cybersecurity

[silent pier]

Fair enough

Would also have to slightly change the protocol I think to allow for "hey you need to validate"

Ooor just have it be passed automatically per request

[upbeat palm]

How can you generate new tokens and remove old tokens,then you have to send the new tokens for validation.

[silent pier]

Hey you need to validate as in please login again

[upbeat palm]

That's some serious stuff for me.
As I said earlier,good idea. Floppy

[silent pier]

Well it is a course in security after all :p

[upbeat palm]

Course in security?

[silent pier]

Software security to be more specific

[orchid notch]

I had a lecture at the TUM about secure system design once, the prof repeated the 4 principles of security again and again
Encryption
Authenticity
liability
Aaaaand uhhh access control I think

Yeah access control

Apparently that's what makes secure software

[silent pier]

Theres more to in depth ofc

but most of it, in a general list

[upbeat palm]

Let me save it in Google keep.

@silent pier Where can I know about it?

[silent pier]

wdym?

The course im taking at my uni

[upbeat palm]

Oh, sorry.
My professors aren't that good so I have to rely on online resources that's why I asked.

[tight abyss]

@silent pier Instead of bothering with a database, it might be another good approach to do something JWT-style ("Java Web Token")

They basically contain the data you need, like user id and expiration date directly in the token, plus having that payload signed with a private key of the server, so that the client can not change the payload without being detected

That way you never have to actually remember the tokens on the server, you just create them when needed and verify their payload and signature when you receive one

[silent pier]

Ill look into it

[valid furnace]

@silent pier in what context?

basically auth tokens

[safe bear]

@silent pier a possibly good place to look would be OAuth implementations

I don't want to make suggestions because that knowledge is two years rusty and I'd probably tell you something very wrong

[tight abyss]

OAuth is more about how to log in/authenticate with possibly a third party. It still requires you to manage a token to remember the client.

[thorn obsidian]

@silent pier generally the safest way to do token based auth is to create a token based on some random secret + the client IP and store it with a created-at timestamp and an expire-at timestamp in the database, then set it as a cookie in the client browser and check if the token is valid for that IP and if it's expired or not when the client makes a request

make the expiration like 24 hours of no activity or something, and just bump it whenever an action is done with that token

depending on how much security you need you could also have a maximum use time where it forces a token change once every month or whatever

in saying this I would discourage the use of a hash for token authentication since if someone generates a token using the IP and guessing the secret they basically have login permission for that user

but i'm not sure what a good alternative is

I just know it's a good idea to lock it based on IP

probably good to store the client IP in the DB next to a UUID or something

buuut, dynamic IPs are a thing and that's annoying so you could drop it all together and just do UUID with expire times, which is generally accepted as a good enough solution

JWT stuff is also a good idea

[silent pier]

@thorn obsidian I dont really have ips so I guess ill generate a uuid for each user. But how do I give that to the user in a safe way

Actually lemme quote the protocol im creating

[thorn obsidian]

ah

the tokens will most likely need to be a UUID https://en.wikipedia.org/wiki/UUID

you give them to the client via browser cookies, making sure to set the HttpOnly flag and Secure flag to ensure they're only transferred by the client over HTTPS and not accessible via javascript

if it's not on a website then just try to make sure it's transfered in a secure way and stuff

[silent pier]

Its not browser based either, its just over sockets

Final version over TSLsockets

[thorn obsidian]

yea that's fair

most stuff still applies

[silent pier]

So upon reading about how create a UUID it says I need a namespace + name (or high + low significance), in my socket protocol what would this namespace be?

[lilac gulch]

@silent pier Try a different uuid function (v4, i think) that doesnt require a namespace

[silent pier]

Are they still unique then?

There wasn't much about them in the article i read, just that they are based on random numbers

[lilac gulch]

https://docs.python.org/3/library/uuid.html

Namespace determines the "format" for lack of a better word

For socket stuff, uuid1 may be handy

[tight abyss]

If all you need is a unique token, and you'll store them on the server anyway, just generate completely random 128 bit (cryptographic RNG preferably) and add an additional check to avoid the almost impossible chance of having handed out that token already

I see no need to bother with UUID standards or whatnot in your case

[simple orchid]

if you want to guarantee uniqueness you could use a 'snowflake' technique like Discord

those are only 64 bits, too

[silent pier]

The one part that doesnt quite stick with me is how the user uses this token to authenticate

do i just pass their token to the user, and they provide it if they have one

and that's it?

Or would I store a combination of a key (that i give the client) and their username into a hash or something

[simple orchid]

the point is

they give their usual credentials (username/password, email/password, already logged in session) to site A, and get the token. site A also sends the token to site B.

the user proves to site B that they have the token, therefore proving that they were able to log in to site A

(also, i didn't realize you were doing token authentication - a discord-style snowflake id is absolutely inappropriate for that. they should be completely random)

[silent pier]

yee, its a token based authentication for a socket client-server

[thorn obsidian]

no that's not the way it always works @simple orchid

that's just the way Oauth works

cookie tokens are different

@silent pier in this case they send the username and password from the user once, then get a token and store that then reuse it instead of password as authentication

it stops the program needing to store passwords or constantly ask the user for passwords

don't use a hash

just use a UUID function and make sure the output is not a duplicate of another token

tokens should expire etc etc as above

[silent pier]

Right

Only one way to get this rolling then, actually try it

[thorn obsidian]

haha good luck

[simple orchid]

i thought we were talking about oauth

[shell gulch]

Hey everyone

I'm kinda stuck in web-dev atm and I want to get closer to doing information security. I'm curious on certs that get taken seriously by medium / large orgs that I could do?

Namely to do with things like pen testing.

[thorn obsidian]

none of the certs are worth it

everything except an OSCP and maybe the NSA information security masters is useless

orginisations require shit like CEH for jobs, but it's the most backwards stuff ever, you might come out knowing less about security than you did going in

https://twitter.com/CroCyber/status/1043146646647914499

Luka Sikic (@CroCyber)

e. Refund Please #CEH

Twitter

command shell trojan

netcat

[upbeat palm]

@thorn obsidian CEH is total crap.

[thorn obsidian]

yeah

it is

[upbeat palm]

I tried online quiz of CEH and tbh a kid can pass it too without knowing anything.

[valid furnace]

@thorn obsidian that shit is why I cannot stand people that say "I got all of the certifications and stuff"

[thorn obsidian]

lol

yea they're dumb

most of them are useless

[valid furnace]

or people that speak in pure jargon

[thorn obsidian]

double fuck those people

they know what they're doing and it's fucking dumb

they need to stop

[thorn obsidian]

https://interc.pt/2Q14FAi

[valid furnace]

@thorn obsidian
/me immediately goes to download said information

[thorn obsidian]

lol

[safe bear]

Can't say I fully agree with the jargon statement

If you're doing anything with networks and systems, which is a huge portion of cybersec, then you're going to use a lot of jargon

You can sometimes reduce the use of acronyms, but not jargon

Didn't realize CEH was such garbage though, that's good to know...

