Nadim Kobeissi

PhD student in applied cryptography. Rides a blue Vespa. Good person.

[nadimkobeissi]

How WhatsApp Needs to Improve Its Encryption

Addendum (10 Jun 2015): WhatsApp cooperated with US intelligence agencies to decrypt user chats, according to Bloomberg.

I recently wrote a post touching upon secure messaging projects falling short from due diligence when it comes to their security guarantees. The most-discussed aspects of that post were my claims that WhatsApp guarantees zero end-to-end confidentiality of your messages. I stand by this. To answer questions I’ve received, I’m going to explain in detail why this is the case.

How is WhatsApp claiming that my chats are encrypted?

WhatsApp and company recently announced that they had implemented an excellent encryption protocol that includes a truly innovative messaging and key scheduling mechanism. According to WhatsApp and Whisper Systems (who worked on the encryption techniques), they actually did implement some pretty nice encryption, and it’s likely that this encryption is working in your WhatsApp client right now if you’re using it on an Android device.

So why isn’t WhatsApp’s encryption effective?

Even though encryption is coded into WhatsApp, it is possible for any WhatsApp employee, for your ISP or other agencies to circumvent and even completely disable its security, at any time, and without you knowing.

The details of why WhatsApp is vulnerable are painfully clear to anyone who’s seasoned in computer security. But the average WhatsApp user has no understanding of the attacks that can render even well-implemented encryption completely useless. Here’s my attempt to explain these simple and important ideas.

When you start an encrypted conversation with with a friend on WhatsApp, you need to communicate to them a special public key that is used to establish secure communications. Your friend’s WhatsApp client will rely on this public key to know who it’s encrypting the information to, so it has to be authentic.

Now, since you don’t have a direct connection to your buddy’s phone, this public key exchange, like regular WhatsApp messages, has to be sent through WhatsApp’s servers, who then deliver it to your friend’s phone.

What’s preventing WhatsApp from forwarding fake public keys to your buddy instead of your actual ones? You guessed it: nothing. In computer security, this is called a Man in the Middle attack.

Anyone who replaces your public key in transit obtains immediate decryption. Zero security, end of story. And nothing is stopping WhatsApp from replacing your public key on demand, to intercept “encrypted” communications at any time.

It gets worse. Due to its excellent multiplatform compatibility, WhatsApp’s servers need to manage connections between many different WhatsApp clients. Some clients are running on advanced phones (such as Android or Windows Phone) while others are running on old Symbian or BlackBerry phones, which can be overwhelmingly popular in India and other parts of Asia. Managing connections between highly capable phones that can handle encryption and phones that can’t is a process that requires the WhatsApp server to get involved.

Because of this, your phone needs to check with WhatsApp’s servers first before being allowed to begin a secure chat, and WhatsApp servers can reject this permission at any time, without you ever being notified. The problem here is that WhatsApp won’t even provide an indicator of whether a conversation is encrypted or not: the server could disable your encryption permit at any time, and you would still completely be under the impression that you’re having an encrypted conversation. This isn’t security. End-to-end security is not selective, it’s always enabled, and enabled by default.

Be constructive. How can WhatsApp fix its encryption?

Man in the middle attacks are not new, and most secure messaging apps have a mechanism to protect against them. They allow you to verify the integrity of encryption keys yourself: for example, by scanning a QR code generated by your friend’s phone. WhatsApp can implement this today; it’s been implemented by countless applications before it.

It’s also clearly possible to implement reliable indicators that inform users when their chats are encrypted and when they aren’t. Calling WhatsApp conversations “secure” in the media and then depending on WhatsApp’s servers not disabling your encryption at any time is misleading to the point of being dangerous.

It wouldn’t be accurate to build a motorbike with just a front wheel and call it a motorbike. You need both wheels. It’s similarly inaccurate to call WhatsApp’s encryption end-to-end secure as it stands today. Until WhatsApp offers means to resist against man in the middle attacks, and ways to verify the encryption status of conversations, it is simply, and undeniably, not secure.

Disclaimer

I’d like to point out that I am an avid, daily WhatsApp user and that I love its reliability, ease of use and multi-device availability. I just don’t use it under the pretense that I’m benefitting from any level of privacy. I mostly use it to send cat photos. Thanks WhatsApp for your bandwidth for my cat photos.

