Its not really clear about how much of your information is encryped. Your passwords yes, but your browsing history? Your bookmarks?
That would be "It's".
Anyway, the source is avalable. All of the relevent code can be found in the file named browserstate.js.
Interesting function names: "SHA1", "ARC4", "RSAEP", "G_Base64".
What worries me is the mention of RSA. SHA1 could be used for many things, but is probably used for hashing the PIN. ARC4 is most likely the encyption used. But what is RSAEP doing there? That is the primative used for full fledge RSA encryption. Is Google perhaps using the PIN to generate an RSA key on the fly, which is used to encypt a random symetric key used for ARC4ing the main data? Except for generating the key on the fly, that is how OpenPGP works, although OpenPGP does not use ARC4.
The major problem seems to be that while either system works, they need fairly widespread implementation to be generally usefull, especially if it is intended to cut back on internet traffic. Method 2 does require potentially excessive levels of state to be maintained by the routers. (Imagine several thousand streams operating concurrently). The other method requires packets to be able to contain large, even potentially enormous aadress lists. It seems like implementation might be easiest if IP was not used, but a complimentary protocol, although clearly IP would be usefull in tunneling, and potentiall as the end-delivery protocol. An advantage of meathod 1 is that it could be implemented transparently, so that the destination only sees notmal TCP/IP or UDP/IP packets. For reliablity, clearly the simplest meathod is to have the routers cache the packets and resend them until they recive an acknowlegement from each interested downstream, much like what is done in TCP. Implemeting bi-directional communication could be difficult, the easiest solution would be to use normal unicast packets in any message to the host. It is surprising that few ISPs support a multicast system. It seems like that could be a very useful system, especial for Tier-1 providers. Clearly multicast, and optionally convering some multi-cast streams into unicast streams could be very useful in trafic balancing.
You mention multicast streaming. There are only two decent ways to implement multicast systems, as far as I know.
Remember that packets normally travel "up" the network to the Tier-1 level (which shall be called the "backbone" for simplicity), across the "backbone", and then "down" the network to the destination computer. There are exceptions, especially when the host and target are both served by the same Tier-2 or Tier-3, but those cases are not important enough to need to worry about. For a multicating system to work well the packets will get to the Tier-1 level, and be more or less broadcasted to all the Tier-1 routers. So the model can be simplied by pretending that the Tier-1 level works as just a single giant router, and that the Server is connected directly to this router.
So now I should outline the two major ways in which multicasting could work.
Way #1: Each multicast packet contains a complete list of adresses to which the packet should reach. When a router recives such a packet, it scans this list, decides which downstream routers need a copy of the packet. It creates a copy of the packet for each of those downstreams, and trims the address list of each of the duplicate packets such that the downstream routers do not need to see or care about recipents other than those to which it is responsible.
Way #2: Multicast packets do not contain a list of adesses, but instead contains a stream id. For each in-use stream ID, a router maintains a list of which downstream routers are interested in that stream. The list only needs to include immedeate neighbors. A router is interested in a stream if a router or computer downstream is intested. The packets would cascade downstream, and should reach all interested computers.
Both of these solutions could include methods for working around routers that have not been updated to support the multi-cast solution, and the systems could even be combined.
There are problems with both solutions, although the second solution's requiring of routers to keep strack of state could become a problem, while the first solution does not require routers to keep any more state information than they currently do.
You want a Linux designed to blow up and destroy ships?
It seems you missed the OP's point. Torpedo-Grade Linux can sink Carrier-Grade Linux.
The air-based assult weapons of Carrier-grade Linux can cripple your communications network
(WIFI-jamming, and causing your Blue Tooth to fall out), so being able to defend your network
is very important.
Except that the particular chip in question can write directly to main memory (this is based on other comments, I did not check this myself). I suspect this chip is not a normal GPU, which has fixed API for communication, and is programmable only via pixel shaders. Clearly a better design is true co-processor, optimized for graphics, but not limited to highly specific APIs. Traditional API based communication could easilly be emulated in software running on such a chip. Of course, this paticular chip might be a normal GPU, implementing an API that supports writing to main memory.
