Slashdot Mirror

I developed a technique for making snapshows, or animated snapshots. The idea is to take a series of digital photographs from slightly different angles and link them into an animation (sample, sample). The effect is similar to "bullet-time" -- a picture with a little bit of motion and 3-D depth.

Snapshows require a format that can store a full-color animation of just a few frames long. GIF is out since it's limited to 8-bit color, and MPEG is too complicated and lossy. MNG worked just about right. The biggest problems were that the available editors were a bit weak and buggy and the only browser with MNG support was Mozilla. I hoped that MNG would catch on in the mainstream. Sad to hear that it's falling out of even Mozilla.

AlpineR

"Thermoptic camoflage? How is that possible?"

________________________________
The Spiders are coming

The CoVirt Project
The CoVirt project is investigating how to use virtual machines to provide security in an operating-system-independent manner. Virtual-machine security services can work even if an attacker gains complete control over the guest operating system.

One hard part of designing virtual-machine security services is the semantic gap between the virtual machine and those services. Services in the virtual machine operate below the abstractions provided by the guest operating system and applications. This can make it difficult to provide services. For example, it is difficult to provide a service that checks file system integrity without knowledge of on-disk structures.

Another potential challenge of using virtual machines is that running all applications above the virtual machine hurts performance due to virtualization overhead. Commercial virtual machine monitors such as VMware achieve excellent performance by executing (mostly) directly on the bare hardware. However, we would like to use a virtual-machine monitor that runs as a user-mode application on top of a host operating system (so-called Type II VMM), and these tend to be an order of magnitude slower than a standalone system. We modified a host OS (Linux) to enable it to better support a virtual-machine monitor. The resulting virtual-machine monitor and modified guest OS (based on UMLinux) runs even kernel-intensive applications at about 14-35% overhead. See our USENIX paper for details.

We have designed and implemented a replay service for virtual machines called ReVirt. ReVirt logs enough information to replay a long-term execution of a virtual machine instruction-by-instruction. This enables it to provide arbitrarily detailed observations about what transpired on the system, even in the presence of non-deterministic attacks and executions.

We designed and implemented a system called BackTracker that will help system administrators understand (and thereby recover from) an intrusion. BackTracker automatically identifies potential sequences of steps that occurred in an intrusion. Starting with a single detection point (e.g. a suspicious file), BackTracker identifies files and processes that could have affected that detection point and displays chains of events in a dependency graph.

Here is an example of BackTracker's output for an attack against a machine that we set up as a honeypot. It shows an attacker gaining access through httpd, downloading a tar archive using wget, then installing a set of files using tar and gzip. The attacker then ran the program openssl-too, which read the configuration files that were unpacked. We detected the intrusion when the openssl-too process began scanning other machines on our network for vulnerable ports.

Project members
Papers
Presentations
Project Sponsors
Source Code

The CoVirt Project
The CoVirt project is investigating how to use virtual machines to provide security in an operating-system-independent manner. Virtual-machine security services can work even if an attacker gains complete control over the guest operating system.

One hard part of designing virtual-machine security services is the semantic gap between the virtual machine and those services. Services in the virtual machine operate below the abstractions provided by the guest operating system and applications. This can make it difficult to provide services. For example, it is difficult to provide a service that checks file system integrity without knowledge of on-disk structures.

Another potential challenge of using virtual machines is that running all applications above the virtual machine hurts performance due to virtualization overhead. Commercial virtual machine monitors such as VMware achieve excellent performance by executing (mostly) directly on the bare hardware. However, we would like to use a virtual-machine monitor that runs as a user-mode application on top of a host operating system (so-called Type II VMM), and these tend to be an order of magnitude slower than a standalone system. We modified a host OS (Linux) to enable it to better support a virtual-machine monitor. The resulting virtual-machine monitor and modified guest OS (based on UMLinux) runs even kernel-intensive applications at about 14-35% overhead. See our USENIX paper for details.