OSCP is good, and I think the SSCP (Systems Security Certified Practitioner) is also good

Personally, I found doing the CCNA (Cisco Certified Network Associate) was invaluable, since you learn a ton of networking fundamentals and architecture

[rigid plinth]

Hi guys, I'm new to the server, so apologize if I'm asking this in the wrong place. I am a Python coder moving from beginner to intermediette. I am still very new to this, but I do know that I'm not going into web dev. From the different things I poked my head around in, I realized I would like to work either in OS dev (so contribute to Linux kernal or any distros in the long run) and/or work in security. What kind of questions should I be asking myself if I'm thinking of these path(s)? There are a lot of resources out there, but I need to know what to ask - so any suggestions would be helpful 😃

[safe bear]

Check the pins

There are several good sources of information in there

Keep up with those, and maybe try your had at OverTheWire

Regularly read stuff in the field for a while. Do that for a bit and see if you're finding it to be more of a slog or if you're thinking "wow that's cool".

@rigid plinth

[valid furnace]

@safe bear haven't taken any CCNA classes, but I can setup network switches and routers just fine, although it takes me a bit longer. 😉

[safe bear]

I didn't take any either, learned by messing around with old equipment from eBay with my father and reading, and lots of practice tests and quizzes

It was difficult for a while but eventually it clicked

[rigid plinth]

thanks @safe bear , I'm looking at them right now. The Bandit game looks very interesting - do you know if it's suitable for a noob?

[safe bear]

Some of the stuff on there is kind of a waste of time though, like token rings lmao

Yes, it is @rigid plinth

The point is to learn as you go

[valid furnace]

@safe bear my last workplace had problems with rogue DHCP servers, it was both stupid and horrific.

I learned to trace those down from CMTS

[safe bear]

There are some clues, but the whole point is you are supposed to Google and search for answers @rigid plinth

Oh wow

[rigid plinth]

that's awesome. I have no problems at all doing that - just making sure that all of a sudden, I don't have to realize that I need a CS degree to figure it out

[safe bear]

Rogue DHCP eek

haha yeah

it may seem like that, but if you get really stuck, sleep on it, give it another try.

If you're still stuck, there's nothing wrong with looking up the answer.

[rigid plinth]

sounds good. thanks for the words of advice :)
this really seems like a great place to start with security 😃

[safe bear]

Yes, that is a good mix of stuff

[upbeat palm]

I thought of taking CCNA course but now I'm not.

[safe bear]

Why's that?

Self-study worked well for me, but structure is helpful especially if you're having to manage time around other things

[upbeat palm]

Institutions offering CCNA course here don't have a good review.

[valid furnace]

@safe bear without breaking NDA, imagine an endpoint modem fucking up and DHCP serving across the entire cable/cmts network.

[lusty flare]

Hmmm

Not possible to use Rogue DHCP detection on the switches?

We had to implement that here after people kept plugging in stupid shit to the network

[safe bear]

If the modem is the perpetrator...how do you deal with that?

Isn't the equipment pretty low level until you hit the distribution layer?

[lusty flare]

DHCP is layer 2 tech

So it's well within a switches purview

[lusty flare]

Well, i mean i was being a tad simplistic there

but you can block it through layer 2 only

[silent pier]

How high of a prime is required for diffie hellman to be safe?

[lusty flare]

largest number to not be vulnerable to attack i guess

what was that attack vs diffie

NFS?

[silent pier]

unsure about the terminology, im just aware of a brute force way, and man in the middle attack

[lusty flare]

Seems like NFS has been demonstrated against stuff towards 2^750ish

[silent pier]

oof

[lusty flare]

so like 250digits

i guess

[upbeat palm]

Nix is Cryptography expert here.

[lusty flare]

tbh i've just seen people using 2048

¯_(ツ)_/¯

[silent pier]

Time to find my prime

[orchid notch]

Apparently I am cryptography expert now

Feels good

[silent pier]

to write a script to test my way down from 2**800, or just google it

[orchid notch]

You want to find a prime?

[silent pier]

ay

Big prime for diffie hellmans key exchange

[orchid notch]

Generate random numbers in the range of numbers you want your prime to be in and use the miller rabin test on them until you find one

For 1024 bit primes around 40 rounds of miller rabin are recommended

So I guess that should be more than reasonable for you

[upbeat palm]

I'm Cryptography noob.

[silent pier]

does it matter how big the secrets are?

if i have a big p prime

and g i suppose matter, but im unsure how big these should be

[native edge]

dfh uses very very large primes that have been checked for safety

Like over a hundred digits large

[orchid notch]

Minimum or 2^1024 is recommended for p although depending on the duration of your connection you can choose smaller ones as the NFS method takes hours to crack an exchange

If you really want to be safe you gotta take more into consideration though

For example the prime factors of p-1

For g choose a number smaller than p which allows g^a mod p to result in as many numbers as possible (ideally all) between 1 and p-1

@silent pier

[silent pier]

Read g is usually 2

for most cases

[orchid notch]

Well then 2 appears to be the number which allows the most results in the range from 1 to p-1 for the above equation

[silent pier]

what about secret a and b? should they be completely random between 1 and n?

or have a specific size as well

couldnt find much om them

Other than examples with numbers below 100 🎟

[orchid notch]

random iirc

but like

huge random

[upbeat palm]

Random Key?🤔

[orchid notch]

yes

[upbeat palm]

Oh,that would be kinda complicated,imo.

[orchid notch]

no

[upbeat palm]

Random Key for each ciphertext,right?

[orchid notch]

no

random keys for symmetrical key exchange

so one message gets transmitted

[upbeat palm]

Oh, I thought of something else.

[silent pier]

so like 2^512 huge?

[lilac gulch]

thats only a 512 bit key

[silent pier]

Private key~

but it's not some big corp security, just proof of concept really

[violet notch]

do you guys think taking a ethical hacker course is worth or are there better courses out there

[orchid notch]

no

better than what lol

[violet notch]

oh why is that>?

[upbeat palm]

@silent pier For PoC 512 is good.

[upbeat palm]

USBKill,one who gives a high voltage of electricity when plugged in PC?

[violet notch]

So like if someone was to try and use a ducky or something?

[safe bear]

Wasn't aware of usbkill until now, that's neat

I'm not at that level of paranoia (yet)

Good idea extending it to other interfaces commonly tampered

My systems are Windows 10 bare-metal with WSL + various Linux distros in VMs

Yeah, it's only going to work on the typical modern Linux distros

Yeah

I was playing around with those a month or two ago for a random project

They're pretty easy to do

https://github.com/GhostofGoes/Astolfo/blob/master/service.py

Would be neat if could run as a Daemon in WSL

Can't recall if it's possible yet to get them to come up on boot yet though

I might take a look tomorrow at adding Windows functionality to this

Might simply be a matter of calling the right APIs through ctypes or whatever

Quick question though, why AGPL?

