« first day (1525 days earlier)      last day (3653 days later) » 

12:03
@Flyk Physical copies don't really work for being pulled through a cable.
what is a "physical copy" ?
what purpose does that serve?
you know what
one of my releases is available on a torrent site
mixed emotions encapsulated in irony
I was on said torrent site ripping off tv shows
I am partially sad because this means I'm not getting paid as much
but I am also partially happy because this means that (a) somebody somewhere did, and (b) those that wouldn't normally still get to hear it
what a moral dilemma
it also means somebody cared enough to share it
hmmm
*channels inner @Simon * Wait, I can download my own copy of @flyk
sure, it's out there... somewhere
@Flyk I have the bad habit of collecting CDs.
@Arperum you mean VDs.
12:12
no he meant STDS
@Flyk thats what I said
Sexually transmitted infections (STI), also referred to as sexually transmitted diseases (STD) and venereal diseases (VD), are infections that are commonly spread by sex, especially vaginal intercourse, anal sex and oral sex. Most STIs initially do not cause symptoms. This results in a greater risk of passing the disease on to others. Symptoms and signs of disease may include: vaginal discharge, penile discharge, ulcers on or around the genitals, and pelvic pain. STIs acquired before or during birth may result in poor outcomes for the baby. Some STIs may cause problems with the ability to g...
I guess they stopped calling it "VD" before your time
but kudos, you ruined yet another joke. And this one was great.
@TerryChia It does ring a bell here as well
Cool, after 3 weeks of work I finally got my interpreter to add 2 numbers. \o/
@TerryChia why didnt you just add them yourself and ask her to interpret the result?
@AviD That's sexist.
12:16
@AviD I guess
@AviD it was, sorry
@AviD That's sexist.
@TerryChia oh I see what is going on. You are implying that because your interpreter couldnt add, I must have assumed it was a she.
not only is that sexist, its racist.
@AviD Heh.
also anti-semitic.
Kidding aside, this is a pretty fun project to work on.
your face is... facist
2
no, that doesn't work
well, it could
12:18
@AviD I enter the DMZ and I find that as the first thing I see
Why am I not surprised?
@Nוnɛfוngɛrϛ so, business as usual.
ssdd
@AviD Basically yep
welcome to the DMZ, moral standards are on the other side of the firewall
8
for the record, @TerryChia and @Flyk, I usually refer to anonymous professionals as "she". It would be sexist to always refer to them as "he". look it up.
12:20
@AviD it's sexist to call them she, as well
use they
@Flyk thats pluralist.
much better than sexism
how good is 256-bit AES?
@Flyk Less good than 128-bit AES.
really?
that doesn't seem logical
@Flyk It's much slower than AES-128 but does not offer any compelling benefits over AES-128.
12:28
ok and how is 128 bit AES vs 128 bit blowfish
@TerryChia thats not accurate.
I believe that should be "necessary benefits... that are worthwhile"
blowfish is also quite slow in comparison. To my mind, AES-256 is only really of use if you think your adversary has a quantum computer that can run Grover's (Correct me if I'm wrong) search over a 2^64 space in a "reasonable" about of time.
i.e. it is slightly better, in theory, but the betterer part is moot.
if I'm VPN'd twice, and both VPNs are 256 bit AES, would that be why my internet connectivity is nowhere near as fast as without the VPNs, even though both VPNs have no bandwidth caps?
@Flyk There hasn't been any (I think, not 100% sure about Blowfish now) attacks that will break the security of either cipher and AES at this point in time is probably much more scrutinized and faster (thanks to AES-NI).
@Tinned_Tuna Aye, and at that point in time you have far worse things to worry about since public key crypto is basically broken.
12:31
@Flyk well you're doing two additional hops and the VPN servers you're using may well be heavily loaded
@Tinned_Tuna the slowest part of blofish is the setup time. After that it tends to be quite speedy (in most circumstances), though not as fast as AES with CPU support.
@Flyk you're putting more layers around your communications + lots of computation. You can't expect to have the same bandwidth
so their ability to process data is a bottleneck
@RоryMcCune well sure, that and my exit point on the internet is geographically thousands of miles from my current location
@AviD AES-NI really is a game-changer for speed.
12:32
@TerryChia yeah. Where available.
@TerryChia indeed, it's more that I'd rather have the justification for completeness.
meh, my VPNs need paying for and I'm just looking at whether I keep the ones I'm currently with or changing them, is all
sorry for starting the on topic conversation in the security channel
let's talk ponies
@TerryChia (thats the CPU command set for AES, right?)
