Still it is amazing he amount of work put into it. Digging a bit more around, I found that the following sites all share the same ip: 64.106.155.30, and share the "Copyright Protect" JavaScript:
But of course, I guess you should have guessed already by looking at the 'Edible Computer Chips' and 'Quantum Chocolate' mentioned at the front page:-)
The big question is: is this a scam or a hoax?
Okay. So the PDF standard has the potential for launching external (or even embedded) files.
In Adobe Reader this will create a warning dialog with the following text:
"The file and its viewer application are set to be launched by this PDF file. The file may contain programs, macros, or viruses that could potentially harm your computer. Only open the file if you are sure it is safe. If this file was placed by a trusted person or program, you can click Open to view the file."
That seems perfectly clear to me. There is really no reason to change this behavior. This is not an exploit.
"The string theorists predict that there are perhaps 10^1,000 [ten raised to the power of one thousand] different types of universes that can be formed that way," Linde said.
Yep. Of course you will probably have to write new device drivers for a different phone.
And judging by this comment it seems that the not all of the source for the G1 is available.
If each tab really was a separate process, they would not be able to share memory (without using some sort of IPC). So for instance each tab would needs its own memory cache for storing objects like images - these could not be easily shared between tabs.
Also, stuff like session cookies would be much more difficult to share between tabs (if you log onto GMail in one tab, the other tabs/processes would have to know about it, otherwise you would have to log in again in the other tabs).
No, Their server does not directly fetch information from news.google.com. They use a double redirect (the last one using a MHTML prefix) which wrongly transmits your google session cookie to news.google.com (a cross domain vulnerability).
It is easy to check: first log in using your google account at news.google.com, and then try the Secunia test. If your system is vulnerable, you will find that your google username is present in the 'view retrieved content' (because it is displayed on the page, if and only if you are logged in). Try it.
This allows you to send mail with no @gmail in the 'from:' field. (You are then asked to verify that you own the account you want to send mail from, probably to avoid mail spoofing).
Then just forward your mail from the selected adress to gmail and all should be fine.
Of course people can still identify the mail server the mail was sent from (by it IP) as belonging to Google, but this is only a minor annoyance to me.
"... Going back to the wmf vulnerability itself, we see number of sites mention that shimgvw.dll is the vulnerable file. This doesn't seem correct as it's possible to exploit a system on which shimgvw.dll has been unregistered and deleted. The vulnerability seems to be in gdi32.dll... "
Web services are excellent supported in C# (much better than anything I know of in Java), and according to the Mono docs they are supported there as well. From http://www.mono-project.com/Webservices_and_GtkSha rp: using System; using System.Web; using System.Web.Services;
namespace GtkWebservice {
[WebService (Description="Our first web service")]
public class RemoteWebService : System.Web.Services.WebService
{
[WebMethod (Description="Adds two numbers")]
public int Add (int firstNumber, int secondNumber)
{
return firstNumber + secondNumber;
}
} }
These lines of code will be able to generate all the plumming neccesary to expose a web service (automagically creating WSDL files and such).
35 million reads per day equals 405 reads per second. (35000000 / (24 * 60 * 60) = 405).
Estimating that the camera does the signal processing (number plate extraction) locally, and that a license plate can be encoding in 10 bytes, and that the camera can be uniqely identified using a 2 byte id, we get a bandwidth to the datacenter of (12bytes * 405/S) ~ 5 kBps or 40 kbit/s.
Neglecting overhead and spike periods, a 56K modem is all that their datacenters need to handle the incoming traffic.
Future generations will be able to run a large number of simulations of the entire human history in short time.
Therefore there will be more simulated humans, than real humans which means we are more likely to be simulated than real.
However, when we do simulations we make gross simplifications of the problem. Otherwise the simulation time and memory requirements would be ridicoulous: in order to store the state of the universe you would need a memory unit the size of the universe itself. Even though compression is a possibility, we dont get around the time requirements issues:
The author mentions the notion of 'Virtual Machines' and the possibility that they could be stacked: a civilization could create a VM simulating a new civilization, which in turn could create another VM and so on.
