This is where the government can step in: they can make it unlawful for ISPs to not disconnect someone after they are notified of illegal activities via a hijacked PC. Right now many ISPs simply don't care what their customers do. This should change. When not handling abuse reports will result in being put out of business, this will wake them up and put people on the task.
That may seem unusual but it probably happens every day, and both in the compromised and non-compromised batch of accounts. I don't see how such monitoring is going to detect any relation between a fraud case and this leak, other than statistical figures for all of the accounts.
Bank of America said it will continue to monitor the accounts on the data tapes and will contact the government cardholders if any unusual activity is observed.
Earilier in the article they said there are 2.1 million accounts and 1.2 million of those have been compromised. How will it be possible to monitor for "unusual activity" on half of your accounts? Unusual when compared to the other half? Not very realistic, I think.
You are considered about the processing required to encrypt the data? Or about the key management involved?
I think the decision not to encrypt backups is normally motivated by the hassle it would cause. When you lose the keys, the encryption is worthless. When the same keys are used every time, it is also almost worthless. So encryption causes extra work, to manage and securely store the keys. The actual encryption of course isn't a problem.
It should be possible to resize the Windows partition, install Linux and VMware, and then run VMware with direct access to the Windows partition.
Of course this limits the possibilities you have with VMware and you have to hope that XP does not see these changes (partition size, memory size etc) as suffcicient to require re-activation.
SSL is not working with shared hosting. You need a dedicated server with a separate IP address to realistically use SSL.
Why? With shared hosting, the virtual host is selected based on the Host: header of the HTTP request. But the request is sent over the SSL connection! So the sequence is: 1. establish secure connection based on certificate (which is attached to sitename) 2. send request over secure connection
But in shared hosting the situation is: 1. connect to shared host 2. decide which site to serve based on hostname sent with request
Unfortunately, those two sequences are conflicting.
Filemon and regmon are also very useful when you are trying to lockdown a system and are confronted with applications that only like to run on a system where the user has Administrative access.
Creating tempfiles in %windir%, keeping datafiles under %ProgramFiles%\Appname, modifying registry keys in HKEY_LOCAL_MACHINE, etc. These can be tracked with those tools.
This black-and-white thinking is typical for American people, and sadly, for European government people it has become more common as well.
However, before you think that a doubling of fuel prices would lead to a sudden loss of infrastructure and collapse of the entire economy please consider that here in Europe the fuel prices are 3 times what you pay in the USA and this effect has not happened.
But what we see now is that the USA is the largest waster of fossil fuel in the world, and the largest dumper of greenhouse gasses and other pollution. And before you say that you are the largest country, this is not true by far.
But then, it comes down to "what do we spend our tax money on? will we improve broadband connectivity for our citizens or will we interfere with other countries' business and pretend it is for a good cause"?
Those days are gone. It used to be that you only needed to scan executables because a virus could only activate when the program was actually running, but: - Microsoft introduced and pushed the concept that many passive datafiles can include executable content like macros - Hackers got very determined and for every case of a buffer overflow or other fault in parsing a datafile, they craft an exploit that actually does damage
So it is no longer safe to exclude data from scanning "because it cannot do any harm".
- write a program that installs a trojan - write documentation that says it handles.whatever files - make sure Google has indexed it - send.whatever files around
People will download and install your trojan all by themselves! Profit!
Of course your scanner should not check filenames (extensions are basically just part of the filename) but it should determine the type of the file by looking at its contents.
Every decent scanner does this. Surprisingly many commercial virus-scanners from "wellknown manufacturers" don't. But who said those were any good?
First, there was "Basicode". This was simply BASIC sourcecode saved as kansas-city-standard modem data (1200 bps 1200-2400Hz coherent FSK). It was developed by Klaas Robers for the Apple ][. Some other people wrote load/save software for other computers of that time. I wrote the TRS-80 model I version.
After some time it was realized that it was difficult to write software in BASIC that would run unmodified on many computers. At first this was not really an issue as every computer owner knew how to program, and how to modify programs that were not working correctly. However, the times started where computer owners were simply consumers, and this had to change.
