Saturday, February 03, 2007
Sunday, November 26, 2006
How to never stop at a red light again
Here's the trick. With a little thought and effort, this trick can make it so that you'll never have to actually stop for traffic lights again. To explain the effectiveness, I first have to explain the opposite. Most people will do the hurry up and wait routine. This means that they are rushing to the next intersection in the vague hope of making it through in time. This works occasionally, but can result in a speeding ticket too. Moreover it results in lots of hard accelerations, and hard braking. It's tough on the car and the driver. Drivers doing this are thinking that they are beating the system because occationally it actually works. They aren't.
The best thing to do is to adjust your timing. Many lights do function on timers, and are staggered so that if you are going at a certain constant speed you'll be able to go through a series of lights without stopping.
The key is to gage how long green lights last, and how long red lights last, which you can do within a couple of intersections. If your approaching an intersection that is already green, and has been for a while, and you're roughly 15 seconds away from the intersection, your best bet is to slow down. It is about to finish it's green cycle, and will hit it's red cycle shortly. If you slow down, it lengthens the amount of time that it takes you to reach the intersection. With a little practice you should be able to reach the intersection without breaking at all.
Thursday, July 06, 2006
UAC on Vista will work wonders
It's fascinating to me that UAC has received so much flack from so many journalists, and testers, but I am going to boldly make the prediction that if it works the way Microsoft states, their flagship Windows Vista will have drastically fewer bugs per year.
Here's why: A little over 3 years ago (probably even more) MS released their latest server OS Windows 2003. One of the newest components in Windows 2003 was IIS6, which was a major departure from previous versions. To set up the new features I need to first describe how it was done in IIS5. IIS5 did not have this feature, and was a tremendously huge attack vector (think CodeRed, and the like). So their enterprise web server was hopelessly vulnerable.
Then along came IIS6, and with it greatly enhanced security. What did they do to enhance the security? Microsoft took a page out of Unix's book, and coded IIS6 so that it ran with guest-level priviliges, which greatly reduced how vulnerable it is to attack. Why you may ask? Suppose there is a buffer-overflow attack in IIS6? If it runs with guest privileges, this would mean that the attacker could only gain access with guest-level privileges. Hackers became disinterested in IIS6. In fact since it has been releaset, there are only 4 vulnerabilites available for it on Bugtraq. There are over 60 for IIS5.What does this mean for Vista? IE has been just as vulnerable to Windows as IIS5, if not more so. I recently did some research on a tool called "dropmyrights." It is an executable that uses the same mechanism as a bult-in policy of windows, and basically makes it so that you can run programs with reduced privileges. What I found is that when IE is run with reduced privileges, it makes it invulnerable to malware. In fact I could not infect Windows, or IE with any kind of malware when I droped its privileges down to the level of guest or even normal user.Each of these examples are reinforcing the idea the MS is on to something with UAC. Droping the rights that the windows desktop, and applications runs under will drastically improve Vista's security.Monday, June 12, 2006
The Korea Times : Anti-Aging Molecule Discovered
[Enter your comment here]
Read more at times.hankooki.com/lpag...
Saturday, March 04, 2006
PBS | I, Cringely . March 2, 2006 - Peering into the Future
This is a good article with good ideas about how to resolve the issue of how to distribute legitimate media through the internet such as TV, music, Movies, and other copyrighted media.