But the authors main premise is that a civilization can run these simulation at very short time (less than seconds). In these simulations the simulated civilization could run another simulation at ridicolous short time and so on.
Converting to the VM metaphor, the VM would need to execute code a lot faster than its host. This is of course non-sense.
The Author himself notes that a complete simulation would be impossible - we cannot hope to include all quantum mechanical degress of freedom, for one. Several AI researchers (i.e. Roger Penrose) however claims that consciousness is linked to Quantum mechanics.
No matter what, the 'simulations in the simulations' would need to become increasingly coarse, until eventually reaching a stage where obviously no consciousness is present (like when we do our own 'Game of Life' simulations). So the process cannot go on "for arbitrarily many iterations" as the author claims.
Other than that, I do not agree that simulating consciousness creates something self-aware. This is an old debate (Strong versus Weak AI), but I do not think it is settled yet.
(Please forgive my syntax and spelling. English is not my primary language)
Slashdot Mirror
Still it is amazing he amount of work put into it. Digging a bit more around, I found that the following sites all share the same ip: 64.106.155.30, and share the "Copyright Protect" JavaScript:
:-)
http://www.anndeweesallen.com/ http://www.nanoscienceinstitute.com/ http://www.glycemic.com/ http://www.glycemicindextesting.com/ http://www.glycemicresearchlaboratories.com/ http://www.nutrilabusa.com/ http://www.genefoundation.com/ http://www.ediblecomputerchips.com/ http://www.trutinadulcem.com/ http://www.skinnyscience.com/ http://www.skinnyscienceedu.com/ http://www.encoderesearch.com/ http://brsei.org/ http://www.humansportsperformance.com/ http://www.arginineresearch.com/ http://www.humanmaximumperformance.com/ http://www.argm2.com/ http://argandobesity.com/ http://arginineanswers.com/ http://glycemiclifestyle.com/ http://lowglycemicchocolate.com/ http://www.skinnyicecream.com/ http://www.skinnysciencecoffee.com/ http://skinnysciencelibrary.com/ http://sweetinfusedfruits.com/ http://encodefoundation.com/
But of course, I guess you should have guessed already by looking at the 'Edible Computer Chips' and 'Quantum Chocolate' mentioned at the front page
The big question is: is this a scam or a hoax?
I did a quick search for this Nutrilab Corp.
Take a look at the following sites, that all link to each other, and revolves about dr. Ann:
http://www.anndeweesallen.com/ [ip:64.106.155.30]
http://www.nanoscienceinstitute.com/ [ip:64.106.155.30]
http://www.glycemic.com/ [ip:64.106.155.30]
http://www.nutrilabusa.com/ [ip:64.106.155.30]
http://www.genefoundation.com/ [ip:64.106.155.30]
http://www.ediblecomputerchips.com/ [ip:64.106.155.30]
http://www.trutinadulcem.com/ [ip:64.106.155.30]
Notice anything peculiar about the IP-number?
Pi - Verification
64 hours (Primary - Bellard's BBP)
66 hours (Secondary - Plouffe's BBP)
The actual code is here: http://github.com/cantrell
The platform specific files are configuration files and icons and such.
Okay. So the PDF standard has the potential for launching external (or even embedded) files. In Adobe Reader this will create a warning dialog with the following text: "The file and its viewer application are set to be launched by this PDF file. The file may contain programs, macros, or viruses that could potentially harm your computer. Only open the file if you are sure it is safe. If this file was placed by a trusted person or program, you can click Open to view the file." That seems perfectly clear to me. There is really no reason to change this behavior. This is not an exploit.
AppLocker restricts which applications are allowed to run, not which are allowed to be installed. See e.g. this review.
Firefox > Internet Explorer
"The string theorists predict that there are perhaps 10^1,000 [ten raised to the power of one thousand] different types of universes that can be formed that way," Linde said.
Yep. Of course you will probably have to write new device drivers for a different phone. And judging by this comment it seems that the not all of the source for the G1 is available.