I’d like to disclose that I also work on other encryption software (which has had its fair share of flaws over the years). I’m stating this because the last time I published a critique of encryption software, an upstanding gentleman accused me of doing so only to “sling mud at competitors”. Writing to point out security flaws that affect real users isn’t slinging mud. It’s called peer review. It’s what you do when you’re a graduate student studying cryptography.

Finally, my description of man in the middle attacks is overly simplified, I’m aware. No mail, please.

[nadimkobeissi]

On the Secure Messaging Community

This afternoon, WIRED published an article praising Telegram, a mobile messaging app, for its privacy features and "hardcore encryption". Telegram is an open source messenger that provides some encryption features -- but it has come under legitimate scrutiny due to the exotic and unstudied nature in which it applies its encryption. Here is how JHU cryptography professor Matthew Green describes Telegram's encryption:

@AlyssaRowan @bascule It's like someone who had never seen cake but heard it described tried to bake one. With thumbtacks and iron filings.

— Matthew Green (@matthew_d_green) March 29, 2015

To be sure, Telegram's encryption is, also by my own standards, highly unusual. Poring through it, I get the impression that it was designed by someone with a strong mathematics background, but with little prior knowledge in applied cryptographic protocols: they know the language of designing protocols, but not the applications or the history of protocol design. Nevertheless, Telegram has continued to tout its "hardcore encryption" without directly taking into account expert feedback.

On the other side of the playing field, Whisper Systems, a notable rival of Telegram that also makes messaging apps, had some developers react angrily to the news:

Telegram's claims to be a safer alternative to WhatsApp is getting on my nerves.http://t.co/py468NJW0e

— Frederic Jacobs (@FredericJacobs) March 29, 2015

Indeed, a claim that has often been repeated by Whisper Systems is that WhatsApp enjoys end-to-end encryption. This is due to WhatsApp hiring Whisper Systems to help implement encryption for their Android clients.

So, having seen Telegram tout its unstudied encryption everywhere for the past year, and having seen Whisper Systems shout from the rooftops that their involvement in WhatsApp makes it end-to-end encrypted and secure, I need to clarify two very important things. Because some claims from both Telegram and Whisper Systems are detracting from the entire effort of secure messaging.

To Telegram: Get some academic peer review.

No, this does not mean that you should hold another one of your crypto cracking contests; they are devoid of any academic value. Yes, that does mean that you should formalize your protocol to the extent where you describe why in the world it includes factoring random 64-bit integers into two primes (yes, this actually happens). It is no wonder that actual cryptographers are so critical when this is the level of scientific legitimacy you've settled for.

Telegram's mistake here is that they have construed any call for academic review as an attack on their business reputation. I'm going to go ahead and say that I understand why they'd think so: First, cryptographers yelling on Twitter might seem discouraging. Second, there's Whisper Systems' repeated attempts to assert that the only possible change to Telegram's cryptography that will be accepted as legitimate is to incorporate Whisper Systems' own technology and federate into its network. Moxie, it's obvious that this is a business tactic, and it's equally obvious that while Axolotl is an excellent ratcheting protocol (in my opinion, the best there is right now), holding community opinion hostage until your competitor implements your design is very tacky. So, stop it.

There are other venues with which Telegram can justify its cryptographic legitimacy. It can sponsor a panel of independent cryptographers to study the protocol, review it, and propose real, actionable modifications that can then be implemented. If we end up with Axolotl or some OTR variant, great! If we end up with something equally good, that's fine too.

Whisper Systems: Stop calling WhatsApp secure.

Ever since WhatsApp hired Whisper Systems to design some of the encryption for its Android app, Whisper Systems developers have been claiming that WhatsApp includes end-to-end security, a claim which they repeat today amidst their flame war with Telegram's CEO. Even after ridiculing Telegram's crypto cracking contest, Whisper Systems' CEO retorted with one of his own for WhatsApp:

@durov How would you like me to send you my number? Then you can send me my WA chats to Russia at the end of the week.

— Moxie Marlinspike (@moxie) March 29, 2015

The real, hard truth is that even though WhatsApp for Android has some end-to-end encryption code built in, the effective added security of WhatsApp's encryption implementation amounts to zero. This is because:

WhatsApp can intercept the encrypted communications of any user at any time. WhatsApp does not offer any method for users to authenticate their encryption identities. This property, known as identity authentication is essential for an end-to-end encryption system to be actually secure. Without it, there is no real guarantee of security against any WhatsApp employee told to intercept the traffic of certain users. Whisper Systems claimed in 2014 to be working on "surfacing" this feature, but there hasn't been any progress.