So, "Basicode 2" was born. It used the same load/save format, but additionally there was a library of subroutines that you could call from your program to accomplish certain things, like clearing the screen, positioning the cursor at some given location, etc. The subroutines all had defined linenumbers below 1000, and your program was supposed to start at 1000.
There was no specific interpreter, each computer used its native BASIC interpreter (usually Microsoft Basic). This meant that the program was only allowed to use a given BASIC subset, also described in the "Basicode 2" standard.
I had problems with an Adaptec SCSI controller in a Dell system recently.
The trouble is that Adaptec seems to think that doing RAID-1 in the device driver is somehow a good idea and worthy to be very secretive about. So they provide binary-only drivers for their card and it is 3 kernel versions behind. Of course we need no Adaptec software RAID-1 as Linux has it in the kernel. After some searching and asking I found a patch that allowed the Adaptec controller to operate as a plain SCSI controller and from then on it has worked OK. Of course this means trouble whenever the kernel is updated, which happens every 2-3 weeks lately:-( So I decided to swap the system with one running Windows this week. I know that the other one, which has a MPT controller, works without such problems.
Adaptec is better avoided. Problems like this are not uncommon with their controllers.
Normally when you sign a contract including a phone supplied to you, you are not paying the full price of the phone upfront. You pay nothing or a small amount upfront, and then pay for your phone as part of the monthly fee (even when this is not specified).
Of course they don't want you to cancel the subscription and walk to a less expensive competitor with the phone they paid but you did not yet pay back. So they lock your phone. You can get the unlock code when some time has elapsed (and you paid back at least some amount), and/or if they trust your story that you are not going to cancel.
It is not reversible because more than one datafile (with different content) will map to the same hash value. So you don't know which of those was the original file.
Slashdot Mirror
This is where the government can step in: they can make it unlawful for ISPs to not disconnect someone after they are notified of illegal activities via a hijacked PC. Right now many ISPs simply don't care what their customers do. This should change.
When not handling abuse reports will result in being put out of business, this will wake them up and put people on the task.
Cisco use Linux in their products as well. For example, they sell a web accelerator module that is a Linux sytem with a proxy.
Try Skype. It connects as directly as possible and encrypts everything.
Why not?
Because this already can be specified in html metadata:
<meta name="revisit-after" content="7 days">
Why not?
Because this already can be specified in html metadata:
That may seem unusual but it probably happens every day, and both in the compromised and non-compromised batch of accounts.
I don't see how such monitoring is going to detect any relation between a fraud case and this leak, other than statistical figures for all of the accounts.
Bank of America said it will continue to monitor the accounts on the data tapes and will contact the government cardholders if any unusual activity is observed.
Earilier in the article they said there are 2.1 million accounts and 1.2 million of those have been compromised.
How will it be possible to monitor for "unusual activity" on half of your accounts? Unusual when compared to the other half?
Not very realistic, I think.
You are considered about the processing required to encrypt the data? Or about the key management involved?
I think the decision not to encrypt backups is normally motivated by the hassle it would cause. When you lose the keys, the encryption is worthless. When the same keys are used every time, it is also almost worthless. So encryption causes extra work, to manage and securely store the keys.
The actual encryption of course isn't a problem.
It will fit in a shirtpocket.
It should be possible to resize the Windows partition, install Linux and VMware, and then run VMware with direct access to the Windows partition.
Of course this limits the possibilities you have with VMware and you have to hope that XP does not see these changes (partition size, memory size etc) as suffcicient to require re-activation.
Maybe when you buy such a certificate they move you over to a dedicated server and this is part of the reason they charge more.
SSL is not working with shared hosting.
You need a dedicated server with a separate IP address to realistically use SSL.
Why? With shared hosting, the virtual host is selected based on the Host: header of the HTTP request. But the request is sent over the SSL connection!
So the sequence is:
1. establish secure connection based on certificate (which is attached to sitename)
2. send request over secure connection
But in shared hosting the situation is:
1. connect to shared host
2. decide which site to serve based on hostname sent with request
Unfortunately, those two sequences are conflicting.
Filemon and regmon are also very useful when you are trying to lockdown a system and are confronted with applications that only like to run on a system where the user has Administrative access.
