A message from Comodo Hacker | Hacker News Reader

A message from Comodo Hacker

pastebin.com

56 points__hudson__37 comments

I crossed paths with the guys behind globaltrust.it years ago. I ended up auditing a code project they had written for a co. in the financial services industry that required good security, audit trails, client authentication etc.

After the third remote exploit that I found, my recommendation was that they throw it out and start again, it was a huge jumble of PHP. This hurt their feelings, and a 6 month long argument ensued where they defended their competency.

When I heard 'Italian certificate provider' last week, I thought it could be them, because they went on to launch a certificate project. I am more surprised that Comodo didn't do any due diligence on their resellers. All they had to do was to email my client to ask how their project went, and they would have found out about the clusterfuck.

> At first I decided to hack RSA algorithm, I did too much investigation on SSL protocol, tried to find an algorithm for factoring integer, analyzed existing algorithms, for now I was not able to do so, at least not yet, but I know it's not impossible and I'll prove it

Huh. He kind of lost all credibility at that point. Breaking RSA isn't something you just decide to do. I'll wait for the day when he announces he's broken it.

I highly doubt that he will be able to do so either.

However, that does not reflect on whether or not he is the one that is behind the attack on Comodo, which has no real indications of being difficult.

Comodo claims that it must have been an organized, planned out attack because they knew which domains to get certificates for. That does not explain why 3 were generated for one domain and one for 'global trustee'. Nor does it take a genius to figure out a set of domains you would want certificates for depending on what you are planning to do (in this case, it seems like attacking large webmail providers).

Hasn't everyone woken up some day and said "hey, I'm gonna break RSA today"? I guess that only happens when you have the experience of 1000 hackers...

You'd think someone with the experience of 1,000 project managers would at least have said "this week." Just for some "uh oh" room...

Or the polynomials of 1,000 numeric field sieves.

pfft call me when it's over 9000.

LOL. Does this guy think he's really going to scare us with lines that could've come from a Michael Bay movie??

Aren't they just naively translated idioms which don't sound quite so hackneyed in Farsi? Dunno, just guessing.

The math in these algorithms is not the vulnerable part of the process. It is almost always the implementation or the implementers that are exploited.

I came here to post exactly that.

After reading him state how astonishing his skills are and how he'll tackle the integer factorization problem, I thought to myself, "he must be a 20 years old university student." Then he states his age.

Yes, the posturing did not lend credibility to his statements. Besides that, it seems vaguely plausible.

> I'll wait for the day when he announces he's broken it.

And so will a massive part of the Computer Science/Mathematics/Physics community. Answering the P vs NP question is kind of a big deal. :D

No. Integer factorization is not NP-hard (so not NP-complete). (This isn't proven, but it's generally thought to be the case.) So, while doing a polynomial-time integer factorization would be hugely significant (and make all asymmetric encryption in the world useless), it would not prove P=NP.

> So, while doing a polynomial-time integer factorization would be hugely significant (and make all asymmetric encryption in the world useless)

This is wrong in two ways.

First, a polynomial-time algorithm could still be too slow to be practical, either because the degree of the polynomial were high or because the constant factor or asymptotically disappearing overhead were high.

Second, discrete-logarithm-based cryptography does not depend on the difficulty of integer factorization. That includes Diffie-Hellman, ElGamal, DSA, SRP, and elliptic-curve methods.

You're right that integer factorization is not known to be NP-hard, and so a polynomial-time integer factorization algorithm wouldn't show P=NP.

You're right, that statement was rather too broad. Thank you for the correction.

I don't want to turn this into a complexity theory discussion thread but isn't it an NP problem? And does proving that it can be solved in polynomial time mean P == NP (and vice versa)?

Am I the only one to expect the guy who compromised a CA and generated a number of very high-value certs to use one of those certs to sign his message?

So it takes the "skill of 1000 programmers" to write a program to send POST requests to an HTTP API? He must have been writing his client in Java...

I wonder whether that hyperbole of '1000 ...' is due to mistranslating some Iranian idiom. Not that it would make the assertion any better.

I come off with the feeling it is actually composed by an Iranian (possibly a team but could be an individual) who is clearly motivated to make "politically correct" (from the POV of the Iranian govt.) speeches. He goes to great lengths to praise his government, ambassador and president while denouncing all dissidents, separatists, Israel and the US. I cannot help feeling he/they have some connection with the government. Possibly the whole teenage cyberpunk rhetoric was deliberate (see repeated "hard for you easy for me" and "I was so fast" and the absurd "I will factor large integers") and little more than a poorly executed smoke screen to divert attention.

If that is indeed the guy, he certainly sounds like a massive douchebag.

About 1000 times a douchebag.

welcome to lone wolf penetrations

Actually he sounds like a teenager troll, and blatant at that.

Update with part of the decompiled TrustDLL code:

http://pastebin.com/DBDqm6Km

My favorite line: "easy for me, so hard for others"

He sounds like a douchebag because he thinks, he is awesome after hacking a CA.

The most stuff is just lame, but I still appreciate the way how he infiltrated the system.

I think I kind of know, but for the uninitiated, can someone tell us what the potential practical consequences of this hack could have been?

If you combined these evil SSL certs with the right BGP hijack, you could read a lot of people's email and such. And since reading mail lets you reset passwords on everything else, he could have basically owned millions of people.

BGP, or DNS, or your DSL modem or cable modem, or an open access point, or the router in an internet cafe, or...

I thought the most interesting part about this was that he claimed to be a 21yo Iranian who took sole credit, while having a problem with the lack of controversy surrounding the Stuxnet US/Israel project.

Oooh, my favorite part is on line 136, when he claims "RSA 2048 was not able to resist in front of me". That's a pretty, um, "interesting" characterization of the level of sophistication of his attack.

That's some serious Miyamoto Musashi rhetoric there. The guy's either a grade A rationalist or a massive douchebag. I put odds at 5/95.

Can you elaborate on the "grade A rationalist" remark? Does attempting to compensate for one's cognitive biases generally blind one to how one sounds to others?

Well, it's finally arrived the time to rethink the whole CA system.

I LOVE the fact that I live in a world that I get to read some hackers manifesto online:

My Rules as I rule to internet, you should know it already...

Although, I do have to agree with his points about Echelon.

EDIT: His command of the English language is irrelevant -- I stand by my comment above, which is the fact that we are even reading stuff like this makes the 16-year-old-cyberpunk-playing self from the '80s quite happy.

Reply on news.ycombinator.com