AutoHotkey Community

[Sakurako]

Note: this script is only sample of how one feature in Sakurako's encryption method would work (but that of the method would probably use something as Cos instead of simply adding the password), for security the sample would be only recommended for personal use
update version might use improved encryption and not likely to decrypt the code from previous version

[Sakurako]

[Sakurako]

Special Version

[Sakurako]

Original Version

[SMOKE 1990]

I got a question for you that you could pm me about so i can find it easier, where do enter this into my script in order for it to encrypt after i complie it. I dont mind letting a few people use my program i just dont want them to know how its made.

[tic]

just put a password and /nodecompile switch when you compile

[scottmeyer]

Hello,

Your script autohotkey for encryption is excellent!

I want to use part of your program in my program, in order to give information to my program without the user does not know what it is.

I would need a GUI requesting the code (encrypted) and an OK button.

I also hope that the code (key phrase) is placed directly in my AutoHotKey program, we should not box asking GUI code.

Sorry I do not speak English well, I am french ^ ^
J'espére that I managed to make myself understood

Thank you in advance for your help

[Sakurako]

[Rhys]

Thanks for sharing, this is really fun to use. The GUI makes it very simple to play around with. I wonder how secure this is, but I don't know enough about crypto to find out (yet). I've read some really interesting discussions on this forum about it, though.

Can someone tell me what this says?
662deab67f4e1ae2a7604dceefb084ff0edbb4354f10d8a32f4e0cd7af36f213d7a91c2503c5477b2cf8727e511beead45
_________________
[Join IRC!]

[scottmeyer]

Hi !

I actually wish to remove "Gui, Add, Edit, vCode"!

But if I do, and I try to replace this line:
Code = Key Phrase

It is no longer working ...

An idea?

[ EDIT ]
Thank You for your help, it's ok, this program is excellent ^^

[Rhys]

@ScottMeyer: It doesn't like certain key phrases for some reason. I didn't look too closely, but there's a function that checks the key phrase you use for something.

This worked for me (commented out two lines and added a definition of the Code variable:

[Sakurako]

If password not long enough to encrypt the text, decryption attempt would probably be as simple as detecting the number of invalid characters. But if conversion method using only valid characters, the number of invalid would be 0, and how secure it is probably would depend on ENum^PL (ENum = the number of valid characters in the conversion list, PL = the length of password but not more than that of text). Another advantage of valid character conversion is to avoid the double file size problem ~

Conversion method similar to following would be recommended if text using only Chr(9) and Chr(32) ~ Chr(126) ...

[Rhys]

As I said, I'm really not well-versed in crypto theory, but I had an idea today that may or may not have merit. To me, a potential way to defeat (or at least make much more difficult) decryption attacks is to make your original message less like 'normal' text.

For example, eliminating vowels (which often results in semi-understandable messages) or even scrambling the middle letters of words.

Example 1:
Fr xmpl, lmntng vwls (whch oftn rslts n sm-ndrstndbl mssgs) r vn scrmblng th mddl lttrs f wrds.
Example 2:
For elampxe, eilniitmang vlwoes (wchih oeftn rltuess in semi-urdtnsbnldaeae messages) or even smalncibrg the mdilde lettres of wrods.

Obviously these methods aren't without their drawbacks - Short words are harder to understand in the first example and are not scrambled in the second. This could be minimized by using longer words, but neither is a perfect way to obfuscate a secret message prior to encryption. This type of obfuscation is also (for lack of a better term) lossy, and could create ambiguity where it isn't welcome (maybe my recipient would laminate all of his vowels instead of eliminating them!).

In the case of the missing vowels, it might even aid decryption if the attacker is able to detect a reduced character set (or increased occurance of 1-3 letter words) and deduce that there are no vowels in the message.

I thought it would be cool if there were a way to easily obfuscate the text of a secret message that is perfectly readable, even after obfuscation. In that case, decryption attempts (I think) would be much more difficult with very little lost functionality, especially in contrast to the two above.

What I suggest would be a function that runs before the secret message is encrypted that converts most characters to a random character that is identical, except for an accent.

Example 3:
Add random accents to some letters like this.
Âdd řåñďom acċèntś tö śõmë leťŧerş ļîkè tĥĩs.

Wouldn't that make malicious decryption much more difficult but also be very easy to obfuscate / de-obfuscate (sorry if these are incorrect terms!) the text with very little 'loss' (you'd lose any 'valid' accents, señor!)? If de-obfuscation is a weak point, it isn't really necessary to understand the obfuscated message. Maybe this method is already known and used, but if not, why not?

Edit: Another benefit to this method is that the same message would never look the same even if encrypted with the same key, since the characters would be different each time.
_________________
[Join IRC!]

[Sakurako]

Not sure whether vowel encryption would work on US/UK system or not, but probably difficult to support international mode ~
_________________

• ControlClick in game
• Coordinate Color Pick
• Encrypt Utility

[Lawrence]

Your script here is so cooooooool. I used it in my program.. thank you!