I created a highly specialized program to password scramble my files so no one can view them save for me. It scrambles their hex data using an input string that is ideally only known to the scrambler.Suck it, feds.
Never roll your own crypto. You should've used VeraCrypt, dumbass.
>>107780457>Create your own software that has to be specifically reverse engineered, wasting valuable glownigger hours, to maybe discover a vulnerability that may or may not be usable against the targetvs.>Use popular public open source software which the glowniggers have already combed through and found all the vulnerabilities and researched in depth how to attack the system>They can just re-use the same exploits over and over again, saving time and utilizing economy of scale to recoup all their investments
>>107780415Why not just use password-protected PGP private keys?Also, these new Captchas are painful.
>>107780483A low iq anon wrote this.
>>107780765I hope they pay you decently for spreading this retarded "dont roll your own XYZ" fud, because it's been pretty successful.Would be a shame if you're still sitting in a small dingy office in Langley after all that hard work.
>>107780415Retarded, files have known bytes in their headers. You'd wanna have salted the crypto somehow.
>>107780877Think about this. They've had competitions to find new algorithms for different cryptography standards. People from all over the world competed, people with PhDs, professors, mathematicians, people smarter than you or me. And their submissions all still fell apart. Almost none of the proposed algorithms in these competitions passed rigorous cryptanalysis or security review. Then lets go ahead and assume that for some magical reason your algorithm is cryptographically sound. Well, you still have to implement it without totally compromising its security. Since various attacks can be used to attack cryptography indirectly (through side channels) or directly (through software flaws.) Even following the advice of "not rolling your own crypto" doesn't mean you're completely safe. A good example was the recent vuln in Libsodium. The authors forgot to add some edge case checks for ECC and no one noticed. Even though the library has been around for years at this point. Regular crypto algorithms are widely reviewed, based on math, and don't contain anything magical.
>>107780415https://en.wikipedia.org/wiki/Vigen%C3%A8re_cipher
>>107781147The fact remains that in order to attack your custom algorithm the glowies have to:1. dedicate a cryptography engineer to spend several hours to:2. decompile your code,3. figure out what your algorithm is trying to do,4. find weaknesses in the algorithm and/or bugs in the code, and5. write a tool to exploit those weaknesses to decrypt your data.It is very unlikely that even step 1 will be reached unless you are a very high profile target. In all other cases they will just dump you in jail until you give up the key.Meanwhile if you use some open sores library the glowies already have some backdoors and some unpatched bugs lined up ready to go, and they can use each of these exploits multiple times, perhaps hundreds of times, before they get discovered and patched out.And if your threat model isn't glowies, but rather it's Jamal, then he won't be smart enough to figure any of this shit out even if he wanted to spend 20 hours finding your bitcoin wallet.
>>107781147I couldn't have said it better, anon.
Why not use both? Encrypt / scramble with your own software, then encrypt the output with publicaly available software. If you're writing your own software anyway this would be trivial to add, so it won't even require more effort on a day to day basis.
>>107781268Much of this AI can already do quite efficiently. If there's one thing AI's good at is discerning patterns.
>>107780765Unironically this. A person with this supposed level of ability is leaving breadcrumbs and an insecure username hash.You don't advertise this.
>>107784787>insecure username hash.wdym?
>>107781268IMPLEMENTATIONIMPLEMENTATIONIMPLEMENTATIONMany security bugs are timed "backdoors" with lagged patches.https://youtu.be/wwRYyWn7BEo&t=143
>>107784818https://www.4chan.org/faq#sectrip
>>107780415Ha, I know. I use a Caesar cipher to encrypt/decrypt my files. My little sister was able to crack it, but I’m p sure the US government won’t be able to crack it.
>>107784900Kid, my program uses a smorgasbord of mathematical machinery to achieve its goal. You think I'd be dumb enough to use something that could be brute forced by the average Andy with a dusty MacBook?
>>107785906>You think I'd be dumb enough to use something that could be brute forced by the average Andy with a dusty MacBook?yes
>>107780415>scramble according to an input stringLemme guess, XOR?
>>107787624It's a *fuck ton* more complex than XOR you little sniveling shit.
>>107791167Oh shut it geek.
>>107784338>Much of this AI can already do quite efficiently. If there's one thing AI's good at is discerning patterns.Like this?
>>107784787>insecure username hashWhat the flying fuck are you even trying to say.
>>107780415Post a scrambled file
>>107780415$ cat ~/encrypt.sh #!/bin/bashgpg -o "$2" --s2k-mode 3 --s2k-count 65011712 --s2k-digest-algo SHA384 --s2k-cipher-algo AES256 --symmetric --cipher-algo AES256 "$1"$ cat ~/decrypt.sh #!/bin/bashgpg -o "$2" -d "$1"
>>107791167>more complexI don't believe you. Show us, or fuck off.
