« first day (1826 days earlier)      last day (3352 days later) » 

17:00
Either way
-.-
YOU WERE STEALING WOW GEAR
WHICH MEANS YOU'RE GREEDY
GREEDY PEOPLE ALWAYS ROLL NEED
@Simon I only used it on myself
@kalina pls hulp
Because two factor seemed silly against Keyloggers and I wanted to prove it.
@MarkBuffalo yeah, but it raises the resources and sophistication
you can't just run hashcat and get accounts
now you need to compromise the target machine AND scoop another piece of information
17:02
@MarkBuffalo In short, threat moves from MitM to MitB.
Indeed, but I like breaking things just because
@MarkBuffalo ees fun
@Ohnana Yeah. Anything can be broken, circumvented, etc. That's my thinking
Hello . Does anybody know if the NaCl Crypto Library uses Montgomery-reduction for the ECC primitives ? :-) Thx in advance
@ThomasPornin someone's using scary crypto words in chat, can you help?
^^ he knows everything about crypto
17:17
Oh yes . I already read a few comments from him :D
@Ohnana We better call the FBI. Only terrorists use encryption
@SebTu fun fact, he's one of the few people on SE that have been cleared to use a sockpuppet. tom leek is his other account
@Ohnana Why?
@MarkBuffalo he only votes on the pornin account, and that much rep is too much for one account :P
bascially it's a harmless sock puppet
He repcaps too quickly for one account so he needs a second one.
17:21
didn't bears are that clever
*know
All the sudden my kidney is killing me. I'm feeling weird too
@Ohnana lol that article is like 5 years old :op
PC world are lame for not putting a date on it, but try looking at the comments
they start at 1285 days ago...
@RоryMcCune :O
@Ohnana Motherfucker I just showed my boss and we botched at LinkedIn together
@MarkBuffalo they deserve it
17:36
Lol
Get to work from home tonight though, so yay
@RоryMcCune those comments are cancer
@Ohnana heh, it's amusingly trending on twitter, loads of people RT'ing that linkedin has a hack.... I suggested to the pcworld twitter that putting dates on their articles might be a plan :)
@RоryMcCune apparently they updated the article?
timestamps motherpony, DO YOU USE THEM
@MarkBuffalo And my buddy posted it on his work Slack.
Whoopsies
@Simon hue
17:43
@Ohnana sure in 2012
@RоryMcCune /table flip
look at the content says linkedin hasn't responded and has 161 million users
I think they'd respond in 3 years and they have over 300 million users
haha
I tweeted a couple of people who're speading it, but I doubt it'll slow it down
would make an interesting study in how mis-information spreads on the Internet
The date of the article should be the first thing at the top
17:47
@RоryMcCune see also "we did it reddit"
17:57
I'm dying
wot
wot
@Simon what do you want
@kalina They said my WoW joke made no sense.
18:05
@SebTu I have not looked at the code and the documentation is a bit scarce on the internals, but chances are that Montgomery reduction is NOT used.
Montgomery reduction helps with doing computations modulo an arbitrary prime p.
and curve25519 doesn't use an arbitrary prime p ?
NaCl uses the special "curve25519" where everything is done modulo p = 2^255 - 19
@Simon Lol, now @Ohnana has started a riot
@MarkBuffalo wat
@Simon I didn't see a joke, were you trying to be funny?
18:07
pls
That prime is chosen specifically so that modular reductions are efficient without needing to use Montgomery reduction.
@Simon srs -.-
For ECC with prime p of no special form are their three common algorithms for modular reduction ? Classic / Barret / Montgomery ?
i didn't start the fire!
@SebTu Normally people use Montgomery multiplication and Montgomery reduction, because it is simple enough to implement, and it is constant-time.
However, not many implementations support arbitrary curves.