We have designed and implemented a replay service for virtual machines called ReVirt. ReVirt logs enough information to replay a long-term execution of a virtual machine instruction-by-instruction. This enables it to provide arbitrarily detailed observations about what transpired on the system, even in the presence of non-deterministic attacks and executions.

We designed and implemented a system called BackTracker that will help system administrators understand (and thereby recover from) an intrusion. BackTracker automatically identifies potential sequences of steps that occurred in an intrusion. Starting with a single detection point (e.g. a suspicious file), BackTracker identifies files and processes that could have affected that detection point and displays chains of events in a dependency graph.

Here is an example of BackTracker's output for an attack against a machine that we set up as a honeypot. It shows an attacker gaining access through httpd, downloading a tar archive using wget, then installing a set of files using tar and gzip. The attacker then ran the program openssl-too, which read the configuration files that were unpacked. We detected the intrusion when the openssl-too process began scanning other machines on our network for vulnerable ports.

Project members
Papers
Presentations
Project Sponsors
Source Code

The CoVirt Project
The CoVirt project is investigating how to use virtual machines to provide security in an operating-system-independent manner. Virtual-machine security services can work even if an attacker gains complete control over the guest operating system.

One hard part of designing virtual-machine security services is the semantic gap between the virtual machine and those services. Services in the virtual machine operate below the abstractions provided by the guest operating system and applications. This can make it difficult to provide services. For example, it is difficult to provide a service that checks file system integrity without knowledge of on-disk structures.

Another potential challenge of using virtual machines is that running all applications above the virtual machine hurts performance due to virtualization overhead. Commercial virtual machine monitors such as VMware achieve excellent performance by executing (mostly) directly on the bare hardware. However, we would like to use a virtual-machine monitor that runs as a user-mode application on top of a host operating system (so-called Type II VMM), and these tend to be an order of magnitude slower than a standalone system. We modified a host OS (Linux) to enable it to better support a virtual-machine monitor. The resulting virtual-machine monitor and modified guest OS (based on UMLinux) runs even kernel-intensive applications at about 14-35% overhead. See our USENIX paper for details.

We have designed and implemented a replay service for virtual machines called ReVirt. ReVirt logs enough information to replay a long-term execution of a virtual machine instruction-by-instruction. This enables it to provide arbitrarily detailed observations about what transpired on the system, even in the presence of non-deterministic attacks and executions.

We designed and implemented a system called BackTracker that will help system administrators understand (and thereby recover from) an intrusion. BackTracker automatically identifies potential sequences of steps that occurred in an intrusion. Starting with a single detection point (e.g. a suspicious file), BackTracker identifies files and processes that could have affected that detection point and displays chains of events in a dependency graph.

Here is an example of BackTracker's output for an attack against a machine that we set up as a honeypot. It shows an attacker gaining access through httpd, downloading a tar archive using wget, then installing a set of files using tar and gzip. The attacker then ran the program openssl-too, which read the configuration files that were unpacked. We detected the intrusion when the openssl-too process began scanning other machines on our network for vulnerable ports.

Project members
Papers
Presentations
Project Sponsors
Source Code

The CoVirt Project
The CoVirt project is investigating how to use virtual machines to provide security in an operating-system-independent manner. Virtual-machine security services can work even if an attacker gains complete control over the guest operating system.

One hard part of designing virtual-machine security services is the semantic gap between the virtual machine and those services. Services in the virtual machine operate below the abstractions provided by the guest operating system and applications. This can make it difficult to provide services. For example, it is difficult to provide a service that checks file system integrity without knowledge of on-disk structures.

Another potential challenge of using virtual machines is that running all applications above the virtual machine hurts performance due to virtualization overhead. Commercial virtual machine monitors such as VMware achieve excellent performance by executing (mostly) directly on the bare hardware. However, we would like to use a virtual-machine monitor that runs as a user-mode application on top of a host operating system (so-called Type II VMM), and these tend to be an order of magnitude slower than a standalone system. We modified a host OS (Linux) to enable it to better support a virtual-machine monitor. The resulting virtual-machine monitor and modified guest OS (based on UMLinux) runs even kernel-intensive applications at about 14-35% overhead. See our USENIX paper for details.