It's not going to be used over a network

Ah

Just more complexity than what you already have with GPL...I just put everything I do under MIT nowadays 😛

[upbeat palm]

@thorn obsidian Oh,sorry for that.
But great job dude.

Massachusetts Institute of Technology🤔

[safe bear]

Yes, google "MIT License"

[upbeat palm]

Oh.

[safe bear]

It reduces complication and encourages sharing

So, stopped GPL'ing my stuff a few years back

[upbeat palm]

So the MIT license is used for making programs open source?

[safe bear]

KeePass and LastPass

Because I'm lazy and my setup works efficiently while still having adequate security

[upbeat palm]

You encrypted your database with AES-256,iirc.

[safe bear]

KeePass

[upbeat palm]

👍

[north rover]

KeePassXC because it‘s pretty good

[valid furnace]

openssh is bad, they don't fit in with normal key procedures, don't allow x509 and implement their own garbage standards.

[north rover]

oookay

[valid furnace]

I've been working on a universal standard for people/companies to implement hardware security keys.

right now the biggest stick in my craw is that openssh doesn't seriously support x509 and I'm trying to figure how how to get it to work.

I'm probably going to be forced to make another CA with some weird implementation

but that's more hot garbage

thanks to openssh folks

[upbeat palm]

Certificate Authorization?🤔

[valid furnace]

@thorn obsidian so get this, they wanted to avoid x509 for "unacceptable risk" basically, then go implement their own CA and ssh key mechanisms

@upbeat palm Certificate Authority

[upbeat palm]

Oh yeah,now I remember.
Thanks @valid furnace

[valid furnace]

basically it's unrealistic to be expected to keep a constant up to date list of every allowed/banned user on each of your 1000s of servers.
You simply put a company public key on each server that never changes.
Then any user who's "signed" by the company key will be allowed in.

[upbeat palm]

A type of constant key.

[thorn obsidian]

https://www.nytimes.com/2018/09/28/technology/facebook-hack-data-breach.html

[valid furnace]

pretty sure this belongs in "anti-security"

or "assholes that don't know what security is"

[upbeat palm]

Plenty of people don't know what security is.

A Taiwanese guy challenged Mark Zuckerberg to hack his account and said he'll do it live.

[thorn obsidian]

lol

yeah

[upbeat palm]

The fact is "Facebook" is total insecure.

[lusty flare]

aw shit

hopefully that breach comes through into the HIBP db

if it's a serious one

[upbeat palm]

Which one?

[lusty flare]

the 50m breach for facebook

[safe bear]

I deactivated my Facebook beginning of May, but I don't think I deleted it fully... Need to do that now.

Not that it would matter, they still keep the data around for at least a year.

[thorn obsidian]

I don't think Facebook is really insecure

I'm pretty sure the website that handles my taxes in my country is full of critical security holes. It's just that noone cares to exploit them because it would be pointless

yet, people still use this website to pay their taxes

Ignorance is bliss : )

[sick hawk]

why would it be pointless?

[thorn obsidian]

because I come a country that noone cares about : )

My point is that Facebook is surely more secured than most website people use.

[thorn obsidian]

everyone cares about something

if it's got a security problem someone will hack it and sell the information

[thorn obsidian]

https://curl.haxx.se/docs/CVE-2018-14618.html
libcurl contains a buffer overrun in the NTLM authentication code.

[thorn obsidian]

http://phrack.org/papers/escaping_the_java_sandbox.html

[upbeat palm]

Damn, I missed the Taiwanese guy challenge.
Any news on it? @thorn obsidian

[thorn obsidian]

nope, haven't heard anything

[upbeat palm]

Oh,okay.

[upbeat palm]

@thorn obsidian If I may ask where do you live?

[thorn obsidian]

In a beautiful country 😃

[upbeat palm]

[upbeat palm]

https://www.google.co.in/amp/s/nypost.com/2018/09/29/zuckerbergs-own-facebook-account-got-hacked-in-breach/amp/

50M data included his too.

[thorn obsidian]

Is there a technical details on the vulnerability?

[safe bear]

Not yet

[lusty flare]

fingers crossed that ends up in Troy Hunt's db

[safe bear]

If he can track it down

[lusty flare]

yeah

[orchid notch]

Why fingers crossed for that

[lusty flare]

i think the responsible thing for FB to do would actually be to disclose it

[safe bear]

I doubt Facebook is sharing enough to provide an anonymized list he can use

[lusty flare]

to him specifically

[safe bear]

Since iirc he uses hashed emails

[lusty flare]

HIBP is getting some serious grunt behind it

yeah

everything's hashed

[safe bear]

They might not be at the point yet in the investigation to know the full and precise scope of the breach

[lusty flare]

at least they shouldn't have to an equifax

[thorn obsidian]

I never mock companies when they have security flaws, because if I was the developper, there would be 10 times more 😃

[lusty flare]

:D

[upbeat palm]

🤔

[upbeat palm]

Yp.

[thorn obsidian]

😃

[echo burrow]

IT wasn't a password hack.

It had to do with session cookies that were static.

How such a feature could be implemented without a timeout is beyond comprehension.

[lusty flare]

okay but i'm more curious about what data was taken

[thorn obsidian]

https://www.troyhunt.com/pwned-passwords-now-as-ntlm-hashes/

github using pwned passwords now too, nice

people should use this ^

build it into stuff

[upbeat palm]

Yup.

[thorn obsidian]

https://www.troyhunt.com/the-effectiveness-of-publicly-shaming-bad-security/

[thorn obsidian]

Probably doesn’t even get passed down to the tech team

Lol

[thorn obsidian]

lol

[lusty flare]

oh @thorn obsidian

i saw someone was doing the NTLM hashes as a side project of their own

nice that it's integrated

[thorn obsidian]

you build in API calls or lookups to stop users from choosing dumb passwords

and regular dev is security related

fuck me i don't have time for this

[lilac gulch]

thats their problem then

you can say "Not our fault you picked a dumb password, we told you"

/shrug i dont do user support

[native edge]

disallow a list of the most common passwords

if it's a fixed salt it wouldn't be a problem

[valid furnace]

@thorn obsidian I want the raw database leaks so I can correlate metadata to deanonymize accounts.

but that's just me.

[thorn obsidian]

he doesn't release that on purpose lol

[thorn obsidian]

@thorn obsidian ok so i'm less busy and the stuff you were saying was dumb, so here

as a company it's your responsibility to protect users from security risks, both legally and ethically, so making sure your users don't use a password that is known to have been used before is something that they should be doing. It provides a very large security gain to users and stops credential stuffing, which is one of the most common attacks on people's accounts these days. Checking tens of millions of passwords is very fast, especially if you use the API he provides. The password database contains unsalted passwords so you just compare the plain password hash, and hash+salt when you store it in your own backend.