18:09
Montgomery multiplication for the ecc group operation ? e.g. montgomery ladder ?
because im wondering if it is a common way to use a native multiplication algorithm like the schoolbook- or comba-method for big integer multiplication e.g. $ c = a \times b $ passing the result $c$ to one of the modular reduction algorithms mentioned above ?
Most implementations support only a few specific curves, typically the P-256 and P-384 curves from NIST, that use special-form primes.
@ThomasPornin Real men support curves
Using an integer multiplication (e.g. with Karatsuba) and then doing a reduction is not a usual thing to do with arbitrary primes, because this makes the reduction more expensive.
@MarkBuffalo good point ! :D
(Because you start with an integer which is twice bigger)
You still do that if you have a special-form prime that makes reduction easier.
As for the point multiplication algorithms, there are various methods.
18:12
So usually you interleave multiplication and reduction ?
@SebTu That's the point of Montgomery multiplication.
Multiply by one word, then reduce size by one word; repeat.
And this applies for ecc ?
Not only rsa
@SebTu Montgomery multiplication applies to all computations modulo a prime, and ECC computations use a lot of these.
Yeah
ECC tends to use smaller primes (say 256 bits instead of 1024 or 2048) but the principles remain.
This book will give you a lot of information on these subjects: cacr.uwaterloo.ca/ecc
18:15
I always used an not interleaved approache for my implementations
Yes . I habe it in front of me ! :D
*have
That book has a lot more information but it is also a bit harder to read: amazon.com/…
I used Comba-Method for multiplication and then montgomery-reduction for the result and obtained good results for my µC
using the MIRACL Library
I hate how clear communication can be perceived as offensive. I used a choice selection of one expletive phrase in my e-mail to the third party benefits provider, to make sure that the severity of the issue was clearly communicated (ie, you're making the last 4 of my SSN public which also probably puts you in violation of HIPAA since it deals with HSA funds). But, the fact is, it worked properly. It got the message forwarded to the CEO of the company and made sure he was aware of the issue
18:18
Even better results using the µNaCl Library
@SebTu For the point multiplication algorithm (this is independent of what you use for the base field), the algorithm of choice will depend on the curve type.
@AJHenderson I would've just filled every few words with expletives.
so I just let him know "please tell your employees that were upset by it that I'm sorry, but it was necessary to make sure the severity was appreciated."
For binary curves (curves in GF(2^m)), the Montgomery ladder is hard to beat in efficiency and ease of implementation.
@AJHenderson ooooh yeah that's something to watch out for
18:19
@AJHenderson His response?
@MarkBuffalo :O
@ThomasPornin Hm I skipped binary curves due to security concerns
For curves in a prime field, some curves allow for an efficient ladder; others will usually prefer a window-based optimization on double-and-add.
@MarkBuffalo nah, explitives have a purpose, but it's only really effective if you use it sparingly in formal communication. It is valuable for conveying a specific feeling, but it's something to be cautious with
@AJHenderson I prefer to come across as a rabid animal.
18:20
@SebTu There is no real, rationally justified, security concern about normal binary curves.
But third-party support is less common.
@MarkBuffalo basically offered assurances that he'd make sure it was looked in to by their vendor and to have some security people take a look
@AJHenderson Kidding, btw. So what happened? Did the CEO try to make you feel bad?
"some security people"
That worries me.
@MarkBuffalo that is a paraphrase, he sounded serious
this is a small local company
I'm sure they hadn't even thought about the issues
18:21
@kalina this is @Simon I kind of assume he's always trying to be funny (absent explicit srsface -.- )
@RоryMcCune I assume that too, but even when he does "-.-"
schroooooderrrr the diamond man
@MarkBuffalo oh no, srsface is srsbusiness, should not be dismissed -.-
@RоryMcCune Oh no. He's done that to me before
@Ohnana howdy
18:23
but his response was good, actually better than I expected. Basically thanked me for reporting it, pointed out the impact it had on his employees (make me feel a little bad) and sandwiched it with more assurances. I responded similarly that I thanked him for looking in to it and I'm sure they'll get it worked out (they really do a great job in their domain, security just isn't their domain), that I stand by my word choice, but that certainly I'm sorry for the collateral damage along the way...