WhatsApp can completely disable a user's encryption without them knowing. Since WhatsApp is forced to multiplex between multiple user platforms, servers must decide whether encryption is enabled, not clients. This leaves any and all encryption decisions entirely up to the server. To put it another way: individual WhatsApp clients do not currently have any guarantee of end-to-end security against WhatsApp servers being active attackers.

Until WhatsApp addresses these huge, gaping holes in its security, it is simply false for anyone to claim that WhatsApp offers end-to-end security. For Whisper Systems to do so simply because they were hired to implement said security is unscientific and, personally, just a huge let-down.

Get your heads together.

This block Telegram and Whisper Systems are walking through right now? I've been around it a hundred times thanks to Cryptocat.

The secure messaging community's behaviour has been gut-wrenching. Two of the most significant applications are putting science in the back-seat to instead focus on business politics built on false claims. I'm sick of this nonsense. I hope both Telegram and Whisper Systems will get their acts together.

[nadimkobeissi]

Power and Social Gravity

Nadim Kobeissi, Felix Stalder and Bill Budington discuss power and social gravity. David Charles reports from the Elevate Festival.

Someone wrote a blog post about one of the most interesting discussions I participated in last year.

"Nadim Kobeissi, creator of the encrypted chat app Cryptocat, criticises Shoshana’s argument from a different direction. “There is a kind of politics that I have grown to dislike very much, which is the politics of us versus them,” he says. “It’s very effective, when you are wanting to mobilise a political movement, to use absolute terms,” Nadim adds, taking issue with the strong language that Shoshana uses: “usurping the internet”, “they’re conquering the internet” and “they’re taking privacy rights for themselves”."

[nadimkobeissi]

No, the NSA Isn't Breaking AES

If you're experienced with cryptography, you probably cringed at the title of this post. But I've recently received the last straw in emails theorizing whether the NSA has breaks in AES.

This speculation was made worse by the recent Der Speigel articles suggesting cryptographic attacks against SSH, IPSec, and even AES. It's possible that the NSA has discovered a small repertoire of minor flaws in AES that decrease its keyspace.

But for crying out loud, the NSA hasn't "wide-open" broken AES. Breaking AES is radically different from finding breaks in algorithms concerned with public key agreement, such as RSA -- which is far more likely the place where actually frightening breaks could happen. Here's the beef:

Block ciphers are not designed by picking the most "unbreakable" methods and combining them into a cipher. Block ciphers are designed ‎by taking a collection of operations that are simple, efficient and fast for a CPU, and arranging them in a mosaic that takes in input and, using a key, transforms it into a pseudorandom string. The key is pretty much an initial seed for a pseudorandoness function more than anything, and that's why newer ciphers have focused on adding features (such as AEAD) instead of making ciphers "harder to break". The most likely source of weaknesses in block ciphers are likely to be not in the block cipher itself, but in its mode of operation (ECB spot-the-penguin, CBC bit-flipping, CTR nonce-reuse, to name a few). This is why block ciphers are now discouraged, and most new ciphers are moving to the stream cipher model.

Algorithms that are actually suspect to breaks are public key agreement protocols. That's because they're based on NP-hard trapdoor functions, such as the discrete logarithm problem. Efficient mathematical discoveries are thus far more likely there, and could lead to a real break. You can discover how to factor primes faster, but that area of investigation has nothing whatsoever to do with discovering how to bypass bitwise operations faster or how to better detect biases in pseudorandom generators.

[nadimkobeissi]

Thoughts on Apple's iOS8 Encryption and the FBI's Reaction

For my thoughts on the FBI's aggressive reaction to Apple encrypting all iPhone device data by default to be beyond the reach of even government court orders, look no further than this Toronto Star article, in which I was extensively quoted (along with Matt Blaze, cool!)

But encryption expert Nadim Kobeissi said law enforcement critics are exaggerating the potential impact on their work and will still have access to the same data.

“They’re making much ado about nothing because they’re scared of the general trend, where companies are starting to stand up for consumer privacy,” Kobeissi said on Friday.

Kobeissi, a Montreal-based programmer, designed open-source encrypted chat program Cryptocat and works with the World Wide Web Consortium — the organization that sets global Internet standards — on online cryptography.

“We shouldn’t be operating in a mentality where increasing consumer safety is a bad thing for law enforcement. This is the mentality with which the FBI is approaching this, and it’s not a democratic mentality,” he added.

Even with the new encryption standards, law enforcement will still be able to access most — and in many cases, all — information stored on cellphones by requesting it from wireless carriers or companies such as Apple.

