|
I am just reading thru some dot-net samples on cryptography. The samples are very good and to-the-point, man, great step forward from the old-days. There're still a few things...
QUESTION 1: A key is a byte array, okay, I have no problem with this. But is the following Key = array of sixteen HEX number? Meaning, 16x16=256bit key? Is HEX number denoted by "0x"? But shouldnt "0x10" be written as "0xA" instead?
RijndaelManaged RMCrypto = new RijndaelManaged();
byte[] Key = {0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16};
byte[] IV = {0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16};
//Create a CryptoStream, pass it the NetworkStream, and encrypt
//it with the Rijndael class.
CryptoStream CryptStream = new CryptoStream(NetStream,
RMCrypto.CreateEncryptor(Key, IV),
CryptoStreamMode.Write);
QUETSION 2: For RSA algo, public key is not an array of HEX as in case of RijndaelManaged algo. But how do I know how I should construct the public key? I understand you can check "LegalKeySizes". But say Legal key size is:
min 16 (bits)
max 32 (bits)
skip 8 (bits)
Then what? An int32 is 32 bits wide. Does it mean I can have ONE int32 in the byte array?
int32 someint = 12;
byte[] MyKey = {someint};
If so, the following key looks pretty long to me:
byte[] PublicKey = {214,46,220,83,160,73,40,39,201,155,19,202,3,11,191,178,56,
74,90,36,248,103,18,144,170,163,145,87,54,61,34,220,222,
207,137,149,173,14,92,120,206,222,158,28,40,24,30,16,175,
108,128,35,230,118,40,121,113,125,216,130,11,24,90,48,194,
240,105,44,76,34,57,249,228,125,80,38,9,136,29,117,207,139,
168,181,85,137,126,10,126,242,120,247,121,8,100,12,201,171,
38,226,193,180,190,117,177,87,143,242,213,11,44,180,113,93,
106,99,179,68,175,211,164,116,64,148,226,254,172,147};
byte[] Exponent = {1,0,1};
... some code ...
//Create a new instance of RSACryptoServiceProvider.
RSACryptoServiceProvider RSA = new RSACryptoServiceProvider();
//Create a new instance of RSAParameters.
RSAParameters RSAKeyInfo = new RSAParameters();
//Set RSAKeyInfo to the public key values.
RSAKeyInfo.Modulus = PublicKey;
RSAKeyInfo.Exponent = Exponent;
norm
|
|
|
|
|
As for question one: 0x10 in Hex == 16 in Dec, 0xA in Hex == 10.
Question 2: I wrote a little app to output all the keysizes returned from RSACryptoServiceProvider.LegalKeySizes and their max, min, and skip values. It looks like this under .NET 1.1:
Max: 16384 Min: 384 Skip: 8
And those are in bits. Divide by 8 to get bytes and you have:
Max: 2048 Min: 48 Skip: 1
So it looks like you have to have a byte[] array whose length is between 47 and 2047 (because arrays are zero based). And those values can be anywhere from 0 to 255 (the possible values for a byte).
Hawaian shirts and shorts work too in Summer.
People assume you're either a complete nut (in which case not a worthy target) or so damn good you don't need to worry about camouflage...
-Anna-Jayne Metcalfe on Paintballing
|
|
|
|
|
humm... thanks.
QUESTION 1:
STEP 1: say once you have determined the legal size:
max 128
min 16
skip 16
STEP 2: Your key legal size is therefore: {16, 32, 48, 56... 128}
STEP 3: pick one size, say 56 bits
STEP 4: 56 bits = 4 x16bits = 4 HEX number = {HEX1, HEX2, HEX3, HEX4} (Is this correct?)
QUESTION 2:
Also, instead of:
byte[] MyKey = {0x10, 0x12, 0x0A, 0x3B}
Could I have generated it programmatically? Or can I load a key from a key file generated by sn.exe?
QUESTION 3:
Do I have to rely on CryptoAPI to gain access to a key in cert store - and perhaps all other tasks that has to do with cert store?
Thanks!
norm
|
|
|
|
|
RSA.modulus=MyPublicKey;
What about my "Private" key?
Thanks.
norm
|
|
|
|
|
bear with me, but one more question =)
<cryptoclass mysha1hash="MySHA1HashClass, MyAssembly
Culture='en', PublicKeyToken=a5d015c7d5a0b012,
Version=1.0.0.0">
Can be translated to:
byte [] mykey = {0x0A, 0x05, 0x0D...0x01, 0x02}
Is this correct?
norm
|
|
|
|
|
norm wrote:
bear with me, but one more question =)
<cryptoClass MySHA1Hash="MySHA1HashClass, MyAssembly
Culture='en', PublicKeyToken=a5d015c7d5a0b012,
Version=1.0.0.0"/>
Can be translated to:
byte [] mykey = {0x0A, 0x05, 0x0D...0x01, 0x02}
Is this correct?
