Eyeballs: March 20, 2006

More food for thought as I close a bundle of Firefox tabs:

Buried in the Wired profile of Sky Dayton is this great line from SK Telecom's global strategy chief: "The ultrahip crowd and the ultrageek crowd. That's the target SK Telecom was dreaming of. I call it technosexual." I'm ultrageek, but I don't think I'm ultrahip. I guess I don't have to ask which half of "technosexual" that makes me ...

Sky Dayton's new venture in the US is Helio. They're an MVNO, a virtual mobile phone company: they don't actually own the transmitters. They're an end-around some of the grinding slowness and stupidity of the carriers (but not all: how long until I see these types of features on my US mobile phone?). As I see companies like Yahoo! putting all their services out in the form of APIs, I wonder whether they're building the infrastructure to create a new class of companies: VPOs--Virtual Portal Operators. Imagine a world where you have access to Yahoo!'s data and services, but you get to write the code and UIs that filter and shape the portal experience to the needs, technology, interests, and abilities of your users.

Speaking of Wired News, my new favourite journalist Clive Thompson has written a great piece on HUDs. Heads-Up Displays are those ever-present overlays on games that show you your strength, score, remaining ammo, etc. Clive talks about how the quest for reality is leading developers to try and hide the HUD, giving you the essential information through gameplay rather than the artificial display. But paradoxically, that leads to games that break the suspension of disbelief (as, for example, you run out of ammo unexpectedly). The HUD, a glanceable ambient information display, turns out to be something we want in other areas from combat soldiers to email grunts. After all, dashboards are basically HUDs for your car, and we're putting dashboards into all sorts of applications now.

The Economist had an interesting piece on open source. The most interesting piece, though, was this fantastic graph showing the different rates of growth for talkers vs doers:

It might take a village to raise a kid, but it doesn't take a village to paint a bikeshed. Managing the ratio of interested vocal people to talented coding people is a challenge for any governance system.

More interesting open source datapoints from this ZDNet story on open source security defects. Most open source components of the LAMP stack had fewer defects per thousand lines of code than average commercial projects. The only exception was PHP. Disturbingly enough, the Amanda backup system had the highest number of defects.

Projects I've got my eye on: Splunk (will open source their product soon), XORP (the extensible open source router project), and the reshuffle at Xen once they realized that their distributors were also their competitors.

We at the Radar have been watching XUL slowly take off. First Songbird, now these cute applauncher (warning: you'll need Firefox or Mozilla to use these). Songbird's a media player built on XULRunner, an interesting platform that gives you some sweet features like cross-platform deployment and easy upgrades for free.

This squirt-gun tournament looks to be major fun.

I enjoyed Simon Willison's (Re-)Introduction to JavaScript tutorial from ETech. Beautiful slides, using Flickr photos to great effect. JavaScript isn't the only language I've been dipping into: I also moved from an infant's grasp of Python to a toddler's grasp as I began exploring Orange (the datamining toolkit, not the mobile network operator). It's outrageously simple to use. I'm in love!

Only three things so far this month have cleared the "important and generally relevant enough to share with my Dad" hurdle, and this was the best: the Guardian's coverage of how cheap goods affects the way we think and live. There's the destruction of the class system of shopping (something I've noticed here in New Zealand since I've been back), as well as our inability to make rational decisions about cheap goods (I only need one shirt but I'll buy five because they're so darn cheap!). I think it's maybe something like a sweet tooth: historically bargains were rare and so we were taught to savvy shop based on price alone--some day we'll need it, so we'll store it until we do. That worked well when low prices were rare, but now they're ubiquitous we're buying crap we don't need and now having to pay for more space to store it. I know I felt liberated when I moved back to New Zealand and had to live out of a suitcase's worth of clothes for the two months it took for our shopping container to arrive. The sad truth is that while I've unpacked all the clothes I brought back from the US, I'm going to put them up for sale at the next local rummage sale. I don't need them, and I don't want them. They're not just domestic clutter, they're mental clutter as well.

This is the same as our inboxes. My inbox now has nearly 200 messages that I sincerely intend to reply to, but rationally know I can never do so. It's now way easier to create an opportunity than it is to execute on it--email is like cheap t-shirts at Wal-Mart. Not that I'm saying we should literally charge for email, but rather that we need to implement some kind of virtual tariff to impede messages. How can I weight the inbound messages (the way spamassassin does) to reduce some of this opportunity cost? Is there a mental trick I can use, like "I will only send fifteen email messages per day"? Getting Things Done turns you into an effective shirt folder and storer, but I wonder whether there's another way? I'm amazed I managed to write that paragraph without using the term "attention economy" or citing Linda Stone. D'oh!