As for rate limiting, it's mostly ineffective since chances are very high that the user uses the same password on every site, so it will succeed on the first login anyway. IP based rate limits are almost useless, people can use any number of IPs and machines to brute force things, especially with IoT these days.

also lots of companies prevent password reuse, including github, EVE Online, and 1password

they all use the system built by troy hunt

[upbeat palm]

@thorn obsidian I guess that way we/you can prevent "bots" from trying hundred millions of password,but what if they use proxies?

Oh.

[thorn obsidian]

yes

[upbeat palm]

Either way,it won't last forever.

[thorn obsidian]

it's in the NIST standards

people should be doing this

yes

[upbeat palm]

Recpatcha is a good way to avoid bots.

What is preventing password reuse?

Oh.

When I was on telegram people use that old hit and try method to from their so called combos(combination of email and password like this [e-mail]: password123),they use it with proxies and configuration file of a tool and test it on website like Spotify, Netflix etc. That way they earned a shitload of money.
Google password creator is good.

That's like comparing Krillin to Goku SSJ

[lusty flare]

hey i was scrolled up

but seriously

preventing users from using passwords already exposed in breaches

is a fucking brilliant idea

[upbeat palm]

Indeed it is.

Sometime paranoia is good.

[lusty flare]

Everything Xair has said is literally the best practice we could be doing right now

Just because you as a user are technically aware of the threats you face doesn't mean developers or even the company your work for shouldn't actively try to mitigate risk

and that's pretty much what this is

mitigating risk.

tbh now there's NTLM hashes i might see if i can convince work to let me deploy something to our client's environments

it would prevent some really dumb shit from happening

[upbeat palm]

+1

Even though there are so many ways to make secure things no one implement it.

[lusty flare]

i feel like there are bigger problems

people ignoring NIST and NCSC standards for no good reason

Biggest bitch of mine is not allowing pasting into password fields

[upbeat palm]

Biggest bitch of mine?

[lusty flare]

oh, biggest bitch = biggest complaint

:)

bitch being used to mean something that annoys me etc

[upbeat palm]

Oh, I'm not good at understanding puns.

[lusty flare]

it's more like a western turn of phrase

like if someone's complaining about something they're "bitching"

but also if something is actually really cool it's also "bitching"

like "Terry wouldn't stop bitching about how i burned his eggs."

but also "Terry said my eggs were really bitching!"

English is logically broken

[upbeat palm]

Oh, I'm gonna try it.

[thorn obsidian]

english is pretty memed

also yea not allowing pasting into a password field is the worst, anyone who won't allow that i just refuse to interact with tbh

my bank is the most insecure system that i currently use, which says a lot about security in companies

[upbeat palm]

No one thinks that the security will be going to be first priority of digital world in upcoming years.

[chilly elk]

Hi I just released my first big project which is a new penetration testing and exploitation framework. If you are interested in helping add some utilities and plugins feel free to make a PR or get in contact with me. I would really like this to be a community driven project and hope to make it much more usable and updated often! Need all the help I can get. Heres a link to the framework for now 😃 https://github.com/DarkSpiritz/DarkSpiritz

[upbeat palm]

Even if I want to I just can't.

[chilly elk]

@upbeat palm why not?

[upbeat palm]

I don't have any way to pull request, I can't use Git to pull request i.e. no PC.
But your project seems good.

[native edge]

Pull requests are done on the website

[upbeat palm]

I asked Nix and he said after pulling request I have to use Git.

And he said it'll be overcomplicated for me.

[chilly elk]

You can also just send them to me

any plugins you make

or just email and ill upload them myself 😃

[upbeat palm]

Really?
Thanks. @chilly elk

I'm gonna work then.

[chilly elk]

sweet 😃

appreciate it

youll be credited on the README and in the framework

[upbeat palm]

Oh.

[chilly elk]

what?

something wrong w that? what else would u like? @upbeat palm

[upbeat palm]

Nope,that's enough for me.
Thank you, I'll contribute to this project with all my potential.

[chilly elk]

thank you

[thorn obsidian]

@chilly elk what plugins are you looking for

[chilly elk]

@thorn obsidian anything you think would help with penetration testing

[thorn obsidian]

Not anything I can make haha

But good luck!

[chilly elk]

thanks

[orchid notch]

@upbeat palm its the other way round, in order to get to the point where you can create a PR you have to use git, the PR creation itself is done on the website

@chilly elk is there any particular reason you allow python only plugins? For example I wouldve already contributed to metasploit or at least tried to if it wasn't written in ruby, why not use a stdout stdin based plugin system like for example neovim so everyone can contribute in the language he is best at?

If there is nothing against it I would be willing to try and implement it together with a few plugins

[upbeat palm]

@orchid notch That's the reason I dropped the contributing idea.
You're right, I should work on my bot now.

[orchid notch]

I never said that

The last onr

[upbeat palm]

Yup, I said it.

[chilly elk]

@orchid notch well because its a python framework you can use the run() command from plugin_support to run shell code so if you wanted i dont see why you couldnt add your own scripts in whatever language to storage and just have the plugin run that

[orchid notch]

i see

thanks

in that case I'll add a plugin i have in mind soon ™

[chilly elk]

sweet appreciate the support 😃

[thorn obsidian]

https://www.wired.com/story/facebook-hack-single-sign-on-data-exposed/

[orchid notch]

alrighty

next PR is out for now so lets do dis

your installer is somewhat interesting

@chilly elk

[lilac gulch]

oof

[cedar pelican]

@orchid notch where to you get the cool terminal layout?

just a arch feature?

[orchid notch]

powerline9k + config files

im on fedora atm

anyways

lets manually install dis

fun

this installer was sure written for some outdated fedora

this installer just destroyed my softlinks to python2

fkin hell

"fixing python2"

[cedar pelican]

@orchid notch does that work over ssh?

I'm guessing not

[orchid notch]

ofc not

its a prompt

[cedar pelican]

Yes cool

Well fuck you to then

;^)

[orchid notch]

i mean you could install it on the machine and configure it there.......

then it would work

[cedar pelican]

Its a virtual host

[orchid notch]

why is it installing telnetlib

that thing is in stdlib

this code is really damn old

lots of the packages its installing atm arent on pypi anymore

[chilly elk]

@orchid notch what Distro are you on?

[orchid notch]

fedora 28

[chilly elk]

is that dsf package manager?

