well, I follow a forum providing help for new linux user. The sound problem is classical. most of the time it comes from pulseaudio's configuration that is shit in ubuntu. Uninstall/disable it and it works.
Or it came from two sound cards in the machine and the configuration tool picked the wrong one.
It is scandalous that they try to prevent you from using getting debian on you iphone by removing bittorrent. It is not even usefull since you can get it by http...
Do you know that this idea is clearly not new. Intel TBB provide the same programming model (and much more) for smps and Datacutter ( http://www.cs.umd.edu/projects/hpsl/ResearchAreas/DataCutter.htm ) provide it on grids. That are just the ones I know for working in parallel programming and not even on pratical side of it.
Disclaimer : I am not really into cryptography. My understanding of mathematics behind is mainly superficial
You are right for the Diffie-Hellman protocol. But other protocol such as RSA (used in SSH) rely on factorization. The public knowledge is a integer N=pq where p and q are big prime numbers so that it is difficult to find p and q from the only knowledge of N.
Currently the factorization is a difficult problem (however we have no proof of its computational difficulty). Any advance in the prime number knowledge could lead to better factorization algorithm. Most of prime number generation techniques are probabilistic, they do not guarantee that the p and q you generated are primes. They ensure this property with high probability.
A better understanding of the "pattern" of prime number could lead to better probabilistic factorization algorithm as well as better random number generation algorithm.
Is there a mathematician around that can enlight us ?
Well, there is a bunch of flaw in apache/php that are regularly patched which are mainly configuration problems. I acknowledge there seem to be less flaws in apache than in IIS.
I agree with you in general. But I believe
"I'm an ubuntu user, I know the software I run is secure because I install from the repository." is wrong.
It came from the repository so you came suppose the goal of the software is not to deploy malware. However, they are still not bug-free.
I know it was not your point, but linux users tend to believe that repository software are secure and they are not.
Wait, of all the modern OSes out there, tell me again, which have a virus problem?
I guess all of them is the answer. Do not get me wrong, I am a linux faithful user and I am sure that when linux will become mainstream it will be heavily infected by viruses.
pwn2own showed us how easily an attacker can gain control over a web browser (firefox and safari on both mac os x and linux). Once you have that, the door to viruses is opened. I do not even talk about "run this cool software that add more smileys"
Well, I was thinking of a Unix like operating system that allows a user to do anything on HIS account but that never get to an other user account. The user can install anything, it does not have to run under an other privilege than the user privilege.
A process running in windows should not be able to get system privilege. In practice it can because there are bugs in the system or the user enters the administrator password. But no software should need Administrator privilege to run except hardware driver.
I do not see the link to TCP since it will not prevent software flaws...
I completely agree with you. My point is that every protection software will get deactivated by user if they want to see the dancing pig. Thus I believe that anti-virus software are not a solution to the malware problem.
What you need is a robust system (I am speaking about OS level security not annoying anti virus) to prevent machine infection in the first place. A software should not be able to get system/root privilege from a user account (which clearly happen in windows, and is certainly possible in linux). Once this is true, the security policy becomes: 'nothing' for regular user that do not look for dancing pig and 'crash your user account each week' for user that do not care that much. If a malware can not cross the user boundary, you can just run as many crappy software as you want under a derivative account without any care.
Well, I have already read several comment that tends to say the problems comes from windows. I am not sure windows is the (only) problem.
Don't get me wrong I'm am a faithful user of linux, but I believe the problem is mainly user related. And I do not believe linux would do much better than windows if targetted by malware writer:
1/ Users are going to install the crap that they believe is nice such as "emoticon packs" or "the new application to know who banned you from MSN".
2/ Linux users tend to believe their operating system is virus free. The last pwn2own showed that firefox is still not bug-free and those bug can imply arbitrary code execution. If firefox does not run in a sand box (or similar protection), the problem will remain. Linux noob will tend to install package from source or do stupid things they read online. This will install virus as well.
You could say that some virus would be installed but will not reach root privilege. You do not need to have root privilege to be pain in the ass. you can compromise all openoffice document using script (as in MS Excel), you can access password in firefox (even if it is protected by a master password, at some point the user will use them). Well, all your user data are compromized.
However, gaining root privilege is usually easy: First, most people only have one account on the machine which use su, sudo or gksudo to perform administration under their X server. If you have a malware running while you use those, you're screwed. If you use a separate account to perform those task, transfering data between users can become a problem. Even if you do not transfer data, the operating system is not bug-free. Most of the time, several services run under root, there are a couple a suid binary in the system. Or there can still be kernel issues.