@TerryChia but I am quite the stick in the mud :-p
@AviD Yep, on x86 platforms. I think most Intel CPUs has them for quite some time now.
12:33
@AviD AES-NI is the instruction set for accelerating AES work, yes
there's a couple of instructions, for key setup and one for performing a round.
@TerryChia so I should stick with AES regardless, is what you're saying?
@Flyk Yes.
@Flyk yes, AES is much faster, and has had much more academic interest.
@Tinned_Tuna yes, familiar(-ish), just forgot what theyre called.
@AviD fair enough :-)
12:34
@Tinned_Tuna ahem, and governmental.
blowfish always makes me giggle
one day I'll grow up
@AviD ah yes, the ol' govt. Didn't they announce it can be used for all but the very highest classification of data or something recently?
makes me think of blofeld.
or am I mis-remembering?
feels like I need a tshirt making saying "I blow fish"
great, Simon arrives immediately as I say that
12:35
aannd a wild @Simon appears
fml
you ninja'd me!
nope
refresh, I was first
@Flyk happier?
yes, I like being first
12:36
that was my intent not to impune your Internet speed
uhuh
next question
would you use OpenVPN client rather than Windows' build in client
pls all
also, what does this mean:
"For key exchange and authentication 4096-bit RSA keys are used."
@Flyk I was using openvpn for a while. pissed me off.
@Flyk they use a big key
12:37
It means that 4096-bit RSA keys are used for key exchange and authentication.
@RоryMcCune its not the size of the key that counts
@Simon -.-
its also what algorithm you use
Sounds speedy
I can read the words
I need explanation
12:38
@Flyk I thought rewording would help you out.
@Simon well there's the problem; you thought. Don't do that.
pls
@Flyk uhuh :-)
@RoryAlsop well then there's less apprehension about whether you'll finish at all
makes the entire experience more enjoyable
That should not be in question
12:40
WELL C'EST LA VIE
#somemenarenice
can we get back on topic now
@RoryAlsop #notallmen
@Flyk doughnuts?
should I just change VPN providers out of good practice
12:40
ehh? ehh @Flyk? see, it works both ways!
JE N'Y PEUX RIEN C'EST ELLE QUI M'A CHOISIE
@Flyk AES is a symmetric algorithm, that is, if you have the key you can both encrypt and decrypt by doing the same thing in different directions (and sometimes not even that, in the case of DES). For public key, you need to agree a secret over a public channel where you trust absolutely nobody who can listen in.
C'EST LA VIE
For that we have key agreement protocols, like diffie hellman
@AviD unique usage
12:41
basically we exchange bits of secrets and do some magic operations each end, and voila we end up with the same key
but nobody in between did
however, there's still a Q - how do you know you were talking to me in the first place?
oh god what have I done
For that, I "sign" my message using a well known signature scheme, like say RSA
you know that I understand each of the individual words you use but not the sentences you form with them, right?
chat slower
what is "diffie hellman"
@Flyk the most common way of agreeing a secret over an untrusted channel, using maths
12:43
@Flyk that is the name of a key-agreement protocol, named for Diffe and Hellman
^^that's diffie
I give you some numbers, you give me some numbers, we do some maths and end up with the same thing, but nobody else can work out what we came up with
ok
I understand
obviously, there are special restrictions on the numbers we can choose to make that possible, I can't just pick any old numbers
so
once we've got our secret, we can both use this as a key
to a symmetric algorithm (one that only needs one key)
if anyone else knew this key - bang goes security
but we agreed it, and nobody else can know it
HOWEVER
It is possible when we were exchanging secret numbers that I was replaced by an imposter
Let's say by @RoryAlsop
@RoryAlsop invented math, though
he's basically unbeatable
final boss alsop
12:45
If that happens you might think, hey, I'm talking with this guy with Nine fingers when actually you're talking with @Rory
To make sure this doesn't happen you need more maths
@Nוnɛfוngɛrϛ even if you're talking with @Nוnɛfוngɛrϛ, you're always talking with @Rory.
It's like Omnipresent Man.
@Rory = NSA?
I have a public/private key based on RSA (more maths) and I can use the public one to encrypt stuff, and the private one to decrypt stuff, or sign (a special case, but I digress)
but wasn't rsa4096 broken already?