[orchid notch]

dnf

yes

[chilly elk]

dny badf yeah m

i meant dnf

[orchid notch]

but all the packages its installing are outdated/not in the repos anymore /already fixed it btw

and the pip packages

[chilly elk]

ok

[orchid notch]

lots of them are either not their or in stdlib already

[chilly elk]

it works fine on ubuntu and arch

and debian

maybe its just the fedora

weir

thank you btw

did u look at the installer script

and figure out what was broken

[orchid notch]

i mean

im assuming i have to run this under python3?

oh nice internet lag

wooop

[chilly elk]

its python2

so idk why you would do that

[orchid notch]

wait no

i wanted to write 2

i executed it under 2

[chilly elk]

[orchid notch]

misstyped

[chilly elk]

all good

[orchid notch]

im touch typing from a 60 degree side angle

hard to hit numbers correctly 😐

[chilly elk]

oh boy

why the angle of attack lmao

60 degree

[orchid notch]

uh

fedora on laptop

next to me windows pc because im doing word stuff

so laptop goes left to the screen

[chilly elk]

ah i see

[orchid notch]

fun

so once i chose an exploit for example, how am i supposed to configure it

[chilly elk]

ill try installing a fedora VM to see if i can fix

well you configure it as you would with metasploit

use the 'retarget' command

which will allow you to go through the config and change each value

[orchid notch]

"Unknown syntax: retarget"

[chilly elk]

atm im working on a way to add values without manually changing config.xml

[orchid notch]

you sure?

[chilly elk]

im guessing you already used use

[orchid notch]

yes

[chilly elk]

if so type back

and then retarget

you have to retarget from within the main framework not inside of an exploit

[orchid notch]

wait so

every exploit gets all the variables etc passed

[chilly elk]

only if they call for the

them*

but they can use any value from config.xml

so say they want the target

it would be

ask.target

say they want LHOST

it would be ask.LHOST

using plugin_support

[orchid notch]

ok

[chilly elk]

https://github.com/DarkSpiritz/DarkSpiritz/wiki/Plugins

theres an example plugin there

you can see how the ask feature works

[orchid notch]

oooh

a wiki

didnt see that one

[chilly elk]

yeah i still have to work on it but ill add as ds grows

[orchid notch]

alright

so as i used it recently in a theoretical attack on a web page ill port this drupalgeddon2 ruby script to python now i guess

yay ruby

[chilly elk]

i suck with ruby lol

which is why this is python

[orchid notch]

never even wrote ruby

gwyn is good with ruby, ill just ask her if i have questions i guess

[chilly elk]

you can do anything in python that u can w ruby

just using diff methods

[orchid notch]

yeah i know

[chilly elk]

also try to refrain from using raw input

[orchid notch]

but the ruby syntax

[chilly elk]

just because we would like to be able to just pull from config settings

[orchid notch]

diamond operator for example

<>

==

!=

[chilly elk]

¯_(ツ)_/¯

try not to use print as well

if you can use text(), success(), fail(), and warning()

[orchid notch]

yeah seen it

luckily enough this is well documented ruby so it should be easy to port

i hope

[chilly elk]

i was going to try and port over somethings from PTF but its all weird

it basically just downloads the tool and runs it as command line

[orchid notch]

and for the python2 dependencies you might want to move them to a requirements.txt file?

[chilly elk]

yeah i just cant be bothered to pip freeze

and pipreqs is py3 only for somereason

[orchid notch]

pipfile?

[chilly elk]

whats that

@orchid notch going to work on a new installer

[orchid notch]

pipfile is this new standard used by pipenv

not sure if it likes python 2.7 though

[chilly elk]

oh

well i fixed the install script

go ahead and type reload it should fix @orchid notch

[orchid notch]

im running on a forked version

wouldnt it be better to update from remote and stuff?

[chilly elk]

well its assuming you followed the getting started instructions and cloned from the repo

i dont have a server to be able to host

[thorn obsidian]

https://whynohttps.com/

[orchid notch]

to be fair

lots of those pages are in china

[thorn obsidian]

yea, and china is dumb

but lots of them aren't china too

[tidal oasis]

china is very far from dumb, naive to say

[tawdry hornet]

Hey guys. I've just started to learn source control and I've written a basic program.

I need to use API keys in this code, so I've put the keys in a dotenv file and added that to my gitignore.

And I'm using python-dotenv to pull in the variables.

Is that prudent or should I be doing something better.

When I move this onto AWS serverless infrastructure I'll be using AWs Secrets Manager and a API private key stored in s3.

[tidal oasis]

that's fine

[chilly elk]

https://github.com/DarkSpiritz/DarkSpiritz3
Dark Spiritz Base Framework in Python3. Everything in the original wiki still works 😉

[lean skiff]

here's a CTF challenge

Tip - There's a reason it's a .txt and not .asm file.

[upbeat palm]

You made it on your own​?

[upbeat palm]

https://www.wi-fi.org/news-events/newsroom/wi-fi-alliance-introduces-wi-fi-6

[safe bear]

lmao

I never realized there were Wi-Fi numbers

I just always knew them as 802.11x

ax is new to me tho

and apparently they're skipping over ad

which is 60ghz

[upbeat palm]

Article says ax going to be the fastest, skipping over ad?

[safe bear]

802.11ad

[upbeat palm]

Oh.

[lean skiff]

@upbeat palm yes

[upbeat palm]

@lean skiff Looks good.

[lusty flare]

so many wifi letters

it's hard to keep track

a, b, g, n, ac, ad, af, ah, ai, aj, aq, ax, ay

plus the 2007, 2012 and 2016's

lol, they're calling ax "WiFi 6"

and yeah @safe beari think they're skipping over ad because other technological improvements happened

that were more significant i guess

802.11ay will be 60ghz again though

peak transmission of 20gbps apparently :o

[upbeat palm]

Can't do anything.

[upbeat palm]

Peak transmission is 20gbps,damn.

[thorn obsidian]

What’s this ?

[lusty flare]

WiFi standards

802.11ay supposed has 20gbps peak transmission

[tight abyss]

ayy

[upbeat palm]

@lean skiff Is it a reverse engineering challenge?

@thorn obsidian Read above article.
Or this one
https://thehackernews.com/2018/10/wifi-version-6.html?m=1

[thorn obsidian]

so its about hacking

lol

i wanna become a pro hacker

[lusty flare]

They're called security researchers

[upbeat palm]

Kind of,it's related to wireless security.
I want to become security engineer.
Checkout the above CTF.

[lusty flare]

and it's more boring than you think

[upbeat palm]

Yup.

[lusty flare]

pen testing is a subset of security researcher and it comes with a hell of a lot of paperwork

[upbeat palm]

Yup,yup and yup.
Sometimes it can be boring and sometimes it can be challenging.

[thorn obsidian]

You know,I know a guy who did this hacking competion thing where he had to hack into some computer or something and he won,now he's in the MI5 or MI6 whatever it is.

[upbeat palm]

You mean CTFs at DEFCON?

[lusty flare]

MI5/6 don't employ people off of competitions alone o_O

He was probably a solid researcher already

[upbeat palm]

MI?🤔

[thorn obsidian]

Well idk lol

[lusty flare]

The UK intelligence services Robin

[upbeat palm]

OMG.