Kobeissi noted that, as the recent iCloud hack of nude celebrity photos demonstrated, many iPhone users also allow their phones to automatically upload photos and other data to Apple’s backup cloud service, which police can still access.

Police can also still tap a suspect’s phone, track their calls and text messages or track their phone with GPS or cell tower triangulation. “Your own refusal to unlock your phone is now guaranteed, but that doesn’t mean they don’t have this immense set of tools,” Kobeissi said. Kobeissi added that the new encryption standards are among several introduced since media reports revealed large-scale surveillance of phone and Internet data by American intelligence agencies.

“The three-letter agencies know that it’s their responsibility that this has happened in the first place,” said Kobeissi, referring to organizations such as the FBI, the Central Intelligence Agency (CIA) and the National Security Administration (NSA). “If they weren’t so overly aggressive in using their surveillance tools and exploiting surveillance laws, people wouldn’t have been so panicked.

“A lot of people in the security community are saying, ‘Cry me a river, FBI — you are responsible for this.”

[nadimkobeissi]

Surface Pro 3 Sleep/Hibernate Problems and Solution

My current main computer is a Microsoft Surface Pro 3. It's an excellent machine for my personal needs.

However, I recently ran into an issue where the sleep function completely disappeared, and the computer would only hibernate once the lid is closed. Hibernation is a far slower and more expensive process whereupon RAM contents are saved to disk so that the computer can be completely powered off so as to use zero power during hibernation. Sleep is the more common and much, much faster standby mode in which the computer remains minimally powered on for quick session resumption once you reopen the lid.

After some research I finally discovered that the reason this occurred was due to installing Windows Phone development tools, which in turn enabled the Hyper-V client-native hypervisor, which is used for running virtual machines during development operations. It seems that computers equipped to run Hyper-V cannot, for some reason, access the sleep function.

In order to re-obtain sleep, you must disable the Hyper-V hypervisor. This is possible through the following steps:

Run the optionalfeatures command either in the Windows Run dialog or in a command-line.

Find and deselect the "Hyper-V" feature.

Restart your computer.

After restart, you should be able to select Sleep as a default action in your power management settings for once the computer lid is closed, etc.

[nadimkobeissi]

Letter to the Mayor of Montreal Regarding Fighter Jet Parade

Fighter jets flew over Montréal today without warning in order to promote a football game. They were extremely, extremely loud and frightened pretty much everyone I know. I am sending this letter via snail mail to the Mayor of Montreal and posting it here as an open letter as well:

Dear Mr. Coderre,

I write to you today in good faith as a proud citizen of Montreal and as one of your constituents.

I must express my deep disapproval and personal shock at Montréal's decision to deploy fighter jets over the city today as some sort of aesthetic show to support a local football game.

Montréal residents were given no warning of the fighter jets and the entire city seems to have been rendered half-deaf by how loud they were. I've heard from a handful of people panicking and wondering about the fighter jets.

But more importantly, fighter jets are not a happy image, or something to display as a point of pride. Fighter jets are instruments of war used exclusively to deliver deadly payloads. They are a realist self-defensive necessity at best, and generally just are instruments towards such a deeply grim end that a culture should be ashamed to have the depravity of using them as a parade to raise people's spirits. If Canada has fighter jets it should not be proud of them, it should tuck them away as an unsightly but necessary evil.

I understand the necessity of fighter jets as a pillar of Canadian defence and deterrence. I believe that Canada is justified in owning these jets. However, parading them as a point of pride is morally depraved. They are like prisons: a necessity, not something to be proud of. Many of my friends in Montreal share my concerns. I urge you to take our collective concerns seriously and to do your utmost to make sure this does not happen again.

With my sincerest regards and thanks for your service towards our city,

Nadim Kobeissi

[nadimkobeissi]

On Cryptography and the Golden Age of Islamic Civilization

The following is a beautiful excerpt from Simon Singh's The Code Book: The Science of Secrecy from Ancient Egypt to Quantum Cryptography:

Between the appointment of Abū Bakr in 632 to the death of the fourth caliph, Alī, in 661, Islam spread until half of the known world was under Muslim rule. Then in 750, after a century of consolidation, the start of the Abbasid caliphate (or dynasty) heralded the golden age of Islamic civilization. The arts and sciences flourished in equal measure. Islamic craftsmen bequeathed us magnificent paintings, ornate carvings, and the most elaborate textiles in history, while the legacy of Islamic scientists is evident from the number of Arabic words that pepper the lexicon of modern science such as algebra, alkaline and zenith.