I am not sure that Linux is currently much more secure than windows. I believe it is more secure, but in only a marginal way. Notice that I did not said anything about Mac OS X, since I do not know it well. But I am sure the same question arise.
that is securing operating systems and educating users so that they don't install viruses ? This can also be push forward with tax dollars and would be more useful IMHO.
Yeah that why I love linux too. It's so fucking small and efficient. And you can strip all those Graphical Useless Item you don't want and keep the exact things you need.
Since you see, well informed. I am looking for an hand-held (not neccessarily cell phone) such as the nokia N800 you talked about or an ipod. Do you know some place where there is kind of a review of such items I can read ?
Consider that switching is USLESS if all hosts connect to a common point (like a file-server, or the internet)
That is because the bottleneck is the network card (if not HD bandwidth) of the server and not the network. So you'll consider setting a second server up and then your argument fall.
if the internet connection is faster than the internal network (connecting to fiber or something), switching still is useful.
well, most company do not even turn their 150Watt computer off for the week end. So I believe, the difference in power consumption is relevant for them.:)
At home you usually use wireless because you don't want to have a wire, so there is no option.
Charging people for using the internet "too much" is ridiculous. The problem is bandwidth on the pipe, not the number of bits it can handle in a month. Offer them speed tiers, not usage tiers.
To the ISP, what is costing money is not the pipe between the ISP and the customer but the pipe between the ISP to other networks. This pipe cost is billed to the ISP as a function of the usage and NOT the bandwidth. So the ISP is really paying that.
I am not trying to say they should cap or anything. Just stating that getting out of the ISP network cost them money which is not bandwidth related. That is why AOL wanted to proxy everything or why proxad was looking for p2p caches.
A French association called FDN is currently investigating solutions where you have fiber with unlimited intra-FDN communication but 30Mbps if you go through the tubes they pay.
Slashdot Mirror
well, I follow a forum providing help for new linux user. The sound problem is classical. most of the time it comes from pulseaudio's configuration that is shit in ubuntu. Uninstall/disable it and it works. Or it came from two sound cards in the machine and the configuration tool picked the wrong one.
I agree parent is a flamebait, but I believe it should be a 5, insightful, flamebait. :)
It is scandalous that they try to prevent you from using getting debian on you iphone by removing bittorrent. It is not even usefull since you can get it by http...
Do you know that this idea is clearly not new. Intel TBB provide the same programming model (and much more) for smps and Datacutter ( http://www.cs.umd.edu/projects/hpsl/ResearchAreas/DataCutter.htm ) provide it on grids. That are just the ones I know for working in parallel programming and not even on pratical side of it.
link : http://www.cs.umd.edu/projects/hpsl/ResearchAreas/DataCutter.htm .NET...
except data cutter does not rely on
Disclaimer : I am not really into cryptography. My understanding of mathematics behind is mainly superficial
You are right for the Diffie-Hellman protocol. But other protocol such as RSA (used in SSH) rely on factorization. The public knowledge is a integer N=pq where p and q are big prime numbers so that it is difficult to find p and q from the only knowledge of N.
Currently the factorization is a difficult problem (however we have no proof of its computational difficulty). Any advance in the prime number knowledge could lead to better factorization algorithm. Most of prime number generation techniques are probabilistic, they do not guarantee that the p and q you generated are primes. They ensure this property with high probability.
A better understanding of the "pattern" of prime number could lead to better probabilistic factorization algorithm as well as better random number generation algorithm.
Is there a mathematician around that can enlight us ?
well, my argument was not that the most popular is the most exploited. More that "something that is almost not used is likely not to be exploited".
The market share IIS/apache is nothing like the market share windows/linux.
I agree popularity does not explain everything.
Well, there is a bunch of flaw in apache/php that are regularly patched which are mainly configuration problems. I acknowledge there seem to be less flaws in apache than in IIS.
I agree with you in general. But I believe "I'm an ubuntu user, I know the software I run is secure because I install from the repository." is wrong. It came from the repository so you came suppose the goal of the software is not to deploy malware. However, they are still not bug-free. I know it was not your point, but linux users tend to believe that repository software are secure and they are not.
Wait, of all the modern OSes out there, tell me again, which have a virus problem?
I guess all of them is the answer. Do not get me wrong, I am a linux faithful user and I am sure that when linux will become mainstream it will be heavily infected by viruses.
pwn2own showed us how easily an attacker can gain control over a web browser (firefox and safari on both mac os x and linux). Once you have that, the door to viruses is opened. I do not even talk about "run this cool software that add more smileys"
Well, I was thinking of a Unix like operating system that allows a user to do anything on HIS account but that never get to an other user account. The user can install anything, it does not have to run under an other privilege than the user privilege.