[thorn obsidian]

But he did get into mI something from knowing how to "hack"

[upbeat palm]

India doesn't have a intelligence service, afaik.

[lusty flare]

Yeah you do

All countries do

pretty much

[thorn obsidian]

your indian? @upbeat palm

[lusty flare]

The Indian international intelligence agency is called "R&AW"

[upbeat palm]

But I guess it's confidential.
Someone hacked(a government website) and posted Doraemon photo.
Yup.

Yup, Research and Analysis Wing.

[thorn obsidian]

Anyways could some of you guys like hack discord n stuff??

[upbeat palm]

Hack discord and stuff?

[lusty flare]

Errr, probably not.

[thorn obsidian]

Yes like are you pro hackers?

[lusty flare]

Security really isn't as exciting as you think

[thorn obsidian]

Should i watch out whaty i say or will a bullet will go through my window in next 30mins?

[lusty flare]

I don't specifically work in security but I do secure networks / systems for a company

And just yesterday i had to prove to a bank that a client wasn't hacked

[thorn obsidian]

BUT HE WAS

HEHE

[lusty flare]

Erm, no.

It was a fraud investigation and the write up was boring

[upbeat palm]

If were one, I should've been hacking something right now.But I need to learn a lot i.e. Cryptography(learning now) after that networking then reverse engineering etc.

[lusty flare]

Like 2-3 pages of A4 explaining why the specific fraud couldn't have occured resulting from a data breach

After i did meticulous evidence gathering

[thorn obsidian]

omfg

i cant write 1 page at school

[lusty flare]

Good luck working in IT.

:P

[lean skiff]

@upbeat palm crypto

[thorn obsidian]

im doing ITC so like html etc

[upbeat palm]

Security involves lot of writing and communication.

@lean skiff Looks good.

[lusty flare]

Pretty sure the C stands for communication in that

¬_¬

[silent pier]

Hacking software like discord is really not something this channel is for. Keep it clean please

[lusty flare]

^

[thorn obsidian]

Sorry b

[lusty flare]

Also it's a super dumb thing to even try to do.

[thorn obsidian]

How come?

[lusty flare]

It's illegal and what do you even expect to gain out of it even if you succeed?

A crap load of login credentials that you then have to get rid of somehow?

[thorn obsidian]

yes

[native edge]

A bunch of hackers just go caught in my country

[thorn obsidian]

where u from

[native edge]

netherlands

[thorn obsidian]

do you smoke weed?

[native edge]

They were a group of russians trying to steal toxic gass

No

Weed is illegal here

[thorn obsidian]

IS IT?

Sorry off topic

[upbeat palm]

Recently Korea Hacking team came into the show.
They tried to hack world bank.

[lusty flare]

APT8 or whatever

North Korea have apparently been breaking into a lot of national banks recently

for smaller nations

[upbeat palm]

https://thehackernews.com/2018/09/linux-kernel-exploit.html?m=1

Google security engineers are awesome.
Extraordinary minds.

@lusty flare Yup,in your country?

[lusty flare]

Hell no.

The UK has a fairly robust banking sector for now. ;)

[upbeat palm]

I always want to visit UK for vacation, will visit after getting job.

[thorn obsidian]

Robin it's pretty boring unless you go to london but that is wayy too busy for my liking.

[lusty flare]

Cardiff is a nice place. Plus there's lots of historic stuff to see all over the country.

It's not all down in London ;)

[thorn obsidian]

Well if you like historic things go to Nottingham(home town BIG UP) see like robin hood,castle,museum,some caves(more like museum)

[upbeat palm]

Let's move to #ot2-the-original-pubsta

[upbeat palm]

@thorn obsidian

[thorn obsidian]

ermm

[upbeat palm]

Tip - There's a reason it's a .txt and not .asm file.

[thorn obsidian]

i'd rather not download it

[upbeat palm]

😕

[thorn obsidian]

i dont get it

[upbeat palm]

It's not a virus/worm/trojan dude,it's a CTF.

[thorn obsidian]

i downloaded it

but idk what it is

00000001 27 DAA
00000002 60 PUSHA
00000003 5F POP EDI
00000004 4D DEC EBP
00000005 2338 AND EDI,DWORD PTR [EAX]
00000007 6E OUTS DX,BYTE PTR [ESI]
00000008 3C7D CMP AL,7D
0000000A 7C46 JL 00000052
0000000C 45 INC EBP
0000000D 56 PUSH ESI
0000000E 305353 XOR BYTE PTR [EBX+53],DL
00000011 64743E JE 00000052
00000014 722F JB 00000045
00000016 6F OUTS DX,DWORD PTR [ESI]
00000017 4B DEC EBX
00000018 4B DEC EBX
00000019 256C472234 AND EAX,3422476C
0000001E 7E67 JLE 00000087
00000020 317B63 XOR DWORD PTR [EBX+63],EDI
00000023 636161 ARPL WORD PTR [ECX+61],SP
00000026 762A JBE 00000052
00000028 294C7877 SUB DWORD PTR [EAX+EDI*2+77],ECX
00

wtf is that

[upbeat palm]

How are you going to be a pro hacker then?🤔

[lusty flare]

that's assembly

[thorn obsidian]

idk

its just numbers and letters

[upbeat palm]

Yup.

[lusty flare]

I got a great book on ASM

It's from like the 90's

boss gave it to me

[upbeat palm]

Oh,man
I want to learn assembly but it's freaking complicated.

[lusty flare]

that's why i haven't read the book

¬_¬

[thorn obsidian]

it's easy asf

[upbeat palm]

tears of blood
Wonder if it was.

[thorn obsidian]

0000004A 58 POP EAX this means that is a assembly of POP EAX and POP EAX is 1512

Easy

[upbeat palm]

👏

Solve it and do a write-up.

[thorn obsidian]

