Nerd time Issue 14

I use to work at a research lab, and while they do know what’s going on with military tech, they don’t really get news of consumer web technologies. I read up on it alot while I was there, and I still do now. I started a little mailing list after I left to keep them updated, and that’s what became Nerd time. Here is the latest for your enjoyment.

Hey all,

Happy 2008! It’s been over a month since the last nerd time. At first, it was because not much was happening. I’ve found that Silicon Valley(especially the software/web side of things) is much like Hollywood in the sense that there are fads and fashions that people make ruckus about and then ditch after a month. And it’s like sensationalism news when nothing noteworthy is happening, but people still find lots of trifle things to make noise about. So even though we attempt to filter it all through our favorite feeds and social news sites, it’s still a chore.

I met up with Ray the first time since College, and he said he feels out of touch with the tech world, even though he’s a chip designer. All he has to do is read nerd time! You get my own biased, hand-picked version of what’s noteworthy.

Mike and Nick, check out Johnny Lee and Buglabs, being hardware freaks that you are.

As usual, the easy stuff is up top. If you want to stretch your brain, go further down.

Johnny Lee’s wonderful world of HCI
Wiiremote VR tracking, Wii multitouch, Automatic registration of projection screens. It’s this researcher’s venture into HCI. more melding of computing into everyday lives. The videos are more self explanatory. If you skip everything else, I suggest you don’t skip these.

Facebook, Google, and Plaxo join DataPortability Workgroup
I strongly believe that a users’s data belongs to the user, and they should be able to take it around with them. I was pretty excited about Google’s announcement about OpenSocial platform around October or so. I haven’t personally taken a look at it myself, but according to pips here and there, it’s been said it’s incomplete and there was a case of a security hole. But this announcement of fb joining a data portability workgroup is good news, as making your data portable with you across applications will go a long way in making using computers a lot easier.

This is a startup that you redirect all your snail mail to, and it opens and scans it in for you. You basically can read snail mail like you do email. You can even choose to recycle junk mail (good for the earth), and shred important documents. Neat idea, though of course, it all has to rely on trust of the company. As the social-political climate of the US over foreign energy dependence gains awareness with the push for ‘green’, we might see more companies like this that tries to move things into digital to save on green.

Google’s philanthropic arm going to develop clean energy cheaper than coal.
Just because something is digital doesn’t necessarily mean it has a small environmental footprint. The servers it’s running on might be burning coal. I think it’s rather big, actually. It remains to be seen whether Google will democratize energy production, to shake up a very traditional energy industry, or simply join as another player. Regardless, this venture makes sense for them, as they run gigantic server farms.

Prediction Markets at GooglePlex study organization information flow
Google uses internal prediction markets to gather data on what’s going on. Prediction markets are basically using the mechanics of stock markets to predict things like, “is google stock over valued?”, “Will the U of I win the Rose bowl?”, “Will Comet win mindshare with developers?” This particular article talks about mapping information flow in organizations. I think as the world moves faster and faster, mid-sized companies are exploring more flexible organization structures to be able to respond quicker to changing markets. You’ll probably see more studies into company organization in the future.

This is interesting because it’s another indication of how real-life is merging with digital. Any website monitors its traffic to know which pages are more popular than others. It uses that as implicit feedback to what customers like. This does the same for physical storefronts. Using bluetooth on customer’s cell phones to track foot traffic, they can see which racks and displays are selling more. As more and more sensors enter into the world, we’ll be able to do analysis and better understand our lives. I won’t talk about privacy concerns, cuz that’s just a can of worms.

Amazon SimpleDB and computation on tap
I wish I brought Amazon stock. The more I find out about Jeff Bezos, the more I understand that it’s not really a bookstore that he built, but a philosophy. Anyway, Amazon, in this past year has been releasing “computation on tap” They basically sell distributed computing and storage services to developers and startups. Lots of people have taken advantage of it, for good reasons you can read elsewhere. In the past month, they released simple db, which is really just a fast index db. With computing on tap, it enables small teams to do large things. If computing and storage is a commodity, I suppose the value becomes what you can build with it.

WiTricity wireless power
Wireless power has been in development for some time now. I had mentioned it in one of my earlier nerd times. This one, although just experimentally demo’d, is rather neat. We probably won’t see wireless power more ubiquitous until maybe 5 years later. I don’t quite understand it myself (should have paid attention in PHYS 112), but apparently, it’s not radiation like we’re familiar with. Instead, it uses magnetic resonance, so that only two electromagnetic devices with the same resonance will transfer energy efficiently, everything else, like humans, won’t.

This is a startup making mobile devices that you can hack. They plug together like lego pieces and run linux. People can prototype and experiment with different mobile form factors and fulfill their niche needs.

Quagmire 2D programming language
This is a 2D virtual machine, and you can watch it as it uses its memory visually on a 2D surface. What’s interesting is that the latter examples looks like chaos, which reminds me of the pictures you see in “A New Kind of Science” In it Wolfram claims that basic computers nowadays just have simple loop patterns rather than the complex ones you’d see in class 3 CAs. Problem is, we don’t know yet exactly what class 3 CAs compute better, and how to harness it, other than CA120 is turing complete.

Comet – AJAX’s cousin
Comet is using AJAX to push information to browsers, rather than having browser poll the server. The term was coined about two years ago, in reference to the bathroom cleaner. So you can get stuff like chat working in browsers, or like a stock market feed.

Sunspider – javascript benchmarking
Javascript is getting faster and faster with work being donw on the underlying engines underneath popular browsers. I’ve seen “lemmings” implemented completely in javascript. You’ll continue to see this sort of thing. However, flash, silverlight, and flex are still far ahead of the curve when it comes to interactivity of the web.

Version control in emacs
The new emacs 22 has svn version control built it! *high fives Ian*

Tumble logs are the half cousin of twitter and blogs. They represent a flow of conciousness digital media tidbits from the poster’s daily life. We might start to see more of this if it catches on. However, other than personal expression, I don’t exactly see a future for it. But then again, I was wrong about Friendster when it first came out.

Next gen ANNs
A talk on next generation Artificial Neural Networks. I haven’t watched it yet, but will this weekend. It promises to be pretty exciting.

Ycombinator in Ruby
I finally figured out what ycombinators are. If I had to take a stab at it, it’s a way to implement recursion when you don’t have named recursions, loops, iterators, etc, and all you have are functions and a couple substitution rules. It’s pretty neat, and will stretch your brain a bit to understand it. Traditionally, you only do this in lambda calculus, but having the code in Ruby made it easier to understand.

Advertisements Detecting and reducing power consumption in Linux Detecting and reducing power consumption in Linux

Power consumption was usually something of secondary concern for desktop computer engineers for a while. But not so today. When you’re cramming all those transistors in such a small space, operating at high speeds, power definitely becomes an issue. Now that mobile and embedded devices are experiencing their slow infiltration of our daily lives, power should be on the table for improvement. Google has been doing some work for this, and they now build their own power supplies, and even build their data centers where power is cheaper (old news, so I won’t link it).

Battery life is one area of computing that really is way behind. I’m still hoping for some feat of chemical engineering to save us…but hopefully, in this time of scarce energy for mobile devices will drive some creative hardware engineering.