We have designed and implemented a replay service for virtual machines called ReVirt. ReVirt logs enough information to replay a long-term execution of a virtual machine instruction-by-instruction. This enables it to provide arbitrarily detailed observations about what transpired on the system, even in the presence of non-deterministic attacks and executions.

We designed and implemented a system called BackTracker that will help system administrators understand (and thereby recover from) an intrusion. BackTracker automatically identifies potential sequences of steps that occurred in an intrusion. Starting with a single detection point (e.g. a suspicious file), BackTracker identifies files and processes that could have affected that detection point and displays chains of events in a dependency graph.

Here is an example of BackTracker's output for an attack against a machine that we set up as a honeypot. It shows an attacker gaining access through httpd, downloading a tar archive using wget, then installing a set of files using tar and gzip. The attacker then ran the program openssl-too, which read the configuration files that were unpacked. We detected the intrusion when the openssl-too process began scanning other machines on our network for vulnerable ports.

Project members
Papers
Presentations
Project Sponsors
Source Code

Once Solaris 10 comes out, its zomes might be able to provide a better infrastructure for what the original article was talking about, but the last presentation i saw on zones was unsure about the monitoring capabilities to be built in.

The performance boost this article mentions sounds nice, but i don't understand what else they accomplished that a tcpdump on the host machine couldn't do.

links:
honeyd
user mode linux
Bait and switch honeypots
Solaris 10 zones

Poster is missing the point. Fvwm is not a minimalist WM! There are several minimalist WMs out there, and many of them are fairly nice, if that's your cup of tea. I think larswm is a pretty nice one, and the grandaddy of them all is 9wm. And there are a bunch of others, including, apparently, EvilWM. But Fvwm is not a minimalist WM! It's a full-featured WM that happens to use an amazingly small amount of memory. It does this by being highly modular, so that only the features you actually use get loaded. It's also amazingly configurable, considering how little memory it uses. (Another amazingly-powerful-considering-how-little-memory-i t-uses WM is Window Maker -- I'm always amazed at how little memory this feature-filled WM uses.)

And looking at evilwm's web page, I have to say, there is no way I'd consider switching from fvwm. Their choice of hard-coded defaults do not match what I want. If someone wrote a minimalist WM that did have all the defaults set to what I want, then I might consider switching, but these guys aren't even close. (And even then, I'd have to find third-party equivalents for the fvwm modules I use, like the buttonbar.)

No really, I'm serious. This robot's playing style is more akin to what we call "whackball." Better foosball players remain in contact with the ball a lot longer, "palming" the ball with the face of the man to do quick changes in direction, fakes, and more. There's some video available here on the right side of the page.

Of course, there have been robots interacting with dynamic environments in similar ways for a long time, such as juggling and running. It's a big jump to go to the next level, which requires chaining together sequences of difficult actions, such as palming the ball, passing, and shooting. But I think foosball is a great place to explore such dynamical interaction and action composition, and I'm jealous that they beat me to building a table.

No really, I'm serious. This robot's playing style is more akin to what we call "whackball." Better foosball players remain in contact with the ball a lot longer, "palming" the ball with the face of the man to do quick changes in direction, fakes, and more. There's some video available here on the right side of the page.

Of course, there have been robots interacting with dynamic environments in similar ways for a long time, such as juggling and running. It's a big jump to go to the next level, which requires chaining together sequences of difficult actions, such as palming the ball, passing, and shooting. But I think foosball is a great place to explore such dynamical interaction and action composition, and I'm jealous that they beat me to building a table.

I was recently involved in a class debate on whether it is necessary to increase nuclear power production threefold to meet a carbon free economy by 2100. It seems many of the topics raised in this thread deal with points we covered in our project, e.g. safety and efficiency concerns, hydrogen production, economic feasibility, etc.