Creating tempfiles in %windir%, keeping datafiles under %ProgramFiles%\Appname, modifying registry keys in HKEY_LOCAL_MACHINE, etc. These can be tracked with those tools.
This black-and-white thinking is typical for American people, and sadly, for European government people it has become more common as well.
However, before you think that a doubling of fuel prices would lead to a sudden loss of infrastructure and collapse of the entire economy please consider that here in Europe the fuel prices are 3 times what you pay in the USA and this effect has not happened.
But what we see now is that the USA is the largest waster of fossil fuel in the world, and the largest dumper of greenhouse gasses and other pollution. And before you say that you are the largest country, this is not true by far.
What I always wonder about is why I get all this Korean spam (100 messages a day) and no Korean understands that I will not be able to read that!
.kr addresses on it.
Let them trim down their spamlist and only leave
But the environment would live up!
Of course, some countries value the economy much higher than the environment.
But then, it comes down to "what do we spend our tax money on? will we improve broadband connectivity for our citizens or will we interfere with other countries' business and pretend it is for a good cause"?
Those days are gone.
It used to be that you only needed to scan executables because a virus could only activate when the program was actually running, but:
- Microsoft introduced and pushed the concept that many passive datafiles can include executable content like macros
- Hackers got very determined and for every case of a buffer overflow or other fault in parsing a datafile, they craft an exploit that actually does damage
So it is no longer safe to exclude data from scanning "because it cannot do any harm".
So what you could really do is:
.whatever files .whatever files around
- write a program that installs a trojan
- write documentation that says it handles
- make sure Google has indexed it
- send
People will download and install your trojan all by themselves! Profit!
I hope that served to teach you that e-mail is not a sensible mechanism to exchange executables.
Of course your scanner should not check filenames (extensions are basically just part of the filename) but it should determine the type of the file by looking at its contents.
Every decent scanner does this. Surprisingly many commercial virus-scanners from "wellknown manufacturers" don't. But who said those were any good?
Actually it worked like this:
First, there was "Basicode". This was simply BASIC sourcecode saved as kansas-city-standard modem data (1200 bps 1200-2400Hz coherent FSK). It was developed by Klaas Robers for the Apple ][. Some other people wrote load/save software for other computers of that time. I wrote the TRS-80 model I version.
After some time it was realized that it was difficult to write software in BASIC that would run unmodified on many computers. At first this was not really an issue as every computer owner knew how to program, and how to modify programs that were not working correctly.
However, the times started where computer owners were simply consumers, and this had to change.
So, "Basicode 2" was born. It used the same load/save format, but additionally there was a library of subroutines that you could call from your program to accomplish certain things, like clearing the screen, positioning the cursor at some given location, etc.
The subroutines all had defined linenumbers below 1000, and your program was supposed to start at 1000.
There was no specific interpreter, each computer used its native BASIC interpreter (usually Microsoft Basic).
This meant that the program was only allowed to use a given BASIC subset, also described in the "Basicode 2" standard.
I had problems with an Adaptec SCSI controller in a Dell system recently.
:-(
The trouble is that Adaptec seems to think that doing RAID-1 in the device driver is somehow a good idea and worthy to be very secretive about. So they provide binary-only drivers for their card and it is 3 kernel versions behind.
Of course we need no Adaptec software RAID-1 as Linux has it in the kernel. After some searching and asking I found a patch that allowed the Adaptec controller to operate as a plain SCSI controller and from then on it has worked OK. Of course this means trouble whenever the kernel is updated, which happens every 2-3 weeks lately
So I decided to swap the system with one running Windows this week. I know that the other one, which has a MPT controller, works without such problems.
Adaptec is better avoided. Problems like this are not uncommon with their controllers.
Normally when you sign a contract including a phone supplied to you, you are not paying the full price of the phone upfront. You pay nothing or a small amount upfront, and then pay for your phone as part of the monthly fee (even when this is not specified).
Of course they don't want you to cancel the subscription and walk to a less expensive competitor with the phone they paid but you did not yet pay back. So they lock your phone.
You can get the unlock code when some time has elapsed (and you paid back at least some amount), and/or if they trust your story that you are not going to cancel.
It is not reversible because more than one datafile (with different content) will map to the same hash value. So you don't know which of those was the original file.