VB Encryption Program :: Need BETA testers

Hey,

I'm in the process of making a basic encryption progam and would appreciate it if you could take a few moments just to download the program, and try and test it for bugs.
The program can currently only encrypt alpha numerical passages of plain text. Special characters are not supported.

So, please download, run and try enter extreme values or incorrect values.
Basically, just mess around with different values to encrypt and report back on any bugs you may find.

It's a great help. Also, if you have any ideas for improvements to the algorithm or program in general. Please shout out.
Everyone who tests here will be mentioned in the program credits, (this is a smaller part of a much larger program) ;)

I will do a new post with the project

Many thanks,

Jord

Here is the project in full

works great..except if you hit enter..then it just messes everything up

Quote:

---

Originally Posted by kfcSmitty

works great..except if you hit enter..then it just messes everything up

---

Okay, keep the comments coming guys, they all help towards the assembly of the program. :)

Jord

Works fine with me, at first i thought this program ecrypted visual basic coding and was excited ;) lol

Worked! I like the way you coded it and found no bugs in your code ;), I think you should work on the GUI a little more because the colours are a little hard on the eye :afrog:. Other than that, nice work. ;)

hi

ok - will have a play...

i encrypt "123456" and get "61-?#48-?#27-?#13-$56-?#34-?#" - fine..
i try "654321" and get "34-?#56-?#13-$27-?#48-?#61-?#"

similar pattern sort of..

I try a 1 and get "34-?#"
I try a 2 and get "56-?#"
etc...

looks like a straight subsitution cyfer to me but reverses the order..

i try "abcdefghijklmnopqrstuvwxyz" and get

"26-?#59-?#32-?#50-?#39-?#60-?#54-?#10-$64-?#45-?#65-?
#49-?#35-?#43-?#9-$14-?#18-?#47-?#24-?#44-?#21-?#57-?#15-?#36-?#28-?#5-$"

so i guess that "a"="5-$", "b"="28-?" "c"="36-?" etc

so if use the above guess work to encode "test123" I would probably get an encryption of 54-?#, 57-?, 60-?, 54-?#, 34-?#, 56-?#, 13-$

now string them together and reverse them to make the encryption "13-$56-?#34-?#54-?#60-?57-?54-?#"

so if i go into your app and try this as the encypted value and click decrypt I should get test123 back - I try and get the value "teut123"

whoops - look like I miscounted somewhere.... "s" is "10-$" not "60-?" - doh, my fault!!

tried again - got test123 :)

ok now given a cyfertext I can decode using above subsitutions and without using the code...

never use your own design for an encryption process - used a well known and proved algorithums :) (you will aqlways be able to create a process you yourself cannot break but that doesnt necessarily mean its secure)

cheers AJP

Quote:

---

Originally Posted by jpritchard

never use your own design for an encryption process - used a well known and proved algorithums :) (you will aqlways be able to create a process you yourself cannot break but that doesnt necessarily mean its secure)

cheers AJP

---

Nah, i'm too innovative for that ;)
Thanks for the comments, i'll definetly work on the GUI and, yes, it's a basic substitution cyfer but I am working to include encryption keys which relate to a different key comparison string.
I tried it but found quite a few bugs so i'm trying to think of new ways to do that.
It's strange though, try entering quite alot of text - it's fast at first and then seems to slow down.

Why is this?

Thanks,

Jord

Quote:

---

Originally Posted by intraman

Nah, i'm too innovative for that ;)
Thanks for the comments, i'll definetly work on the GUI and, yes, it's a basic substitution cyfer but I am working to include encryption keys which relate to a different key comparison string.
I tried it but found quite a few bugs so i'm trying to think of new ways to do that.
It's strange though, try entering quite alot of text - it's fast at first and then seems to slow down.

Why is this?

Thanks,

Jord

---

Nice encription technique.

One thing though I could not get it to work very well on encripting already encripted text...

Cheers,

RyanJ

That's not encryption, it's an encoder... and one that makes the result very big :eek2:

I don't understand why people like it, it can be decifered easily.

Make it to require a password, and make the result different if the pasword is different for the same string to encrypt, THEN you can label it as "encryption".

found another error...if you enter a comma, when it translates back, its a period

hi

:)

no problems

if your are still wanting to write your own algorithum then using some sort of key as you suggest is the way to go - good processes have all the security in the secrecy of the key used and not via the actual mechanics of the encryption process.

(meaing that you should be able to give people source code for your algorithum and they wont be able to do much with it as the key is the critical part)


at the moment each character always maps to single value (eg "1" is always "34-?") - what that means it that if U encrypted a block of normal english, if you broke the encrpyted text into its 4chr blocks and counted them up the most common 4chr would almost always be the letter "e" the next most popular will probably be a "t" etc etc (this is because e it the most common letter in "normal" english etc) - this is called frequency analysis.
(most home written stuff is undone via this technique)

what I would do is as you go thru the text string encoding using your process have part of 1 chr send an input into the encryption of the next character.

also I would calculate a hash from the message and pad that into the message somewhere - doing that well would have the effect of making 2 near similar message have very different cyfer texts (which would stop idiots like me breaking your encryption by just playing :) )

its a big subject right :)

if you are using a key - I often see people just repeativilty xor the key/password against the cleartext to produce an encryption. These type of thing fail where there is an area of low-values (hex 00) - at this point the encryption texts IS the password used!

password xor low-values = your original password in the encrypted text
(you probably were not going to do that but I though I would mention it)

have fun - I hope works out well

:)

Well i had a list of VB keywords, and i encrypted the lot, took a while, but its all gd. As i type it is decrypting to make sure it all goes back. It looks as if it is doing some pretty decent decrypting, but the progess bar isnt going any where?? Maybe you need to look at that.