As my portion of the project dealt with safety and proliferation, I can say that at least from safety standpoint, building newer nuclear plants is a better solution to accomplish these goals than sticking with fossil fuels. For example, existing coal plants cause 15,000 premature deaths annually in the U.S. alone. Now, given the probability of 400 deaths in the event of a nuclear meltdown, this would require over 25 meltdowns per year for nuclear power to be as dangerous as the coal industry. Currently the probability of a meltdown is 1 in 20,000 reactor years, or once every 30 years.

But even if you doubt these conclusions, you can rest assured that the effects of greenhouse gases would be far more severe than an incident involving localized exposure of nuclear waste (however unlikely that may be). Keep in mind the last ice age occured when the average global temperature was as little as five degrees (C) less. And currently the global temperature is rising at a rate that tops all previous historical trends.

I was recently involved in a class debate on whether it is necessary to increase nuclear power production threefold to meet a carbon free economy by 2100. It seems many of the topics raised in this thread deal with points we covered in our project, e.g. safety and efficiency concerns, hydrogen production, economic feasibility, etc.

As my portion of the project dealt with safety and proliferation, I can say that at least from safety standpoint, building newer nuclear plants is a better solution to accomplish these goals than sticking with fossil fuels. For example, existing coal plants cause 15,000 premature deaths annually in the U.S. alone. Now, given the probability of 400 deaths in the event of a nuclear meltdown, this would require over 25 meltdowns per year for nuclear power to be as dangerous as the coal industry. Currently the probability of a meltdown is 1 in 20,000 reactor years, or once every 30 years.

But even if you doubt these conclusions, you can rest assured that the effects of greenhouse gases would be far more severe than an incident involving localized exposure of nuclear waste (however unlikely that may be). Keep in mind the last ice age occured when the average global temperature was as little as five degrees (C) less. And currently the global temperature is rising at a rate that tops all previous historical trends.

The Torricelli amendment. It was passed as part of the outrage following an incident in Guatemala where the husband of a United States citizen was murdered by a CIA paid informant.

More about that here.

It wasn't so much that the he was murdered, but that the US Govt knew what had happened to him and tried to cover it up. She tried for years to get information, information that the authorities had but kept denying.

Finally in 1995 Representative Torricelli of New Jersey revealed the information publicly.

Most people recognize the need for human intelligence, but it is very difficult to justify the US working with individuals who are murdering US citizens. How would you feel if you found out that the CIA was still in contact with, and was still paying Osama bin Laden?

Some intelligence and balance needs to be there. In the case of the Guatemala incident, the CIA had apparently severed ties with the colonel responsible for the murder, which was the correct thing for them to do. But they should not have tried to cover up the murder. Heads don't need to roll, we just need to be willing to admit when we made a bad choice.

This could prove to be the most important use of this technique, as most proposed Grand Unified Theories have interactions that can turn quarks into leptons, so that a proton would be expected to eventually decay into a positron and a meson. Unfortunately, this process has never been observed (well, only somewhat unfortunately, as high proton stability is definitely a Good Thing in most ways), and experiment and theory have thus set a lower bound on the lifetime of a proton of roughly 10^33 years, about 23 orders of magnitude greater than the estimated current age of the universe.

As you can see, compared to the suggested lifetime of a proton, even Bi-209 seems unstable. The expected extreme rarity of a proton decay event, however, is somewhat balanced by the overwhelming abundance of protons in the universe.The "lifetime" for an individual proton is more like a life expectancy, an average figure- given a suitably large collection of protons, odds are good that at least one would decay in a reasonable timeframe. If you carefully watch 10^33 protons for a year, for example, and reality agrees with theory (big if), then it is likely (certainly not guaranteed though) you will see at least one decay event. Now, 10^33 may sound like a tremendous amount, but remember that each proton has a mass of only 1.67*10^-27 kilograms, so that 10^33 protons would have a mass of about 1,600 metric tons- a lot, but not outrageous.