If each tab really was a separate process, they would not be able to share memory (without using some sort of IPC). So for instance each tab would needs its own memory cache for storing objects like images - these could not be easily shared between tabs. Also, stuff like session cookies would be much more difficult to share between tabs (if you log onto GMail in one tab, the other tabs/processes would have to know about it, otherwise you would have to log in again in the other tabs).
No, Their server does not directly fetch information from news.google.com. They use a double redirect (the last one using a MHTML prefix) which wrongly transmits your google session cookie to news.google.com (a cross domain vulnerability). It is easy to check: first log in using your google account at news.google.com, and then try the Secunia test. If your system is vulnerable, you will find that your google username is present in the 'view retrieved content' (because it is displayed on the page, if and only if you are logged in). Try it.
Under GMail try:
Settings/Account/Send Mail As...
This allows you to send mail with no @gmail in the 'from:' field. (You are then asked to verify that you own the account you want to send mail from, probably to avoid mail spoofing).
Then just forward your mail from the selected adress to gmail and all should be fine.
Of course people can still identify the mail server the mail was sent from (by it IP) as belonging to Google, but this is only a minor annoyance to me.
From http://www.viruslist.com/en/weblog?discuss=1768925 30&return=1:
"... Going back to the wmf vulnerability itself, we see number of sites mention that shimgvw.dll is the vulnerable file. This doesn't seem correct as it's possible to exploit a system on which shimgvw.dll has been unregistered and deleted. The vulnerability seems to be in gdi32.dll... "
Web services are excellent supported in C# (much better than anything I know of in Java), and according to the Mono docs they are supported there as well. From http://www.mono-project.com/Webservices_and_GtkSha rp:
using System;
using System.Web;
using System.Web.Services;
namespace GtkWebservice
{
[WebService (Description="Our first web service")]
public class RemoteWebService : System.Web.Services.WebService
{
[WebMethod (Description="Adds two numbers")]
public int Add (int firstNumber, int secondNumber)
{
return firstNumber + secondNumber;
}
}
}
These lines of code will be able to generate all the plumming neccesary to expose a web service (automagically creating WSDL files and such).
Consuming services is also extremely easy. See http://www.mono-project.com/Web_Services for an example of creating a proxy class from a WSDL file.
35 million reads per day equals 405 reads per second. (35000000 / (24 * 60 * 60) = 405).
Estimating that the camera does the signal processing (number plate extraction) locally, and that a license plate can be encoding in 10 bytes, and that the camera can be uniqely identified using a 2 byte id, we get a bandwidth to the datacenter of (12bytes * 405/S) ~ 5 kBps or 40 kbit/s.
Neglecting overhead and spike periods, a 56K modem is all that their datacenters need to handle the incoming traffic.
Future generations will be able to run a large number of simulations of the entire human history in short time.
Therefore there will be more simulated humans, than real humans which means we are more likely to be simulated than real.
However, when we do simulations we make gross simplifications of the problem. Otherwise the simulation time and memory requirements would be ridicoulous: in order to store the state of the universe you would need a memory unit the size of the universe itself. Even though compression is a possibility, we dont get around the time requirements issues:
The author mentions the notion of 'Virtual Machines' and the possibility that they could be stacked: a civilization could create a VM simulating a new civilization, which in turn could create another VM and so on.
But the authors main premise is that a civilization can run these simulation at very short time (less than seconds). In these simulations the simulated civilization could run another simulation at ridicolous short time and so on.
Converting to the VM metaphor, the VM would need to execute code a lot faster than its host. This is of course non-sense.
The Author himself notes that a complete simulation would be impossible - we cannot hope to include all quantum mechanical degress of freedom, for one. Several AI researchers (i.e. Roger Penrose) however claims that consciousness is linked to Quantum mechanics.
No matter what, the 'simulations in the simulations' would need to become increasingly coarse, until eventually reaching a stage where obviously no consciousness is present (like when we do our own 'Game of Life' simulations). So the process cannot go on "for arbitrarily many iterations" as the author claims.
Other than that, I do not agree that simulating consciousness creates something self-aware. This is an old debate (Strong versus Weak AI), but I do not think it is settled yet.
(Please forgive my syntax and spelling. English is not my primary language)