ok

00000001 27                              DAA
00000002 60                              PUSHA
00000003 5F                              POP EDI
00000004 4D                              DEC EBP
00000005 2338                            AND EDI,DWORD PTR [EAX]
00000007 6E                              OUTS DX,BYTE PTR [ESI]
00000008 3C7D                            CMP AL,7D
0000000A 7C46                            JL 00000052
0000000C 45                              INC EBP
0000000D 56                              PUSH ESI
0000000E 305353                          XOR BYTE PTR [EBX+53],DL
00000011 64743E                          JE 00000052
00000014 722F                            JB 00000045
00000016 6F                              OUTS DX,DWORD PTR [ESI]
00000017 4B                              DEC EBX
00000018 4B                              DEC EBX
00000019 256C472234                      AND EAX,3422476C
``` These mean bob is going bankrupt becuse he lost 159313£ and he only has 50003$

dun

[upbeat palm]

What the 😕

[thorn obsidian]

yh ikr i was confused at the start

[upbeat palm]

Tag the creator of this CTF.

[thorn obsidian]

me?

[upbeat palm]

LappySheep.

[thorn obsidian]

@lean skiff

ok

done

am i getting hacked now?

[upbeat palm]

Where is flag?🤔

[thorn obsidian]

upstairs right room

[lusty flare]

Taking the piss is a great way to not get taken seriously

[thorn obsidian]

Sorry?

[upbeat palm]

I'm out now.
I can't take this anymore.

[lusty flare]

Same.

[thorn obsidian]

Same.

[orchid notch]

what is going on here???

[thorn obsidian]

just teaching some CTF

[upbeat palm]

@orchid notch Dude,there is a crypto CTF,see if you can solve it or not.

[orchid notch]

where when what

[upbeat palm]

What should I explain first?
Where?
When? or
What?

See above chats.

[orchid notch]

scrolling

roles eyes

is it that asm file

[lusty flare]

yes

[orchid notch]

and whats the challenge?

[thorn obsidian]

its pretty easy tbh

[orchid notch]

if there is no challenge its not easy

[lusty flare]

Well I think the guy was being cryptic in saying it's a .txt file not an ASM for a reason

idk what that reason is

[upbeat palm]

Same.

@thorn obsidian Solve it then.

[thorn obsidian]

i alreeady did

[upbeat palm]

Oh,yeah.

If none of us will be able to solve this, I'll ask for a write up from that guy.

[lean skiff]

@thorn obsidian hi yes

i was pinged

[thorn obsidian]

I got told too ping u by @upbeat palm

not sure why

[lean skiff]

i can give another hint

[upbeat palm]

That's a good thing.

[lean skiff]

tip - the offsets are very important

[upbeat palm]

O_o

@lean skiff One thing,the reason it is not .asm file is because it can't compiled?🤔

[lean skiff]

no

[upbeat palm]

Gotta say,you did a good work on it. @lean skiff

[lean skiff]

thank

this was a mean challenge tbh

i wont be surprised if no one does it

[upbeat palm]

😂

I'm a beginner (most of us here), don't expect much.

[lean skiff]

oh thats fine

[thorn obsidian]

no me a pro

i solved

https://gyazo.com/1f64b343b651fc88bfe2ac64f92e9234

[lusty flare]

You know the topical channels are used for topical chat and not just dicking about, right?

[upbeat palm]

I'm done with this challenge, should I ask for a write up or you guys are solving?

[valid furnace]

https://www.reddit.com/r/news/comments/9lac9k/china_used_a_tiny_chip_in_a_hack_that_infiltrated/

[upbeat palm]

That's a good info, thanks @valid furnace

[valid furnace]

The illicit chips could do all this because they were connected to the baseboard management controller, a kind of superchip that administrators use to remotely log in to problematic servers, giving them access to the most sensitive code even on machines that have crashed or are turned off.

BMC modification

fucking evil

[lusty flare]

oh

DRAC type stuff

[valid furnace]

as I understand it, forging microchips in a way that they wind up on computer motherboards and allow remote attacks

[lusty flare]

You're aware of what DRACs do?

[valid furnace]

nope

[lusty flare]

There's a similar flaw posted here earlier i think about flaws in Dell's iDRAC

Well BMC's have their dirty little fingers in almost every part of a server

Meaning that you can completely control stuff through the management port

Including full OS re-installs etc etc

[valid furnace]

OH, yeah, of course

[lusty flare]

So being able to get into the BMC allows you to have total control of the server

through any OS reinstall

[valid furnace]

management ports on servers. I tend to leave them disconnected

but I don't typically deal with that scale.

[lusty flare]

Well, you can't really do that if you've got a lot of shit to manage

Or if you've got a server in someone elses building

[valid furnace]

sure.

[lusty flare]

Usually the management networks are completely closed

but if a sys admin gets popped they might be connected to the management network also

[valid furnace]

it allows you to remotely access the monitor, start it, shutit down, access other control features

[lusty flare]

well, less so access the monitor

[valid furnace]

direct access to the machine basically.

[lusty flare]

it's more like a router's web panel

you're thinking of Network KVMs

[valid furnace]

sure, my servers have a remote console that allows me to access the video

through the BMC

[lusty flare]

i dunno how advanced they've got now, but last time i messed with Dell's iDRAC it was all shells and web panels

[valid furnace]

oh yeah, it's full on video rendering in browser.

[lusty flare]

Nifty

Anyway

If you can exploit a BMC you've pretty much got persistence

and you can't really audit the BMC

except by monitoring traffic on the management network itself

[valid furnace]

yeah. :/

[lusty flare]

It's a similar issue to most firmware type persistence

There's not really AV for firmware

Closest we have iirc is stuff like TPM to check integrity

but in many cases that's poorly implemented and allows for some dirty tricks with BIOS exploits

That sort of attack is going to be hella rare though

super targeted super nation-state

[upbeat palm]

In a nutshell,it'll give full access to it,man security indeed requires a lot of brainf*cking.

For making BIOS exploits , I need to borrow your assembly book ,Bisk.

[lusty flare]

;D

[lean skiff]

@upbeat palm I cannot provide you with the writeup yet

as the challenge is still open on a different server

once the challenge is completed on the main server either the completer or I will publish the writeup

[thorn obsidian]

whats a writeup? @lean skiff

[lusty flare]

It's where you write things up about what you did to do the thing.

A.k.a. a report.

Which is.... as I've said before... A shit load of what security research (HACKING) is.

[tepid tundra]

Holy shit that’s insane

And also wow they must be smart af

Like literal geniuses

[valid furnace]

@lusty flare oh, also, I figured I should tell you the full major features of current BMCs
You can fully remotely:

• Attach USB/CD Drives or load ISOs
• view the vga/screen console and access keyboard/mouse
• control the server power functions, monitor some things

That's the jest. You don't need to sit in front of a server to set it up or wipe it. You can manage them all remotely entirely.

automatic boot, automatic deployment, automatic setup, automatic power management

[lusty flare]

I'm aware of what they do

It's my job

The VGA thing is new to me, to be fair.

[valid furnace]

ah sorry. 😉

[upbeat palm]

@lean skiff Oh,no problem but ping me when you'll provide write up.

Lord Bisk is indeed the Lord of networking here.

[lusty flare]

eh, don't say that

people will raise their expectations of me

that's a tricky position to hold

;D

[upbeat palm]