The real problem lies in that "carefully watching" part. So many other forms of radiation are much more prevalent, and so might mask the signature of proton decay. Cosmic rays, naturally occuring radioisotopes in places you'd never think to look, solar neutrinos, that sort of thing. Ah, why yes, this is one of those experiments they do in a salt mine and uses a gigantic tank of ultrapure water (your proton source). However, as of yet, no one has found concrete evidence for proton decay from one of these experiments. Go here for a excellent site about a proton decay detector that ran in the 80s, and here for one currently in use.

Perhaps this process will detect this very rare event, lending profound support to one of the many supersymmetric models out there. Unfortunately, if it does not detect proton decay, it will be much more difficult to say just what the result means, it being difficult to prove a negative and all.

These movies are metaphorical - entirely representational - of many things in good storytelling and in life, and point to philisophical and epistemological concepts that everyone is not perceptive enough (or open-minded enough) to grasp. Two things may fix this: 1) learn to widen your perceptions - see what things are truly there, whether you agree with them or not, and 2) educate yourself - learn of what may and might exist so that you may make an informed decision as to whether you believe those things do or do not exist. (Ironically enough, isn't this the summation of the first movie?)

As a Christian, I have learned that the same story (or parable) may have different depths of interpretation, which lead to new and progressive applications in life. Most movies are not like this, but instead are straightforward WYSIWYG chatter, though the story they tell may be something that appeals to us. The difference between these and the Matrix Trilogy is in the core of the story, which is inherently (and entirely) the clash of philosophy with perception. I heard a philosophy professor once state, "This is the essence of what it means to live as a sentient being." What do you believe?

I have a problem with those who condemn these types of movies because the action did not suit their preferences or because the CGI wasn't perfect. True, they help solidify the credence of the story in our minds, but if you understand the ideas that these things represent, then hasn't the story-writer accomplished his purpose? Feel free to e-mail me at zuriel@umich.edu with personal comments.

• Free your mind.

-Zuriel 7

Umm, OK.

"Maybe NFS4 is your answer?"

• NFSv4 home page
• NFS Version 4 Open Source Reference Implementation, for Linux and OpenBSD

--Bruce Fields

"Maybe NFS4 is your answer?"

• NFSv4 home page
• NFS Version 4 Open Source Reference Implementation, for Linux and OpenBSD

--Bruce Fields

The following is hearsay, of course. I have no transcripts, but I do know some people who could probably back me up.

The year was 1984. I was a freshman at RPI, busily ruining my life by not going to class and spending all my time in night mode on MTS, the local timesharing system. One of the most popular ways for me to waste time was by using CZSC:CB, a primitive chat room type of thing. One night I was experiencing a craving for cheese, and started babbling about it on the channels. Nobody seemed to pay attention, so I became more and more vociferous, asking people if they had any cheese, did they know anyone who had cheese in their dorm, did anyone know of an all-night supermarket where I could some cheese. Finally I started just typing CHEESE CHEESE CHEESE over and over, filling the line with CHEESE.

Some people found this funny, some people found it annoying. It was dubbed a cheeseyell. It became my trademark, over the next few months, to perform a cheeseyell or six from time to time, especially when bored.

Many of us were indeed Python freaks. Someone *may* have mentioned the relationship to the Spam sketch at the time.

After the Fall 1985 semester, I flunked out (surprised?). Around 1987, my life still a shambles, I started connecting long distance to RPI's systems. By now CB had been banned from MTS, but its heir was called CONNECT. I believe it ran on the same hardware, but you connected to it differently, and it was much less wasteful of resources. I revived the cheeseyell. I am positive that at this point the term "Spamming" was applied to it, as well as to similar repetitious, zero-content remarks. What is now referred to as "scrolling" was not an essential factor, just repetition.

As for html being used in art, that's what the second show at http://art.by.arena.ne.jp/ (1995) was all about. Plus some art shows have featured websites as part of their exhibitions for a while - nothing major that i can think of, but groups like http://entity.ummu.umich.edu/. Then there were (are? can't find link) the minimalist competitions - designing in under 5k pages - and the like. If you want pictures made from html then maybe my http://www.blackant.net/code/oth/img-html-src.html will suffice.

