can put files on servers with faster full-time connections while benefiting from an almost 100 percent increase in download speeds—for $200 less a month. In less than a decade, my home broadband expenses have fallen by $19,080—93.5 percent. That’s pretty phenomenal and awfully bad news for the folks who build and provision broadband services, since if this pace continues it suggests that by 2014 the typical home broadband connection will cost somewhere on the order of $16 a month.

No one has a single connection to the Net anymore, because we maintain many virtual connections through hosted services. There’s an important distinction in this: I am not talking about keeping separate accounts to connect, as we did with dial-up and dedicated circuits in 1996 and still do with wireless, dial-up, dedicated and roaming services today. Today, we also maintain presence—access to our knowledge and ourselves—through Web services.

The cool trick now is learning to distribute your upstream communication to maximize the performance for a lower price. At a time when connectivity is completely commoditized, plainly evident in the drastic slope of my cost curve, it’s not how fast your connection is that matters; it’s how you use multiple connections to stay jacked in to the things you need to know and, more importantly, how you allow others to stay jacked in to you.

Personal Grid Computing

Luc Julia, the founder and CTO of Orb Networks has engaged me in a good-natured dialogue about my posting last week saying that the coverage his company is winning is setting it up for a promise to the consumer it can’t keep. He says I need to see the technology work to believe it and so I must, but I also found out more that should be put out for consideration.

The “streaming live TV” demo touted by the company and by AlwaysOn’s Tony Perkins—the crux of my concern that the company is over-promising because it is dependent on too many network hops out of Orb’s control—requires a Media Center PC and works with a PC with an unsupported Hauppage tuner card, but it is more or less the fluff on what is a meaningful system for sharing media and handling DRM issues. The demo described by Tony Perkins did take place over a 44 Kbps wireless network connection from Cingular, according to Luc Julia.

As I’ve explained, there are other players moving into this market, some with service ambitions similar to Orb’s and others with the goal of being acquired by a Cisco that wants to build media sharing services into their products.

I also made the point, which Luc Julia agreed with, that describing this system as “distributed computing” is somewhat problematic, because Orb’s server does play a role—both as a sort of router and providing key management services that can allow secure access to subscription content. The Orb network offloads encoding of audio and video to the user’s PC, but that is not truly distributed computing because the system doesn’t work if Orb’s server isn’t available.

Here’s where I think we are going with all these systems: Personal Grid Computing. All the computational devices we own will work in concert with a conductor to orchestrate the process (Orb’s server plays that role here) of allocating processor cycles to specific tasks, such as calculating optimal screen dimensions, bit depth and encoding the file to get a video file from a PC to a handheld.

As I’ve written here and here, this is a market that will quickly commodify and I wonder about Orb’s business model. But as Luc Julia said, it’s not likely the first business model at any company will exactly right and I’ll give him that. Just don’t hang the business on live TV streaming, there’s so much more to what a personal grid computing will do. And keep an eye on this business, because it will grow—both in terms of the number of competitors and the amount of money invested—during 2005.