To understand the whole system, you should understand the concept of the "flex-jack". The Flex Jack is intended to be installed anywhere where a normal Power over Ethernet jack (which is basically a normal ethernet jack) would go. The flex jack is basically a very large socket. So if you use PoE you replace your normal jacks with these. Then you have a selection of inserts for the jack, which include a plastic cover for unused jacks, a simple Ethernet Jack, a quadruple ethernet jack that contains a tiny Ethernet Switch, a 802.11g insert that creates an instant wifi hotspot, or the Jack PC, which is intended as only a thin client. Employees that need more than a thin client would be using one of the ethernet choices.
Sorry about the previous post, Accidentally posted it before being finished.
I should pay state income taxes according to the state in which i am Employed, regardless as to where the work is done, because that is the state that provides the services such as unemployment benefits.
However I should pay the local income tax to the locality where I was physically present while doing the work, as they provide services such as Emercigency Medical, etc.
Proprty taxes should stay as-is.
State sales taxes should be payed only to the state in which I reside, regardless of the location of the store. Local sales taxes should belong to the locality in which I purchased the item if the sale was in-person, but should belong to my home locality if purchased out-of-person such as Electronicly, by phone, or by mail-in catalog.
I base all of those statements on the types of services that are provided, and whether I receive them.
As for the taxation without representation claims of some other posters,
at the state level I should be able to vote for state level matters in any state in which I am employed or reside, as obviously I would have a legitimate interest in those matters. Note that any state representative to Congress would not be a state level vote, because Congress is Federal.
I should pay state income taxes according to the state in which i am Employed, regardless as to where the work is done, because that is the state that provides the services such as unemployment benefits.
However I should pay the local income tax to the locality where I was physically present while doing the work, as they provide services such as Emercigency Medical, etc.
Proprty taxes should stay as-is.
State sales taxes should be payed only to the state in which I reside, regardless of the location of the store. Local sales taxes should belong to the locality in which I purchased the item if the sale was in-person, but should belong to my home locality if purchased out-of-person such as Electronicly, by phone, or by mail-in catalog.
I base all of those statements on the types of services that are provided, and whether I receive them.
As for the taxation without representation claims of some other posters,
at the state level I should be able to vote for state level matters in any state in which I am employed or reside, as obviously I would have a legitimate interest in those matters. Note that any state representative to Congress would not be a state level vote, because Congress is Federal.
Electromagnetic radiation in all forms appears to be correlated to cancer. There are studies indicatiting exposure to "Light at night" appears to increase the odds of cancer.
WINE is not a GNU project. WINE Is Not an Emulator is not a GNU's Not Unix project. (WINE Is Not an Emulator) Is Not an Emulator is not a (GNU's not Unix)'s Not Unix project....
If so, would you insist on keeping *all* the icons, even the ones which haven't been used for more than 5 years, and are never likely to be used in the future?
Of course! To do otherwise would break the archives.
Google did not do anything ethically wrong. Blocking the results ensures that chinese people can use Google. It is not teribly difficult for a chinese citizen to bypass the firewall, but guess what? It is also fairly easy for a chinese person to bypass the google censorship too!
Those who cannot figure out how to bypass the google censorship would likely have trouble bypassing the Great Firewall. Therefore the censored results are all that they have a use for.
The address given belongs to "Gille, Hrabal, Struck, Neidlein, Prop & Roos".
Gille, Hrabal, Struck, Neidlein, Prop & Roos is a German patent/trademark law firm. I would assume that if the registration is valid, the firm is acting on Linus's behalf.