Ummmm....that's a really good question. The answer is that I have no clue. I'm guessing that it's actually split up like:
{0xa5, 0xd0, 0x15...} etc, because otherwise why would that zero be there? It looks like it's just a placeholder so that the 8 byte (look, it's 16 digits) key can be translated easily.
Flight to Redmond - £200
Bulldozer Rental - £100
Destroying the MS campus single handedly for not doing an Academic upgrade, PRICELESS!
-Jonny Newman
|
|
|
|
|
norm wrote:
What about my "Private" key?
That's given to the constructor of the CspParameters object:
CspParameters myCspParams = new CspParameters();
myCspParams.Flags = CspProviderFlags.UseMachineKeyStore; //Or CspProviderFlags.UseDefaultKeyStore
myCspParams.KeyNumber = myKeyNumber;
RSACryptoServiceProvider rsa = new RSACryptoServiceProvider(myCspParams);
I passionately hate the idea of being with it, I think an artist has always to be out of step with his time.
-Orson Welles
|
|
|
|
|
norm wrote:
STEP 1: say once you have determined the legal size:
max 128
min 16
skip 16
STEP 2: Your key legal size is therefore: {16, 32, 48, 56... 128}
No, you're mistaking the size for the possible values. The size is the length of that array. The possible values for System.Byte are still going to be 0 to 255. That just means that the legal key size has to be an array with between 16 bytes and 128 bytes in it.
norm wrote:
STEP 4: 56 bits = 4 x16bits = 4 HEX number = {HEX1, HEX2, HEX3, HEX4} (Is this correct?)
Again, no, you're mistaking the possible values in the array for the length of the array.
norm wrote:
QUESTION 2:
Also, instead of:
byte[] MyKey = {0x10, 0x12, 0x0A, 0x3B}
Could I have generated it programmatically? Or can I load a key from a key file generated by sn.exe?
Yes! In fact, that's what I always do. I usually do something like this:
string myPassword<br />
byte[] passBytes = System.Text.Encoding.Default.GetBytes(myPassword);
And then I take passBytes, trim it down to the legal size (actually, it always fits within the legal size for RSA), and then give that to the RSA CSP. As for pulling from an sn.exe generated file, sure. I don't see why not, especially since that's generated (I believe) for RSA.
norm wrote:
QUESTION 3:
Do I have to rely on CryptoAPI to gain access to a key in cert store - and perhaps all other tasks that has to do with cert store?
No. I don't believe so. At least that's not what they say http://www.gotdotnet.com/team/clr/cryptofaq.htm[^]:
10. Does the crypto library have a key management system different from CryptoAPI?
No, we internally use CryptoAPI key storage and you can reference keys stored in CryptoAPI key containers by using the CspParameters object. For a sample of how to store and retrieve your own asymmetric key pair in CryptoAPI key storage see the sample on www.gotdotnet.com/team/clr/about_security.aspx.
However, I couldn't find that sample. You may have to go hunting in the MSDN docs. Although, the most valuable thing I saw on that "Sample" page was this:
The Source Code for the .NET Framework Cryptography Classes[^] 
I passionately hate the idea of being with it, I think an artist has always to be out of step with his time.
-Orson Welles
|
|
|
|
|
Thanks for the patience!
David Stone wrote:
STEP 1: say once you have determined the legal size:
max 128
min 16
skip 16
STEP 2: Your key legal size is therefore: {16, 32, 48, 56... 128}
>> No, I think I agree with you. {16, 32, 48, 56... 128} are just "possible" or "legal" value for key. And sure, 1 byte=8 bit=256 possibilities.
One HEX digit = 4 bits --> So, if key size=56bit, we can define key with 14 HEX digit. we can therefore write:
byte[] MyKey = {0x10, 0x12, 0x0A, 0x3B, 0x2D, 0x81, 0x7D}
Is this correct?
David Stone wrote:
string myPassword
byte[] passBytes = System.Text.Encoding.Default.GetBytes(myPassword);
That's a good trick.
norm
|
|
|
|
|
norm wrote:
One HEX digit = 4 bits --> So, if key size=56bit, we can define key with 14 HEX digit. we can therefore write:
byte[] MyKey = {0x10, 0x12, 0x0A, 0x3B, 0x2D, 0x81, 0x7D}
Is this correct?
