START TYPING KEYWORDS TO SEARCH OUR WEBSITE

Too much is not enough

Posted: 06.14.04

Ever heard me say that? Ever been near me for more than 20 minutes? And no, I'm not talking about supersizing or the world's growing problem with obesity; I'm talking about technology.

In computer graphics it's always been true, and probably always will be. But it applies to other things as well. Take standards, for example. Ever hear the comment, "I love standards; there's so many to choose from"? It's a double entendre. It could be taken cynically and sarcastically to mean that there are too many standards, and on the other hand it could be taken to mean that it's a good thing we have lots of standards so things will work together. Sure, we'd all like life to be simpler and easier, but stop and consider that we, in this industry, are one of the primarily forces of making things more complex, so you can't have it both ways.

Nuts and bolts

SellersIn the process of making things more complex we need to at the same time tame them and keep them under control. I'm talking about more than just Asimov's three laws for robots; I'm talking about all the interfaces, both hardware and software and hardware to software. One of the defining moments in time was when in 1864 William Sellers (1824-1905) gave a speech at the famous Franklin Institute in Philadelphia (and if you've never visited this place I strongly urge you to go). The speech was titled, "On a Uniform Standard of Screw Threads." Sellers argued that nuts, bolts, and screws were all individually made by hand, and that they still had to be hand-matched until a fit was found. A bolt or a machine screw made in one machine shop would not fit a nut made in another machine shop. And nothing made in America would fit anything made in England or anywhere else. By 1886Ð87 we had the "Standard for the Diameter and Overall Dimensions of Pipe and Its Threaded Ends," and the industrial revolution took off. Back then, the machine tool industry was like Silicon Valley is today—a driving force of the most modern technology of the day.

Today we're building the nut-and-bolt standards of the cyber world. And yes, there are a lot, and at the same time there aren't enough. I could fill this entire issue (with small type) of the standards we don't have, and need. So anytime someone proposes a standard your first reaction should be welcome and thank you. Your second reaction should be, and what special interest group does this serve? And if you can't find a conspiracy or an egregious self-interest behind it, then embrace and support it—oh, and also don't fight it if it doesn't serve your egregious self-interests.

New standards by the truckload

This week we saw several new standards offered. There was the PCI Express from Intel and its buddies, a company spec dressed up to look like an industry standard—and that's OK as long as the majority of us agree to it. Khronos has got some things cooking that will be revealed in the fullness of time, and the Intel-led consortium Digital Home Working Group (DHWG), now known as Digital Living Network Alliance (DLNA), came down the mountain and delivered us a 204-page specification book, which, my cuteness put aside, is very much needed and wel-come. And, were I not so lazy, I'll bet I could do a little web search and find a dozen more standards introduced this week. (I actually did type in "new standards" and got 10.5 million hits—no joke.)

I was asked at a recent speech I gave called "Convergence is here now, really" how I could say that when I've been saying it for 10 years? The answer is easy: convergence is here now because of Moore's Law and standards. For 15 years at least we knew what we wanted to do: marry TV, audio, computing, communications, and ease of use into one consolidated box. We just couldn't figure out how to do it. But boy, did we try—talk about shelfware.

In the last year or so all the pieces started to fall in place like Legos, and one of the things to come out of it was Microsoft's Media Center. iPod was another, DVD recording another, and so on.

The big frontier is the handheld thingies, the market that has so many people excited because of the huge numbers it represents. Everyone is going to own a handheld multimedia device that can also be used to make phone calls—our new personal companion (the real PC). And if the industry doesn't come together with some real open standards it's going to be a bigger zoo than TV or the (old) PC ever was. There are some rays of hope with MIPI, and mobile-phone and chip makers are under pressure to give up cherished proprietary technologies in favor of generic interfaces for tying cell phones to multimedia peripherals such as higher-resolution displays and cameras. Open standards made the computer industry grow, and although it's hard to talk to the handheld industry about growth, the market will fragment if we can't get more open standards into it.

So standards: "I love standards; there's so many to choose from"? I say, Standards—give me more—too much isn't enough.

Epilog: And I'd like to lay to rest once and for all that the honorable Mr. Sellers never did say, "Screw this" and walk out in a huff making perhaps one of the first non-sequitur puns.