Also note that they registered in their own name also. A bit strange. If they are acting on behalf of Linus, they will most likely (or hopefully, anyway) be granted the domain.
ummmmm interesting....
time ago, i found that when someone turns on the tv (specially the big ones), i heard a silly high, pitched, noise....i tought "is ok, is a high frecuency sound for the tv-speakers"....but years latter i noticed i keep hearign that noice, even if i cover my hears, or if im in another room, even another floor or even outside the house....
i begun to make tests, like using a control remote to turn on and off the tv... or telling a person in another room to randomly turn on and off the tv and i said if was on or off (with the sound switched off) and i changed the distance from 1 meter to 8 meters (after that... the effect is near zero)...and i found that : or is a sound in a certain frequency that can pass the walls and that only i can hear...or is a EM radiation and my brain is very sensitive to that frequency...
i found too that a old VCR makes the same effect...but seems only tvs and certain electrodomestics make that effect...
I actually suspect that the noise is a very high amplitude, very high fequency noise. Due to the absurd frequency you are not able to accurately determine the amplitude of the sound.
The frequency is near the edge of the human hearing threshold. Indeed most people are not sensitive to it. Adults do not often notice it which seems consistant normal gradual loss of sensitivity to high fequencies.
I suspect that this is not a particular fequency but a range of frequencies. Devices that have been known to cause this for me include Certain TVs (generally inexpensive tvs), my cellphone's battery charger (it is poorly contucted), and I belive when I was younger some heat lamps had the same effect.
I've noticed that how badly the noise affects me seems to vary, and that if I ignore it I generally tune it out. I may try a frequency analysis of the noise emmited by my cell phone charger. I expect to find an amplitude spike somwhere near the threshold of human hearing, possibly even slightly beyond the the of human hearing. I may post the results of such anaysis in a my Slashdot journal.
Weird. If I dial like this from a landline, I get an error saying to dial a 1 first. On a cell phone, the 1 is not required. I'm curious if your problm is still possible?
Sure it is still possible. The GP live in Place_A which has an echange of '555'. Nerby is Place_B who has an area code of '555'. There is also place_C, from which Place_A is local, and Place_B is long distance. Place_C is one of those places in the country where area codes are not yet mandatory. If Somebody in place_C is trying to call place B it is crutial to dial the 1, because if they fail, the call goes to place_A.
Obviously somebody was an idiot when they decided that exchange '555' could be located nearby exchange '555'.
When the hell are you fucking morons going to realize that intellectual property protects the EXPRESSION of an idea, not the idea itself?
You seem to be confused. Let's look at the different types of intellectual property, and what they officially protect:
Copyright: Expression of Ideas
Patent: Idea (officially an applied idea)
Trade Secret: Idea (officially an applied idea)
Trademark: This may protect an idea or its expression in some cases, although that is not the primary use.
Mask rights: Expression of an idea.
So it seems like some IP rights do protect ideas.
On annother topic, I don't see the need for mask rights. The claim here is that copyright is not sufficent since the design of semi-conductor masks is mostly functional. The flaw in this line of reasoning is that in an average chip design there are many thousands of slight variations that are functionally identical. Slight differences in routing are often possible with no adverse affect. Placement of subsystems can be somewhat arbitrary. Chip manufacturers don't always stive for the most efficient layout. So it seems to me that copyright is most certainly sufficient in most cases.
The search engines all seem to be borrowing from each other. The look and feel of the Yahoo search result page is clearly stolen from google. The image search system of Yahoo also looks quite a bit like Google's. Too much to be a coincidence.
On the otherhand Google Groups2 seems to be inpired by yahoo grous and the like.
And of course Google Pages seems a bit like Geocities.
Nevertheless, Google is innovative. I'm maintaining a nearly complete list of google services, and the length of the list is shocking!
Google Video uses flash for the web interface, but it is possible to download the file in other formats, and even the original format (the format as it was upload) if you look. Of course that only applies to the free videos.
If you are realy serious then Yahoo! just committed suicide. Yahoo! has been workingvery hard to look just like Google. Look at the search results page. Damn near identical. Look at the image search. Yahoo! again copied Google's look. Google's distictive styles almost certainly qualify as "Trade Dress", and Yahoo! is deliberately infringing. Yahoo! probably infringes on some Google Patents anyway. So if Yahoo! sues Google, the counter-suit will kill Yahoo!.