I'm sure i'm missing plenty of other sites and competitions but it's only 7am in my TZ.

Then:
I believe Watson and Crick's solution to DNA structure was a fabulous achievement, but press should also be given to Hershey and Chase's 1952 experiment proving DNA as the genetic material. Of course, they too rested on the shoulders of giants in chemistry and biology, but their work has equal claim to initiating an era of reverse engineering hereditary mechanisms.

Now:
Biology has come a long way reverse engineering life, but still has a long way to go. Unlike systems composed of similar components interacting to create a complex and often unpredictable outcome, life is composed of a huge variety of components which can interact to create stable outcomes (homeostasis). As we identify the individual components and subsystems, a new field is emerging. This field, called systems biology, is about modeling this complexity.

Now/Next:
Perhaps most exciting, there now exists enough information to begin forward engineering life. In living systems we have the ultimate collection of both components and subsumption architectures for making complex systems. Rodney Brooks was brilliant for modeling his robots after living systems, but a living system can be the starting point for further engineering. This work has begun, but consists mostly as limited applied science with pharmaceutical, agricultural, or industrial enzyme goals. Is anyone (else) engineering life for the sake of engineering?

How can we take you seriously when you affirm that Noriega was a candidate in the 1990 elections in Nicaragua - when almost anyone who knows anything about Central America knows that he wasn't. (Hint: check out Panama instead...)

Basically, the 1990 election was a sham, as the U.S. made it know that they would withdraw all help to Nicaragua if the Sandinistas (who overthrew Somoza, a brutal dictator who had the support of the U.S.) were elected again. Nicaragua couldn't stand up to the U.S., its citizens knew that very well. That's why the Sandinistas lost in 1990.

- Serge

Americans call themselves conservatives, but they believe in abortion, and equal rights for women.

Americans call themselves conservatives, but they believe in abortion, and equal rights for women.

Americans call themselves conservatives, but they believe in abortion, and equal rights for women.

Americans call themselves conservatives, but they believe in abortion, and equal rights for women.

Here's an old paper on a way to handle rollback and undo: A Framework for Undoing Actions in Collaborative Systems
It's been tried with a group-modified version of Emacs (DistEdit).

• Bible 1.0 - old and widely accepted standard with good support network.
• Bible 2.0 (NT) - more recent, more widespread, many upgrade paths available
• Bible 1.4.k - branched a while back, contains lots of DRM support, has HUGE annual user group meetings

You mean you haven't heard? It's now official: correlation is causation!

Apparently, it does!

-*{War is Peace}*-

CHEF is another project that is in the same area (much like MIT's OpenCourseware, which has been mentioned). CHEF is a product of the University of Michigan. Michigan currently has something they call Course Tools, but CHEF is a completely new codebase and is supposed to have additional/new/expanded features. I won't bore you with a list right now, if you are interested, the links are above.

Check with University of Michigan's property disposition. They are an all-Mac school.

And they have some pretty good deals, too. I got an HP LaserJet 4M+ with Postscript module, MIO, and almost new toner cartridge for $100.

UofM Property Disposition

Could be the arachnoid mater explained under Spinal Meninges & Spinal Cord

arachnoid mater is a thin membrane which is pressed against the inner surface of the dura mater by cerebrospinal fluid pressure; (Greek, arachnoids = spider ), the space deep to this layer (subarachnoid space) has a spider web-like appearance

--

I heard that there would be a live bush show on tv yesterday, but all I could find was a bad western.