and a few more security related thoughts I had realized since
@AJHenderson yeah it's one thing to use expletives, but being on the recieving end can be kind of harsh
@AJHenderson That's good.
the best kinds of expletives are ones that aren't directed at people or projects
@Ohnana well and in fairness, I always avoid directed explitives
I dropped a "the only response I can have to this leak of data is WTF" or something like that
it was very specifically as my reaction to the situation
@AJHenderson oh jeez
that's weak
18:25
I spelled it out
@AJHenderson I had this problem where an employee kept trying to harass me repeatedly. Instead of saying "f-you," I just said, "f-it," and walked out.
@AJHenderson Then she told everyone I said "f-you," and treated me even worse next time :-D
So I try not to use expletives when talking to people. Gives them a reason to be upset
@MarkBuffalo wooooow what a meanie-pants
idk how people decide that's an acceptable way to treat people
@Ohnana Health care worker with a chip on her shoulder. She got fired for doing that to other people too.
but I never complained to her boss about her
yes, it does, sadly, sometimes that's a neccessary cost of making sure the severity of feeling is communicated accurately too though, especially since a non-security person is not going to understand the scale of issue of last four ssn in a username
without the explitive drawing attention to it, it is just "oh, someone's frustrated", not "WTF the sky is falling, get this fixed now or you're liable to be sued for HIPAA violations by someone because this is totally NOT OK"
i've found a good substitute for WTF is quoting a legal statute
that's sure to get their hearts a'quiverin
18:29
yeah, the hipaa angle hadn't actually occured to me at the time
so I didn't know for sure if there was a legal argument outside of just a general civil suit
and honestly, I would rather swear a bit than threaten legal action. It's a mistake and people are too law suit happy
Everyone ready for some Star Wars spoilers?
cause threats are far worse than non-directed expletives in my mind, but maybe that's just me
Luke....
....is Darth Vader's son!
@MarkBuffalo Nnnnoooooooooooooo!!!!!!!!!
@AJHenderson Gotta be careful in the crybaby sissypants age. Almost anything can be perceived as a threat these days.
18:32
what about that bit at the beginning with the thief?
@kalina or the dinosaurs
and then Han finds a co pilot
and then the resistance forces land
I hear nature finds a way
and the rebel forces attack
@kalina the movie's not even out yet how are you spoiling
18:33
what do you mean how?
how are you spoiling something that's not out yet
because I already know what happens?
@ThomasPornin may you find some time to anwser one remaining question . would be nice ! :)

1) Using Montgomery Multiplication is there any restriction for the multiplication algorithm for the field arithmetic ( e.g. Comba- / Schoolbook- / Karatsuba)
I don't understand the question
simple, she threatend to light JJ Abrams on fire unless he spilled the beans
18:34
@kalina movie is not out
how u spoil
well "how" is I press these buttons on my keyboard and words appear and then I press enter
how did you see the movie
did you set jj abrams on fire?
oh I can't share that information
please wait to do it after the new star trek movie is finished
wait, what?
18:35
@Ohnana There are, ehrm, ways, to obtain the script/movie right now.
anyway they sneak into the deathstar
and then stuff happens
the end
@MarkBuffalo awwww i hate leaks
it ruins things
i don't like star wars, but still
in all honesty if you expected a star wars movie without all of that stuff ^ then you've not been watching star wars movies before anyway
@Ohnana I didn't realize you were so evil
;)
I've never even seen a star wars movie
well I watched like 15 minutes
18:36
not my fault the movies suck
and I've seen other bits walking through the room at various times
and I played one of the games for a few weeks
@kalina I read 'bits' incorrectly for a second there.