Huh? The authenticity stickers are required by microsoft for OEMS. They replace the certificate of authenticity. By being stuck to the computer they are far less likly to be lost.
AFAIK there is no requirement for end users to do the same.
You remembered the final dot on the demain name. (Or you have great luck in having a typo like that). May people do not realize that domain names are actually end in a final dot. This is presumable because many browsers and DNS servers do not care about the difference. See http://en.wikipedia.org/wiki/Root_nameserver for details.
My local community college's "Java" class uses J# 2.0! Talk about tring to lock somebody into Windows. The textbook is for real Java 1.5.0, but J# 2.0 is equivlent to approximately 1.3.0.
I feel bad for the fools who go to that college and think they have learned programming.
I mean it is bad enough that many "programmers" out there are just mindless drones. They are incapable of doing anything that does not closely resemble an example in their textbook. They would never teach themselves anything, or even look up a fuction in the reference guide.
Slashdot Mirror
Anyway, the source is avalable. All of the relevent code can be found in the file named browserstate.js.
Interesting function names: "SHA1", "ARC4", "RSAEP", "G_Base64".
What worries me is the mention of RSA. SHA1 could be used for many things, but is probably used for hashing the PIN. ARC4 is most likely the encyption used. But what is RSAEP doing there? That is the primative used for full fledge RSA encryption. Is Google perhaps using the PIN to generate an RSA key on the fly, which is used to encypt a random symetric key used for ARC4ing the main data? Except for generating the key on the fly, that is how OpenPGP works, although OpenPGP does not use ARC4.
The major problem seems to be that while either system works, they need fairly widespread implementation to be generally usefull, especially if it is intended to cut back on internet traffic. Method 2 does require potentially excessive levels of state to be maintained by the routers. (Imagine several thousand streams operating concurrently). The other method requires packets to be able to contain large, even potentially enormous aadress lists. It seems like implementation might be easiest if IP was not used, but a complimentary protocol, although clearly IP would be usefull in tunneling, and potentiall as the end-delivery protocol. An advantage of meathod 1 is that it could be implemented transparently, so that the destination only sees notmal TCP/IP or UDP/IP packets. For reliablity, clearly the simplest meathod is to have the routers cache the packets and resend them until they recive an acknowlegement from each interested downstream, much like what is done in TCP. Implemeting bi-directional communication could be difficult, the easiest solution would be to use normal unicast packets in any message to the host. It is surprising that few ISPs support a multicast system. It seems like that could be a very useful system, especial for Tier-1 providers. Clearly multicast, and optionally convering some multi-cast streams into unicast streams could be very useful in trafic balancing.
Remember that packets normally travel "up" the network to the Tier-1 level (which shall be called the "backbone" for simplicity), across the "backbone", and then "down" the network to the destination computer. There are exceptions, especially when the host and target are both served by the same Tier-2 or Tier-3, but those cases are not important enough to need to worry about. For a multicating system to work well the packets will get to the Tier-1 level, and be more or less broadcasted to all the Tier-1 routers. So the model can be simplied by pretending that the Tier-1 level works as just a single giant router, and that the Server is connected directly to this router.
So now I should outline the two major ways in which multicasting could work.
Way #1: Each multicast packet contains a complete list of adresses to which the packet should reach. When a router recives such a packet, it scans this list, decides which downstream routers need a copy of the packet. It creates a copy of the packet for each of those downstreams, and trims the address list of each of the duplicate packets such that the downstream routers do not need to see or care about recipents other than those to which it is responsible.
Way #2: Multicast packets do not contain a list of adesses, but instead contains a stream id. For each in-use stream ID, a router maintains a list of which downstream routers are interested in that stream. The list only needs to include immedeate neighbors. A router is interested in a stream if a router or computer downstream is intested. The packets would cascade downstream, and should reach all interested computers.