I read this article a few days ago and bookmarked most of the links I thought valueable. If anyone else is interested add some more to this thread so I can grab them :)

Exported bookmarks Fingerprint
blackhole(4) - a sysctl(8) MIB for manipulating TCP
Help Net Security OS-FngrPrint article in PDF
Honeyd - Network Rhapsody for You
http://ojnk.sourceforge.net/stuff/iplog.readme
http://www.insecure.org/nmap/nmap-fingerprinting-a rticle.txt
IP Personality - Home
Kernel Options
p0f file listing
PhoneBoys FireWall-1 FAQs: Blocking queSO packets
s0ftpr0ject 2000 Fingerprint Fucker
Security Technologies
SourceForge.net: Project Info - SING
Sys-Security.com - Because Security is not Trivial
USENIX Technical Program - Abstract - Security Symposium - 2000

honeyd is able to do this already for quite a long time. With honeyd you can basically create "virtual hosts", running on another computer, with their own IP address, their own IP personality (it comes with a large database of them), and their own services (basically, every inetd-capable program can be used as server with it). You can even create a "virtual network" of them, with configurable routes, latency and packet loss. Indistinguishable from real computers and networks.

Of course the hydrogen-centric system won't happen for quite some time for numerous reasons. But there is something that we could do now that would make the road to hydrogen later much, much easier.

What is needed is a stable, clean national electricity supply that is independent of fossil fuels. I am talking about nuclear power. New reactors, such as high temperature gas cooled pebble-bed reactors, can compete with fossil fuels while creating far less waste and no risk of meltdown. They can even be used to burn off the 38 thousand tones of nuclear waste we have (including plutonium that could be used to make bombs).

Of course, this needs to be coupled with other technologies as well to provide a complete end-to-end energy production solution. We need to use fuel reprocessing in order to extend our fuel supply and eliminate wastes. We can use neutron transmutation to decrease the life of radioactive wastes from tens of millennia to tens of years.

There is no way that we can implement even a limited hydrogen energy storage and transport system until we have a stable and abundant source of electricity. Currently, Bush wants to build 2000 more coal fired power plants over the next 20 years (check http://www.energy.gov/). Even if hydrogen goes through 100%, we'll just be burning that much more coal.

There is a proposal that outlines one possible implementation of a national electrical energy policy using nuclear energy online that is worth a read. It goes over some pros and cons. United States of America: Energy Policy for the 21st Century and Beyond (PDF - 115KB)

P.S. - The Wired article also makes reference to hydrogen as an abundant supply of energy. If we don't eliminate this notion, we will be in sore trouble. Hydrogen can only be a transport of energy.

Real men don't use braces. Or degenerate Algol-derivative languages that use braces!

Here is a better page with a link to the actual paper, and here is another one To actually get the published paper, it requires free registration though...

I would have to say that VMWare is a pretty heavyweight solution for most needs. If you've got the time to properly make use of a honeypot, maybe you've also got the resources and skills to make VMWare worthwhile. On the other hand, check out Honeyd, a small daemon that can emulate an entire Honeynet easily on one box. This may be a better solution for you, depending on your needs.

(Note, incidentally, that Sun is sponsoring an implementation of NFSv4 for Linux (here).)

Overall, given over a decade of experience with Sun NFS, to me, it is not obviously preferable to other NFS offerings. I suspect that whatever your needs are, you can probably do just as well based on one of the open source implementations.

I dare you to mix up "Michigan State University" and "University of Michigan" around my family.

IMO you need a clue bat application.

Here, use mine.

1. As far as I know, Sun tried to license NFS. Failed. For various reasons.

Sun made the NFS protocols available in several RFCs. ( RFC 3010 NFS version 4 Protocol, RFC 1094 NFS: Network File System Protocol Specification, RFC 1813 NFS Version 3 Protocol Specification. )

Anyone was free to do a clean implementation based on the RFC, or license Sun's code. Apparently this was such a failure that NFS is used by: Sun, IBM, HP, SCO, SGI, Apple, Microsoft, Hummingbird, *Linux, *BSD, ...

Do not try to pull that "give to the community crap" at least as far as NFS is concerned.

You are either uninformed or trolling. Sun is paying the U of Michigan to port NFSv4 to Linux and OpenBSD. NFSv4 porting project

2. Solaris (not SunOS) NFS support until 2.6 was crap.

Hmmm. I was part of a team running a large engineering site using Solaris 2.5.1 and HP/UX 10.20. Solaris 2.5.1 without patches did suck. But with a reasonable patch set (you did patch, right? Even once at installation?) 2.5.1 was very solid (in my opinin, much more so than HP/UX, especially under version 3 - shudder).