The document went back in the same order, but did not retain new lines. for example after encryption and decryption this.

Blah
Balls
Belly

Would be,

Blah. Balls. Belly.

Wow, this is all great help, i'll confront each point individually:

Quote:

---

Nice encription technique.
One thing though I could not get it to work very well on encripting already encripted text...

---

Right you are. Thats because the program does not support special characters, the characters '?#$' are special characters and so decryption is currently impossible. Glad you like the technique. Although simple - it works.

Quote:

---

That's not encryption, it's an encoder... and one that makes the result very big
I don't understand why people like it, it can be decifered easily.
Make it to require a password, and make the result different if the pasword is different for the same string to encrypt, THEN you can label it as "encryption".

---

Encryption means the use of an algorithmic process to transform data into a form in which there is a low probability of assigning meaning without use of a confidential process. Now, I undertstand its currently easy to break but its a start and would fool the average computer user. I did clearly mention that I was going to use a key system. I don't know where you got the encoding idea from because it is nothing like encoding.

Quote:

---

found another error...if you enter a comma, when it translates back, its a period

---

This is by no means the "finished product" but that wasn't actually a bug. You'll notice its programmed that way, again, I refer back to my original quote - "The program can currently only encrypt alpha numerical passages of plain text. Special characters are not supported". All special characters and spaces are returned as periods at the moment.

To jpritchard, your message is too long to quote but thanks for that analysis. I'm going to be using a key and possibly having the characters encryption procedures directly effect the features of the next character. (But thats a bit tricky so it might take me to donate a bit of brainpower and time to!).

"if you are using a key - I often see people just repeativilty xor the key/password against the cleartext to produce an encryption. These type of thing fail where there is an area of low-values (hex 00) - at this point the encryption texts IS the password used!"

Yeh, thats the problem I had. :)
The program returned an error when the character to encrypt turned out to return a number which was the same as the key because then my program would try to search for negative characters in a positive string!

Do you know why it starts fast and slows down??

Quote:

---

Well i had a list of VB keywords, and i encrypted the lot, took a while, but its all gd. As i type it is decrypting to make sure it all goes back. It looks as if it is doing some pretty decent decrypting, but the progess bar isnt going any where?? Maybe you need to look at that.
The document went back in the same order, but did not retain new lines. for example after encryption and decryption this.

---

Thanks. Yes, again whitespace is something I have yet to account for and in the encrypted key each piece of whitespace = '3-$'. (You may have noticed. It won't be hard to convert this to spaces in the decrypted text).

As for the progress bar, I'm not sure it's possible to have it increment accurately, as it is decrypting an unknown number of characters. hmmm., i'll keep pondering this one..

Keep the great comments coming. Many thanks,

Jord

By the way, with this encryption: VB - 128, 160 and 256 Bit File Encryption/Decryption with MD5, SHA1 and SHA256
You can encrypt a file containing ONLY 0's and still you won't find a pattern, and you also won't find the key (or password). And it DOES use XOR to encrypt

MD5, SHA1 AND SHA256 are established algorithms. There is nothing innovative about that. I understand that they may be secure but I don't want the highest level of security, I want the highest level of balance between innovation and security.
As I have already said quite a few times in this thread, I will be changing the alogarithm to make it more secure in the near future. It is just in the prelimiary testing stages. If people used old technology all the time, nothing new would ever be made ;)

Jord

Quote:

---

Originally Posted by intraman

MD5, SHA1 AND SHA256 are established algorithms. There is nothing innovative about that. I understand that they may be secure but I don't want the highest level of security, I want the highest level of balance between innovation and security.
As I have already said quite a few times in this thread, I will be changing the alogarithm to make it more secure in the near future. It is just in the prelimiary testing stages. If people used old technology all the time, nothing new would ever be made ;)

Jord

---

Keep the comments comin'

Jord

Quote:

---

Originally Posted by jpritchard

i encrypt "123456" and get "61-?#48-?#27-?#13-$56-?#34-?#" - fine..
i try "654321" and get "34-?#56-?#13-$27-?#48-?#61-?#"

similar pattern sort of..

I try a 1 and get "34-?#"
I try a 2 and get "56-?#"
etc...

looks like a straight subsitution cyfer to me but reverses the order..

i try "abcdefghijklmnopqrstuvwxyz" and get

"26-?#59-?#32-?#50-?#39-?#60-?#54-?#10-$64-?#45-?#65-?
#49-?#35-?#43-?#9-$14-?#18-?#47-?#24-?#44-?#21-?#57-?#15-?#36-?#28-?#5-$"

so i guess that "a"="5-$", "b"="28-?" "c"="36-?" etc

---

Looks more like character shift encoding to me rather than true encryption, why not just use blowfish encryption etc, or are you working on your own functions?

Quote:

---

Originally Posted by Madboy

Looks more like character shift encoding to me rather than true encryption, why not just use blowfish encryption etc, or are you working on your own functions?

---

Yes it is - I don't want to use a standard function, i'm trying to think about developing my own.