Both of these solutions could include methods for working around routers that have not been updated to support the multi-cast solution, and the systems could even be combined. There are problems with both solutions, although the second solution's requiring of routers to keep strack of state could become a problem, while the first solution does not require routers to keep any more state information than they currently do.
Except that the particular chip in question can write directly to main memory (this is based on other comments, I did not check this myself). I suspect this chip is not a normal GPU, which has fixed API for communication, and is programmable only via pixel shaders. Clearly a better design is true co-processor, optimized for graphics, but not limited to highly specific APIs. Traditional API based communication could easilly be emulated in software running on such a chip. Of course, this paticular chip might be a normal GPU, implementing an API that supports writing to main memory.
To understand the whole system, you should understand the concept of the "flex-jack". The Flex Jack is intended to be installed anywhere where a normal Power over Ethernet jack (which is basically a normal ethernet jack) would go. The flex jack is basically a very large socket. So if you use PoE you replace your normal jacks with these. Then you have a selection of inserts for the jack, which include a plastic cover for unused jacks, a simple Ethernet Jack, a quadruple ethernet jack that contains a tiny Ethernet Switch, a 802.11g insert that creates an instant wifi hotspot, or the Jack PC, which is intended as only a thin client. Employees that need more than a thin client would be using one of the ethernet choices.
I should pay state income taxes according to the state in which i am Employed, regardless as to where the work is done, because that is the state that provides the services such as unemployment benefits.
However I should pay the local income tax to the locality where I was physically present while doing the work, as they provide services such as Emercigency Medical, etc.
Proprty taxes should stay as-is.
State sales taxes should be payed only to the state in which I reside, regardless of the location of the store. Local sales taxes should belong to the locality in which I purchased the item if the sale was in-person, but should belong to my home locality if purchased out-of-person such as Electronicly, by phone, or by mail-in catalog.
I base all of those statements on the types of services that are provided, and whether I receive them.
As for the taxation without representation claims of some other posters, at the state level I should be able to vote for state level matters in any state in which I am employed or reside, as obviously I would have a legitimate interest in those matters. Note that any state representative to Congress would not be a state level vote, because Congress is Federal.
I should pay state income taxes according to the state in which i am Employed, regardless as to where the work is done, because that is the state that provides the services such as unemployment benefits. However I should pay the local income tax to the locality where I was physically present while doing the work, as they provide services such as Emercigency Medical, etc. Proprty taxes should stay as-is. State sales taxes should be payed only to the state in which I reside, regardless of the location of the store. Local sales taxes should belong to the locality in which I purchased the item if the sale was in-person, but should belong to my home locality if purchased out-of-person such as Electronicly, by phone, or by mail-in catalog. I base all of those statements on the types of services that are provided, and whether I receive them. As for the taxation without representation claims of some other posters, at the state level I should be able to vote for state level matters in any state in which I am employed or reside, as obviously I would have a legitimate interest in those matters. Note that any state representative to Congress would not be a state level vote, because Congress is Federal.
Electromagnetic radiation in all forms appears to be correlated to cancer.
There are studies indicatiting exposure to "Light at night" appears to increase the odds of cancer.
WINE is not a GNU project. ...
WINE Is Not an Emulator is not a GNU's Not Unix project.
(WINE Is Not an Emulator) Is Not an Emulator is not a (GNU's not Unix)'s Not Unix project.
Google did not do anything ethically wrong.
Blocking the results ensures that chinese people can use Google.
It is not teribly difficult for a chinese citizen to bypass the firewall, but guess what? It is also fairly easy for a chinese person to bypass the google censorship too!
Those who cannot figure out how to bypass the google censorship would likely have trouble bypassing the Great Firewall. Therefore the censored results are all that they have a use for.
The address given belongs to "Gille, Hrabal, Struck, Neidlein, Prop & Roos".
Gille, Hrabal, Struck, Neidlein, Prop & Roos is a German patent/trademark law firm. I would assume that if the registration is valid, the firm is acting on Linus's behalf.