Many patchlevels even as late as 2.5.1 had quite a few data corruptions bugs. As a result most old non-academic installations actually used NetAppliance when they needed NFS.

Network Appliance was founded in 1992, shipped their first product in 1993 (a 7Gb appliance), and in 1995 their total revenue was $45million. Even given their rapid growth, there is no way that "most old non-academic installations" were using Network Appliance for their NFS needs in the Solaris 2.5.1 timeframe. Network appliance history

Good grief, Sun shipped 1.6 petabytes of fibre-channel storage alone in 1998. Sun ships 1.6PB

3. I had to be a design authotity on something like 100+ Netra T1s with Solaris running the most elementary services like DNS, news, mail, etc. None of them running more then one service so they were not even loaded.

Its not the number of services that run, its how heavily they are used. DNS isn't likely to be big load, but it could be as you move up the ISP food chain; news could definitely be a heavy load depending on your feed; mail - depends. I'm also curious, if your servers "were not even loaded," why did you use so many as the "design authority?"

And frankly I have not seen so many hardware failures and memory leaks in the core OS anytime before and anytime after.

Hardware - Maybe you had a bad batch, a lemon model, just plain bad luck. I've generally had good experience with Sun kit.
Core OS - You were following that best practice known as patching your systems, right?

most linux kernels in the 2.3.x and 2.5.x series were more reliable.

Linux NFS protocol support has generally been both limited and inferior to Solaris. (Little surprise - Sun invented it.) Linux also had many problems with stability and corruption prior to 2.2.17. It has greatly improved since then, but is still limited in terms of full protocol support. Since Sun is paying for the port of NFSv4 to Linux, it will no doubt continue to improve. As to the kernel in general, the Linux kernel today isn't fully the equal of Solaris. If you want to assert that it was 4 years ago, I don't think that you are making judgements based upon facts.

4. If you have created a website that needs one 100+ CPUs box instead of having the load spread across several redundant systems you should be fired on the spot.

Strawman/flamebait. Read the post. It didn't say one box with 100+ CPUs, it said "When you want to run a giant website with 100's of CPU's." In other words, a site like you claim to have designed.

IMO you need a clue bat application.

Are you done with the clue bat yet?

Funny, I just finished a fun programming assignment here at umich on a limited version of nethack--help Hugo find the stairs!

The Swarm Development Group

SFI

Complex Systems

Actually, it damn near happened, as it was brought up for debate and passed in the house. The only thing that killed it was the lucky presence of a (real) mathematician who was there for other reasons, who had the time to "educate" the senators.

Some things never change.

Also, the math the sponsor introduces is convoluted and wrong, and he came up with 3.2.

Links: Here and Here

Isn't there something in chaos theory about snychronicity between dynamical systems - it may be the case that they can get an earth simulator to run in close sync with the real thing with enough real time inputs and then be able to 'fast forward' in time enough to do some useful predictions, altho, of course, without the synchronizing inputs the simulated future and the real one will diverge to varying degrees, depending on the state of the earth modeling art...

Analog signals can't be compressed nearly as well as digital - so stations would need big new allocations of the spectrum to broadcast (fat chance). It also means good bye, interactive TV, good bye multicasts, etc - there just wouldn't be the room for them. (you can get more info about this here, if interested.)

According to this website, certain shrews produce venom. Also, the Cuban shrew-like animal Solenodons also produces venom in its mouth.

And, just to set the record straight, only male Platypii have venomous spurs. Lastly, Platypii are one of three still-living members of the mammalian subgroup known as "monotremes."

An excellent online resource for information about the animal kingdon is the University of Michgan's Animal Diversity Web.

The article states that the platypus is the only venom-producing mammal. There are actually a few others.

The European water shrew and the North American short-tailed shrew are venomous. They use their poisonous bite to kill frogs, mice, and whatever other little creatures they eat. The bite of the solenodon (Solenodon paradoxus) of Haiti is poisonous as well.