Don't know about other, but I'm the one.
:P

https://securingtomorrow.mcafee.com/mcafee-labs/mcafee-labs-threats-report-highlights-cryptojacking-blockchain-mobile-security-issues/

@tepid tundra Yup.

[upbeat palm]

There are two Scott?
Ransomware is most common cyber attacks these days.

https://www.redhat.com/en/blog/what-you-need-know-about-first-ever-dnssec-root-key-rollover-october-11-2018

[safe bear]

Wi-Fi Alliance may have skipped over ad, but IEEE didn't

Hence the existence of 802.11ad

The whole SuperMicro thing smells fishy since all three parties, including the two impacted, actively denied it

Which as was mentioned on Ars Technica, is not what you'd normally see if they were simply under a gag order

iDRACs are handy

Obviously, you put them on a separate network with a different set of security controls (read: much stronger)

[upbeat palm]

802.11ad is the fastest as of now.

[safe bear]

Yes

Also the lowest range since 60ghz

[upbeat palm]

Yup,and kinda expensive too.
I guess you did some research on it,right?

[safe bear]

No

[thorn obsidian]

lol

i mean, that seems crazy

they'd have trouble going through people, like it would distort the signal when you walk past someone LOL

[fervent patio]

stop breathing, my video is buffering!

[thorn obsidian]

literally yes

[safe bear]

It's intended for small spaces, like a living room, conference room, office, FBI surveillance van, etc.

Which is not a whole lot, and it rather reduces mobility

I'm quite experienced with this

[fervent patio]

i would have ethernet to my phone and tablet if i could tbh. fuck wi-fi :D

[safe bear]

I've definitely had 3 ethernet connections to the same laptop on multiple occasions, and possibly 4 once sometime

One

And two USB-Ethernet adapters

And I think they were either directly mapped or bridged into VMs

All told I probably have close to a dozen USB-ethernet adapters at this point

They're so handy

[safe bear]

Nope Windows

VMs are usually Kali or a Ubuntu devbox or whatever

[upbeat palm]

So much VMs he can't even remember the names.

[safe bear]

Yes actually

Which is annoying when I forget what computer a VM is on

Wish I could just magically sync everything

[lusty flare]

USB ethernet is 😙 👌

hardest part is keeping track of all the networks

my work PC is on 3 different networks

[upbeat palm]

Damn.

@lean skiff Is that challenge over?🤔

By track you mean network traffic?

[lusty flare]

yeah, making sure you don't flow stuff in the wrong place

[upbeat palm]

Oh, gotcha.

BTW have you ever used Vega or Burpsuite?

[lusty flare]

nope

[lean skiff]

@upbeat palm no

its open until its completed

[upbeat palm]

😖
There's no time limit or something?

[safe bear]

Those are web app exploit tools, know of but haven't used

[lean skiff]

@upbeat palm no

my most recent "long and difficult challenge" lasted a few weeks i think

[lean skiff]

reposting CTF challenge

https://www.virustotal.com/#/file/0d03ab589d01ba2f172defcda613552a13c05b2f7058a04d1f535c2c825c58a6/detection

[inland oyster]

would anyone be able to explain how encrypting and decrypting using integers works in further detail? my teacher has not explained it to where i can understand it.

[lean skiff]

@inland oyster using integers as in

the integer is to be encrypted/decrypted

[inland oyster]

yes

[lean skiff]

or the integer is used in the encryption

oh

that makes more sense

[inland oyster]

using modulo

[lean skiff]

I'm assuming you're talking about the RSA encryption/decryption process?

[inland oyster]

i don't know what RSA is

[lean skiff]

or a variation/derivative of it

RSA is an asymmetric-key/password method of encrypting data

[inland oyster]

def encrypt(data):
newStr=""
for i in data:
i=data(i)
newStr+=i%10+7
return data

[lean skiff]

by using the fact that large primes are hard to decompose

i don't understand why it would be modulo before adding

[inland oyster]

thats what i did

not the teacher

[lean skiff]

if the list is 10 long and you add 7 to 5, you get 12 which is still over the 10 limit

it needs to be adding first

i'm assuming from the code that there's a list of 10 items

oh right because its a digit

lmao

okay

so

you need to add 7 before modulo-10-ing it

or actually

all you have to do

is instead of using modulo

just

1 sec

[inland oyster]

35%10 = 5.

is basically what the teacher wants us to use

[lean skiff]

yes

a modulo b = ((a/b)-(a//b))*b

35 % 10 = 35 / 10 - 35//10 * 10 = (3.5 - 3)*10 = 5

you are getting the remainder from a division

[inland oyster]

yeah i understand that process but putting it together in the code is where i struggle

'

[lean skiff]

and you dont need to use modulo if you have the digits

because all you have to do is read the least significant digit

aka the units value

if the number is 14

adding 7 gives 21

and reading the last digit is 1

so using this method you can skip a cycle

[orchid notch]

are we discussing encryptions?

[lean skiff]

not even encryption tbh

i wouldn't call this an encryption

[orchid notch]

im seeing RSA there

[safe bear]

Do you just have a script subscribed to the word "RSA" in this server Nix?

[lean skiff]

def encrypt(data):
    final_string = ""
    for character in data:
        integer = int(character)
        new_integer = ( integer + 7 ) % 10
        new_string = str(new_integer)
        final_string += new_string
    return final_string

[orchid notch]

no

i just came back from dinner

[lean skiff]

you want the addition before the modulus operator

[orchid notch]

and saw oh people are writing in security

[safe bear]

Umm hmm

[orchid notch]

and then i saw modulo

[lean skiff]

since the modulus operator is what is required to make sure it doesn't go out of bounds

[orchid notch]

and then i thought well modulo is quite intersting when it comes to encryption lets read up on history

[lean skiff]

if you modulo first before addition, out-of-bounds is still possible

also if you really wanted a good method of encryption

you should look up on modular exponentiation

used in prime-decomposition related encryption methods like RSA and PGP (iirc?)

[orchid notch]

yes

well

[lean skiff]

cool i'm not stupid then

[orchid notch]

PGP isnt an encrypt method

[lean skiff]

well

[inland oyster]

thats deep 😛

[orchid notch]

its a collection of ciphers (including RSA)

[lean skiff]

PGP uses 2048bit RSA though methought

or OpenPGP anyway

[orchid notch]

but there are more ways than prime stuff to do good encryption

[lean skiff]

@inland oyster if you really want to you can add a cryptographically pseudorandom key generator

[orchid notch]

for example Substitution Permutation Networks

like AES

[lean skiff]

which runs through the string and encrypts each with a different digit

which is better imo

[orchid notch]

that would result in a really long key though

[lean skiff]

not really

you could define an encryption standard and set a salt for a key before running it through

[inland oyster]

all they're asking for is a 4 digit number to convert to another 4 digit number using modulo and print it to you

[orchid notch]

if it comes to modern symmetrical encryption the majority relies on some block cipher nowdays

oh we're not really talking encryption then

[lean skiff]

mhm

[orchid notch]

sad

[lean skiff]

that's what i said earlier smh

[inland oyster]

lol

[lean skiff]

@orchid notch also just curious

did you have a go at my CTF challenge

[orchid notch]

i looked at it and had some thoughts but i eventually decided that the war with word over the colorisation of my encryption paper thingy for school and now my latex table war for the same thing because i gave word up because of the color problems might be more importnat