but hey, now that George Lucas can't lucas them, who knows what could happen
because light saber duels...
pls they're getting disney'd
disney'd is still an improvement over a good old fashion lucasing
lucasing, now that should be an explitive
18:38
anyone posts star wars spoilers and they're on a one way trip to kicks-ville courtesy of my mouse :)
11
@RоryMcCune does that mean I should post spoilers just to see what happens ;)
@RоryMcCune well you should kick me then
since all of that stuff happens
@SebTu Montgomery multiplication is, inherently, the combination of "schoolbook multiplication" interleaved with Montgomery reduction.
in that order
bye chat
@AJHenderson thats' just straight up mod-abuse! given I can't kick the hallowed blues
18:39
@RоryMcCune NO BALLS
"hallowed"
@kalina I didn't know you'd actually spoilered yet
If you use Karatsuba + Montgomery reduction then you do not use Montgomery multiplication; you use Karatsuba + Montgomery reduction.
@Ohnana no I literally can't
oh well if you want an actual spoiler
18:39
@Ohnana the option does not appear in my menu on mods
SHUT UP NO
STOP IT
@RоryMcCune oh lol
@RоryMcCune I guess you'll just have to get kalina to get some good old fashion vigilante burnings
@kalina A spoiler to Montgomery multiplication??? I think we all know how that ends
R2D2 ... at the ... :(
18:40
@AJHenderson "burn the spoiler"
is there a moderator so powerful that they can't kick themselves
@ThomasPornin and it's the implementer decision to do the multiplication and reduction separated or not ?
@Ohnana yes, but they also can't unban themselves
@Ohnana I dunno, perhaps a mod can kick a mod, ...
as Undo discovered
18:41
@AJHenderson wait what
@ThomasPornin There is so much stuff to consider with ECC !
Well this was fun.

1. Found vulnerability, sent to sysadmins.
2. Sysadmins apply configuration changes to fix. App works, but changes fail to fix the vulnerability.
3. Repeat step 2 ad extremis - eventually the configuration was so that there should be no possible way the vulnerability conditions would exist. But it did.
4. Sysadmins tried an alternate solution. Solution fixed the vulnerability but broke some app features.
5. There's no obvious reason the new solution should be breaking the app, so I jumped in to review the docs. And found a step the Sysadmins were missing in the *original
@Ohnana in chat, moderator permissions are tied to which account you have set as your primary association
so when I am mod-status, none of the ban or unban options are available for my user
@AJHenderson so he couldn't undo himself
but if someone changed my primary account to one I'm not a mod on, I become bannable
18:43
@Iszi /facepalm
so someone did that to undo and undo switched back to his mod account, but couldn't unban himself
@SebTu It is up to the implementer to decide on implementation strategies, as long as it keeps things safe.
@Ohnana More like #facepalm.
he could talk despite being banned since it didn't actually apply to him
but he showed as banned
In particular, protection against side-channel leakage requires a lot of care and deep understanding.
18:43
it was pretty hillarious
@AJHenderson lawl who banned him?
You should not implement crypto until you can actually explain why your implementation strategy is safe.
I forget, it was actually started as an experiment
but then it went on for like two weeks because it was just too funny
@ThomasPornin I read that using special modular reduction e.g. with pseudo mersenne primes is not fully protected against side channel attack
@ThomasPornin uh . This will still take a long time thought ! ;-)
anyway thanks for the help :-)
@SebTu that fact that you say that gives me hope that you'll have a great product!
18:46
Haha
well ...
:D
@SebTu It really depends on how you do it, and the specifics of the hardware platform.
@Iszi /palmface
For timing attacks, Montgomery reduction is safe as long as the hardware multiplication opcode is constant-time.
@SebTu yeah, but the alternative is that someone who does looks at your code down the road and wonders why you are using the same key-stream for every encryption operation... oh wait, that was the last major incident I saw with incorrect application of cryptographic principals
"Classic" reduction is much harder to make constant-time because you have to estimate the next quotient word, and, depending on data, fix it afterwards. That fix step is prone to leak data through timing.