@Flyk Nope, rsa 768 has been factored
1024 is probably doable by the NSA if they're in the mood, or you say things like bomb in public internet chat
anyway
I started Windows update before I left work yesterday. This morning: "Preparing to install".
Ah, there's breaking and then there's breaking breaking
@Flyk That's a side-channel attack
Hey @AviD, Windows is a fucking piece of shit.
the maths still holds, but the implementation had some problems
12:47
One break is I give you "n" and you work out my factors
that you can't do
what's a side channel attack
@Simon I'm sorry, I spent all afternoon on it.
@Flyk basically, when you do some computation, you require things like power, time, etc.
@Flyk any way of measuring what's going on so you can infer so info about the key
@Flyk It's a sound played on a side channel with a lot of attack.
12:48
so, looking over somebody's shoulder is a side channel attack?
From your patterns of usage of these resources, an attacker may be able to infer secret things, like your keys
@Simon you're thinking of wind chimes.
@AviD Good, I'm expecting a fix before this afternoon.
@Flyk not quite, those are neither attacks on the implementation or the protocol; they're always present due to "stoopid" users.
@Simon I'm placing you on ignore for the duration of this highly intelligent conversation, my brain refuses to context switch between knowledge absorption mode and silly fuckabout mode
12:49
@Flyk welcome to the club
@Flyk Look at you trying to understand security.
@Simon I will be glad to reimburse you for the day of lost work. I am sending a stick of gum in the mail now.
5
Q: Origin of side-channel attacks

TheGoodUser-AmirWhat is the origin of side-channel attacks? Is it related to the algorithm or to the implementation? In other words, is the software responsible for the attack or the hardware? Is there a cryptographic algorithm that is resistant to side-channel attacks on system A, but is not on System B? Can ...

@Tinned_Tuna right so it would be more social engineering then?
@AviD Not even a donut?
12:50
@Simon tad greedy, arencha
@Tinned_Tuna Go back to the sea.
@Simon is tuna made out of chicken?
@Flyk Or just plain bad operational security. In the words of the mighty Thomas Pornin, cryptography concentrates security into the key. If you input the key in the presence of an adversary...
@Flyk in that class, yea. Basically, an "intelligent" user would shield their keyboard, tell the other person to go away, etc. Intimidating people, using "polite" society tricks, etc. to allow you to monitor people is probably in the class of social engineering.
@AviD Only on Wednesdays.
12:51
@Nוnɛfוngɛrϛ or count the fingers
@Flyk but drawing the line is not necessarily easy.
sorry, before your time.
Right, I'm going to go put food in mah face. Enjoy.
So anyway, the signing bit lets you decide who you're talking to is who they say they are. The exchanging bit lets you agree a key.
Then you can do crypto.
@Tinned_Tuna aim for the big hole on the bottom.
12:52
I am putting delicious food in my face
@Tinned_Tuna right so a side channel attack would be like (and I don't remember any of the specifics, it's just stuff I've read) the thing where if you chill RAM it retains data, or measuring the voltage of the CPU, etc?
... is what yo mom said to me, last night.
I need sugar
brb, ice cream
@AviD Please, tell me that's not a reality show?
You could use public key crypto for the whole session, but it's much less efficient than than symmetric stuff
12:53
@Simon from a reality show, yes. One of the first.... and that bit was damn famous.
Also, agreeing a key per session is nice because once the session is over, the key disappears and there's no key to steal
@AviD SHE GENUINELY ASKED THAT QUESTION?
this is called "ephemeral" cryptography, or forward secrecy.
isnt she also the one that started the antivax?
@Simon oh yes
she is genuinely that stupid
Wow.
12:57
so @flyk to answer your question, the 4096 bit key is the size of the modulus n used in the rsa signing bit of your openvpn connection, meaning it probably can't be directly factored for a while. Then you're using aes-256 for symmetric crypto which is fine also, there's no known attacks that significantly reduce the complexity from "try every possible combination there is and hope"
@Simon I love the husband's look. Not that he is known to be smart, but that is some special level there.
@AviD that does make you wonder what she's doing on telly tho and does the majority really enjoy seeing others more stupid than them and still doing much better than they do, because I find that rather scary
@TildalWave oh absolutely
also one of the reasons I don't watch TV
@Nוnɛfוngɛrϛ so VPN inside a VPN is probably too much?

« first day (1525 days earlier)      last day (3653 days later) »