Sounds logical. And I think that would work. 
Hawaian shirts and shorts work too in Summer.
People assume you're either a complete nut (in which case not a worthy target) or so damn good you don't need to worry about camouflage...
-Anna-Jayne Metcalfe on Paintballing
|
|
|
|
|
I need to find if the word i read from the document is a heading 1 , 2, etc.
I tried to use the method "Style"; for some reason c# doesn't support it.
Any suggestions?
Thanks
keramudugoda
|
|
|
|
|
I try to put some files like attachment from command line like this,
mailto:sarev@sarev.com?subject=my report&body=see attachment&attachment="C:\a.bmp"
but evry time I cann't put it.
|
|
|
|
|
My application checks the latest version of an application from the targeted website and if found then it downloads and installs it.But before downloading or installing i need to check the administrator privileges for these two operations.If not found then it will force the operating system to popup he adminstartor login window and after login the application can be downloaded or installed.after finishing these operations it restore the rights.
I am not able to find out the exact solution & i don't know whether it is possible by Impersonation or Policy files.So please help me.Its very urgent.
Thanks in advance.
Manoj
|
|
|
|
|
Manoj,
Have a look at System.Security.Permissions namespace. You have fine grained security classes for each set of operations.
Also if you want to check for administrative privileges you need to do something like this,
PrincipalPermission myPermission = new PrincipalPermission("Somebody", "BUILTIN\\Administrators");
then you can call the demand method to check if he belongs to the admin group. you can get the current user's name from System.Threading.Thread.CurrentPrincipal and assign it to WindowsPrincipal object and use WindowsIdentity to extract the username.
As for as popping up the dialog and restoring the rights, i have no idea.
hope this helps you getting started.
- Kannan
|
|
|
|
|
You can Impersonate current user with admin. For how to do it chech WindowsIdentity.Impersonate() in MSDN
Mazy
No sig. available now.
|
|
|
|
|
what would be the best way to store user settings, i.e. window locations, options, etc. would i use .ini files, the registry, or something else?
thanks,
Rob
--
There are 10 kinds of people. Those who understand binary and those who don't.
|
|
|
|
|
Have you considered using Application configuration files(yourapp.exe.config) or using isolated storage. I think isolated storage support per user settings.
Regards,
Kannan
|
|
|
|
|
from what i understand the app.exe.config file is readonly. if this is true then i can't use it.
Rob
--
There are 10 kinds of people. Those who understand binary and those who don't.
|
|
|
|
|
|
This may sound like self-advertising,
but I published something similar
that may (or may not) fit our friend's needs:
http://www.codeproject.com/csharp/CFGLitepack.asp
So, now.. if he wants to use XML, he has Nick's class;
else, he can use mine
Ain't CP great?
F.O.R.
|
|
|
|
|
Simply, I'd like to create a get property that would return a list of objects for a class, but I want that the list only allows access to items : adding, replacing or removing items must be impossible.
To do that I first wrote the following code :
class MyClass
{
private ArrayList _AL;
ICollection MyList
{
get { return _AL; }
}
}
Indeed, the ICollection doesn't have Add, Replace or Remove methods.
But I realized later that the ICollection interface returned by MyList could be cast into an IList interface on which it is possible to perform such operations, and then the list can be modified : the data can be acted an unexcepted way, and there is no integrity !
Does anyone know the answer to this architecture problem ?
Regards,
- Éric -
|
|
|
|
|
Did you try making the arraylist
readonly 
Regards,
Kannan
|
|
|
|
|
Unfortunalely that's not the solution.
Indeed readonly attribute will only forbid to replace the ArrayList object AL by another instance, but one can still do anything with the list. For instance, the following code shows that an item can be added normally : it compiles and no exception is thrown.
public class MyClass
{
public readonly ArrayList AL = new ArrayList();
}
public class MainClass
{
static void Main()
{
MyClass MC = new MyClass();
try
{
MC.AL.Add("toto");
}
catch(Exception e) { MessageBox.Show(e.Message); }
foreach ( object O in MC.AL )
MessageBox.Show(O.ToString()+" has been added !");
}
}
What a pity... 
- Éric -
|
|
|
|
|
|
Perfect ! That seems to be a good solution.
This static method returns a readonly copy of the ArrayList. Moreover if the original list is modified (directly or through a binding process), then the readonly copy is updated (synchronized). Very well !
Thanks very much.
- Éric -
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