A process running in windows should not be able to get system privilege. In practice it can because there are bugs in the system or the user enters the administrator password. But no software should need Administrator privilege to run except hardware driver.
I do not see the link to TCP since it will not prevent software flaws...
I completely agree with you. My point is that every protection software will get deactivated by user if they want to see the dancing pig. Thus I believe that anti-virus software are not a solution to the malware problem.
What you need is a robust system (I am speaking about OS level security not annoying anti virus) to prevent machine infection in the first place. A software should not be able to get system/root privilege from a user account (which clearly happen in windows, and is certainly possible in linux). Once this is true, the security policy becomes: 'nothing' for regular user that do not look for dancing pig and 'crash your user account each week' for user that do not care that much. If a malware can not cross the user boundary, you can just run as many crappy software as you want under a derivative account without any care.
Well, I have already read several comment that tends to say the problems comes from windows. I am not sure windows is the (only) problem. Don't get me wrong I'm am a faithful user of linux, but I believe the problem is mainly user related. And I do not believe linux would do much better than windows if targetted by malware writer:
1/ Users are going to install the crap that they believe is nice such as "emoticon packs" or "the new application to know who banned you from MSN".
2/ Linux users tend to believe their operating system is virus free. The last pwn2own showed that firefox is still not bug-free and those bug can imply arbitrary code execution. If firefox does not run in a sand box (or similar protection), the problem will remain. Linux noob will tend to install package from source or do stupid things they read online. This will install virus as well.
You could say that some virus would be installed but will not reach root privilege. You do not need to have root privilege to be pain in the ass. you can compromise all openoffice document using script (as in MS Excel), you can access password in firefox (even if it is protected by a master password, at some point the user will use them). Well, all your user data are compromized.
However, gaining root privilege is usually easy: First, most people only have one account on the machine which use su, sudo or gksudo to perform administration under their X server. If you have a malware running while you use those, you're screwed. If you use a separate account to perform those task, transfering data between users can become a problem. Even if you do not transfer data, the operating system is not bug-free. Most of the time, several services run under root, there are a couple a suid binary in the system. Or there can still be kernel issues.
I am not sure that Linux is currently much more secure than windows. I believe it is more secure, but in only a marginal way. Notice that I did not said anything about Mac OS X, since I do not know it well. But I am sure the same question arise.
that is securing operating systems and educating users so that they don't install viruses ? This can also be push forward with tax dollars and would be more useful IMHO.
Yeah that why I love linux too. It's so fucking small and efficient. And you can strip all those Graphical Useless Item you don't want and keep the exact things you need.
Since you see, well informed. I am looking for an hand-held (not neccessarily cell phone) such as the nokia N800 you talked about or an ipod. Do you know some place where there is kind of a review of such items I can read ?
oh shit. This is one of the most detailled post I read on /. thanks man!
It would be a good solution if you did not have an internet connection back there.
54Mbps should be enough for anybody.
but you never reaches it in real condition. I am not even sure I ever got more than 10 Mbps on wifi.
that mean you do not need electricity. You should consider not using it anymore.
Consider that switching is USLESS if all hosts connect to a common point (like a file-server, or the internet)
That is because the bottleneck is the network card (if not HD bandwidth) of the server and not the network. So you'll consider setting a second server up and then your argument fall.
if the internet connection is faster than the internal network (connecting to fiber or something), switching still is useful.
but, nice argument :)
well, most company do not even turn their 150Watt computer off for the week end. So I believe, the difference in power consumption is relevant for them. :)
At home you usually use wireless because you don't want to have a wire, so there is no option.
No guy you are wrong. Wireless is the futur, the Ethernet 2.0 CAN'T YOU SEE?
wish i had mod point. Exactely what I thought.
My source of information is a conference given by the president of FDN but I am afraid it is in french.... http://www.fdn.fr/Free-as-in-speech-Internet-or.html I'll try to find a english reference.
Charging people for using the internet "too much" is ridiculous. The problem is bandwidth on the pipe, not the number of bits it can handle in a month. Offer them speed tiers, not usage tiers.
To the ISP, what is costing money is not the pipe between the ISP and the customer but the pipe between the ISP to other networks. This pipe cost is billed to the ISP as a function of the usage and NOT the bandwidth. So the ISP is really paying that.
I am not trying to say they should cap or anything. Just stating that getting out of the ISP network cost them money which is not bandwidth related. That is why AOL wanted to proxy everything or why proxad was looking for p2p caches.
A French association called FDN is currently investigating solutions where you have fiber with unlimited intra-FDN communication but 30Mbps if you go through the tubes they pay.