There is also mention of a company called "Peer Impact" which has a different and IMHO better way of distributing media. Details are in the article, but it got me to thinking a little about the TV watching experience. When Desperate Housewives comes on there is almost this shared experience going on, because you know that millions of viewers are watching it at the same time you are (depending on the timezone of course).When you have to download a show, you loose that to some degree. What I mean is, people may not be watching it at the same time because either their download didn't start at the same time as others, or for whatever reason. A sporting event is probably the best example. IT is happening in real-time, and there is no way to fast-forward it or anything. You can pause it and continue, but since it is live, you can't do much more. This brings up the question of how do you distribute sporting events. People want it real-time. They don't want to wait for it to get fully distributed in a peer network before watching it.The best way to handle this kind of thing is to simply multicast. Multicast is beautiful because no matter how many people are watching the stream, the bandwidth required for it remains the same. It is a 1 to many approach, versus the traditional 1 to 1 delivery of several copies of each piece of media to each connection. So as I was saying, this should and does work perfectly for real-time stuff.Here's where stuff gets really special though . . . Without knowing a lot about how the distribution system works with Peer Impact, might I make a free suggestion in that regard? In any P2P network, there needs to be a critical mass of people sharing the content to handle the number of viewers who want to watch it. To get that content out to the peers, traditionally, the system relies on one peer passing its data on to others. So to get a TV show ready to become available on the network may take some time to get cached on all of the necessary clients for distribution. A better approach is to multicast the cached content to all of the clients. This way all the clients are receiving the content they need to cache up at the same time.Now here's an even better idea. Take the same "Peer Impact" P2P idea, and make it into an appliance. That would be the best of all scenarios. The client now has the ability to access their favorite shows as normal, using this appliance, but instead of paying their traditional cable bill, they simply pay per-episode. Now, since their client has been running now all day, servicing content requests, it Peer Impact has been crediting the viewer with cash. The viewer now has the ability to put that cash toward whatever he/she wants to watch, and all from the TV, not the computer.I think the last idea is the best. I haven't been able to find anything on their site about how they distribute their cached content, so they could, in all reality be doing just that already.Read more at www.pbs.org/cringely/pu...
Tuesday, February 28, 2006
| A&S Research
A good way to get rid of global warming. Fertilize the oceans, get phytoplankton going, and they will help suck up some of the greenhouse gasses.
Read more at www.wm.edu/research/art...
Wednesday, February 15, 2006
Using teaming in ESX Server
VMWare ESX server runs virtual machines as services, so you don't have to be logged into the host that is running the virtual machines, and you can access those machines using a console remotely. It is much cleaner and more functional, and FAR less frustrating.
I was a little surprised however to find that most of the cool features of workstation were left out, for instance, the ability to have multiple snapshots, and to manage those snapshots in an organized way. Teams were also a neat concept, but not available in ESX server.
Well I think I may have a workaround. It isn't perfect, but I believe it will work fine under most circumstances. The following assumes that all host OS's involved are Windows-based. You need to have ESX server set up on one of the machines, and Workstation 5.5 on the other machine. Ready? Lets get started (am going from pure memory, so I may be making some corrections soon).
Steps:
1. Copy the "VMware-console-e.x.p-xxxx.exe" file from the [path] directory of the ESX server machine, to the Workstation machine.
2. Now run that file on the Workstation machine.
3. At this point you have now added the ESX Server console to your workstation machine. Create a virtual machine with your VMWare Workstation machine, and configure it the way you want with multiple snapshots if necessary.
4. Make note of the path to the virtual machine you just created. Now be sure the machine is shutdown in workstation, and close out of it in the workstation window. Copy the folder for that virtual machine to the "Virtual machines" folder on your ESX server machine.
5. Now you can open that virtual machine with ESX server console, and run it normally.
6. If you want to revert to do any kind of snapshot management, simply power-down the virtual machine in the ESX Server console and close it out. You can re-open this virtual machine in workstation, but since it is on the other machine, you will need to specify a UNC path to get to it, like \\youresxserver\c$, then browse from there. Once you find the folder for the virtual machine, simply open up the machine. At that point, you can manage snapshots however you see fit. Once you are finished, simply save, and close, and then reopen with your ESX console, and restart it.
There are limitations to this approach. I was never able to get teaming to work at all in VMWare ESX server. It just wouldn't recognize the format. Also the above instructions rely on the fact that virtual machines are shut down in the process of managing them from either ESX or Workstation.
-Schnibitz