Also note that they registered in their own name also. A bit strange.
If they are acting on behalf of Linus, they will most likely (or hopefully, anyway) be granted the domain.
I actually suspect that the noise is a very high amplitude, very high fequency noise. Due to the absurd frequency you are not able to accurately determine the amplitude of the sound. The frequency is near the edge of the human hearing threshold. Indeed most people are not sensitive to it. Adults do not often notice it which seems consistant normal gradual loss of sensitivity to high fequencies.
I suspect that this is not a particular fequency but a range of frequencies. Devices that have been known to cause this for me include Certain TVs (generally inexpensive tvs), my cellphone's battery charger (it is poorly contucted), and I belive when I was younger some heat lamps had the same effect.
I've noticed that how badly the noise affects me seems to vary, and that if I ignore it I generally tune it out. I may try a frequency analysis of the noise emmited by my cell phone charger. I expect to find an amplitude spike somwhere near the threshold of human hearing, possibly even slightly beyond the the of human hearing. I may post the results of such anaysis in a my Slashdot journal.
Obviously somebody was an idiot when they decided that exchange '555' could be located nearby exchange '555'.
Copyright: Expression of Ideas
Patent: Idea (officially an applied idea)
Trade Secret: Idea (officially an applied idea)
Trademark: This may protect an idea or its expression in some cases, although that is not the primary use.
Mask rights: Expression of an idea.
So it seems like some IP rights do protect ideas.
On annother topic, I don't see the need for mask rights. The claim here is that copyright is not sufficent since the design of semi-conductor masks is mostly functional. The flaw in this line of reasoning is that in an average chip design there are many thousands of slight variations that are functionally identical. Slight differences in routing are often possible with no adverse affect. Placement of subsystems can be somewhat arbitrary. Chip manufacturers don't always stive for the most efficient layout. So it seems to me that copyright is most certainly sufficient in most cases.
The search engines all seem to be borrowing from each other. The look and feel of the Yahoo search result page is clearly stolen from google. The image search system of Yahoo also looks quite a bit like Google's. Too much to be a coincidence.
On the otherhand Google Groups2 seems to be inpired by yahoo grous and the like.
And of course Google Pages seems a bit like Geocities.
Nevertheless, Google is innovative. I'm maintaining a nearly complete list of google services, and the length of the list is shocking!
Google Video uses flash for the web interface, but it is possible to download the file in other formats, and even the original format (the format as it was upload) if you look. Of course that only applies to the free videos.
If you are realy serious then Yahoo! just committed suicide.
Yahoo! has been workingvery hard to look just like Google. Look at the search results page. Damn near identical. Look at the image search. Yahoo! again copied Google's look. Google's distictive styles almost certainly qualify as "Trade Dress", and Yahoo! is deliberately infringing. Yahoo! probably infringes on some Google Patents anyway. So if Yahoo! sues Google, the counter-suit will kill Yahoo!.
Huh? The authenticity stickers are required by microsoft for OEMS. They replace the certificate of authenticity. By being stuck to the computer they are far less likly to be lost.
AFAIK there is no requirement for end users to do the same.
Damn typos.
s/demain/domain/
s/are//
You remembered the final dot on the demain name. (Or you have great luck in having a typo like that). May people do not realize that domain names are actually end in a final dot. This is presumable because many browsers and DNS servers do not care about the difference. See http://en.wikipedia.org/wiki/Root_nameserver for details.
You need to enter a URI not a search term.
My local community college's "Java" class uses J# 2.0!
Talk about tring to lock somebody into Windows.
The textbook is for real Java 1.5.0, but J# 2.0 is equivlent to approximately 1.3.0.
I feel bad for the fools who go to that college and think they have learned programming.
I mean it is bad enough that many "programmers" out there are just mindless drones. They are incapable of doing anything that does not closely resemble an example in their textbook. They would never teach themselves anything, or even look up a fuction in the reference guide.