18:49
How to troll nerds: 1) Go see #StarWars 2) Open support tickets 3) Include film spoilers in the ticket
Now that's evil
@RоryMcCune EVIL
@RоryMcCune Ouuuuuuuch
4) get rekt by your it department in 5 different ways
@AJHenderson Sounds like they deserve an award:
hue
18:51
@AJHenderson I tested the implemented code and it works just fine . I used some libraries and did myself some modifications. ECC is really a lot of stuff to consider and partially I know what I am doing. The rest manages the library ;-)
@ThomasPornin Are side channel attacks your research area ?
@SebTu Technically no, but I did look such things up recently.
And more generally, I take care to try to understand how things work, down to the fundamental interactions (most of what happens in computers comes from electromagnetism).
Think there are side channels attacks measuring electromagnetism
You probably spend a life time on this to understand how things work
@SebTu That's not a problem, since being alive appears to come with, indeed, a life time.
@ThomasPornin true ! :D

I leave . Nice evening / day guys :-)
and thanks again for taking time to awnser
*answer
19:08
I'm looking at this new question:
0
Q: ethical hacking question

JohnDoeLets say for instance... X has hacked Y X does not know Y but thought Y looked and sounded suspicious over social media X infiltrated their emails and discovered content that local authorities would be highly interested in. The only issue is how would X inform the authorities about Y being a ...

I'm wondering if this is on-topic? It is a valid question, and I think there can be some good answers to this. But it also feels off-topic - and perhaps too opinion-based.
@S.L.Barth x and y need to take a computer break
i think phillip's answer hit it on the head
you cannot, because legal stuff
although legal stuff is not good
Hm... it's also possible to report crime anonymously, in some jurisdictions. It would not be evidence, but it could lead to the investigation that gets a criminal convicted.
@S.L.Barth i'm starting a close vote, let's see what other people think of it
I'm undecided, so not going to flag. Let wiser people than me decide.
Thanks for thinking along!
@S.L.Barth exactly my thought
20:14
@JourneymanGeek the actual box, the case
@Simon why, wadidimiss?
@AviD scroll up like an inch
@RоryMcCune pls, insiders come to that conclusion
@Ohnana you dont know what resolution I have
or font size, for that matter
@AviD relative inch
@Ohnana @Simon is gonna be offended by that
@Iszi lol offensive use of humor there. Or humorous use of offensiveness.
20:41
@AviD Did you read it?
@Simon read what?
da drama
You did read the first part, since you were involved in it too.
@Simon was I?
Ya the "M" word thing.
@Simon DAMMIT HARDISSON!
20:46
wat
Anyway, I guess I should stop using "bad" words since it makes some people go all bananas.
Never go full bananas.
@Simon hey man, my monkey finds that offensive.
Would pineapple be less offensive?
I assume that could be offensive to apples?
offensive to @Adi.
@Simon "Pineapple" is the usual nickname for fragmentation grenades, which are technically known as defensive grenades.
@ThomasPornin as opposed to the grenades you fight with?
20:51
(Offensive grenades are pure explosive, so that you can throw them while marching on the enemy, while defensive grenades include fragments meant to be used while retreating, i.e. in a "defensive" way.)
oh i see
pull the pin and run at it = offensive, pull the pin and run = defensive
This is an offensive grenade.
This one is defensive:
@Ohnana pull the pin and don't run = suicidal?
@ThomasPornin Interesting, since those are usually the ones that you see in movies and they sure are using them as offensive ones.
As you see, only the defensive grenade features the characteristic "pineapple" pattern.
20:54
@Gilles Only if you also don't throw it.
Or have a shit arm.
@ThomasPornin thats not universal.

« first day (1826 days earlier)      last day (3352 days later) »