The richness of Islamic culture was to a large part the result of a wealthy and peaceful society. The Abbasid caliphs were less interested than their predecessors in conquest, and instead concentrated on establishing an organized and affluent society. Lower taxes encouraged businesses to grow and gave rise to greater commerce and industry, while strict laws reduced corruption and protected the citizens. All of this relied on an effective system of administration, and in turn the administrators relied on secure communication achieved through the use of encryption.

As well as encrypting sensitive affairs of state, it is documented that officials protected tax records, demonstrating a widespread and routine use of cryptography. Further evidence comes from many administrative manuals, such as the tenth-century Adab al-Kuttāb (“ The Secretaries’ Manual”), which include sections devoted to cryptography. The administrators usually employed a cipher alphabet which was simply a rearrangement of the plain alphabet, as described earlier, but they also used cipher alphabets that contained other types of symbols. For example, a in the plain alphabet might be replaced by # in the cipher alphabet , b might be replaced by +, and so on.

The monoalphabetic substitution cipher is the general name given to any substitution cipher in which the cipher alphabet consists of either letters or symbols, or a mix of both. All the substitution ciphers that we have met so far come within this general category. Had the Arabs merely been familiar with the use of the monoalphabetic substitution cipher, they would not warrant a significant mention in any history of cryptography. However, in addition to employing ciphers, the Arab scholars were also capable of destroying ciphers. They in fact invented cryptanalysis, the science of unscrambling a message without knowledge of the key. While the cryptographer develops new methods of secret writing, it is the cryptanalyst who struggles to find weaknesses in these methods in order to break into secret messages . Arabian cryptanalysts succeeded in finding a method for breaking the monoalphabetic substitution cipher, a cipher that had remained invulnerable for several centuries. Cryptanalysis could not be invented until a civilization had reached a sufficiently sophisticated level of scholarship in several disciplines, including mathematics, statistics and linguistics. The Muslim civilization provided an ideal cradle for cryptanalysis, because Islam demands justice in all spheres of human activity, and achieving this requires knowledge, or "ilm".

Every Muslim is obliged to pursue knowledge in all its forms, and the economic success of the Abbasid caliphate meant that scholars had the time, money and materials required to fulfill their duty. They endeavored to acquire the knowledge of previous civilizations by obtaining Egyptian, Babylonian, Indian, Chinese, Farsi, Syriac, Armenian, Hebrew and Roman texts and translating them into Arabic. In 815, the Caliph al-Ma’mūn established in Baghdad the Bait al-Hikmah (“ House of Wisdom”), a library and center for translation. At the same time as acquiring knowledge, the Islamic civilization was able to disperse it, because it had procured the art of papermaking from the Chinese. The manufacture of paper gave rise to the profession of warraqīn, or “those who handle paper,” human photocopying machines who copied manuscripts and supplied the burgeoning publishing industry. At its peak, tens of thousands of books were published every year, and in just one suburb of Baghdad there were over a hundred bookshops. As well as such classics as Tales from the Thousand and One Nights, these bookshops also sold textbooks on every imaginable subject, and helped to support the most literate and learned society in the world.

[nadimkobeissi]

We've all been there, but none of us remember. As we develop into fully self-aware beings, our subjective experience of the world shifts dramatically. Once we leave infanthood behind, that early window on the world – and what it's like to look through it – is closed to us.

[nadimkobeissi]

miniLock

miniLock was released this Monday on the Chrome Web Store after making it through the peer review and public review periods. The contributions made during the two weeks of public review were truly something to make the open source cryptography community proud. People were wonderfully productive, and individuals like Dmitry Chestnykh, Trevor Perrin and Jenna Fox contributed deeply valuable ideas to improve the protocol design as well as the general usability of the project.

Here's miniLock in a nutshell, from the website:

miniLock lets you encrypt any file quickly and easily, and share it securely with any friend that has a miniLock ID.

Open miniLock and enter your email and passphrase. miniLock uses your passphrase to generate a miniLock ID. You can then send your miniLock ID to friends and they will be able to encrypt files and send them to you. You can send files to your friends once you have their miniLock IDs.

miniLock IDs are very small and easy to communicate. They fit inside a tweet, business card or text message. Entering your passphrase on any computer with miniLock installed will immediately give you the same miniLock ID.

miniLock's also received very positive press coverage, and the reception on Twitter has also been great. Of all the articles written about it this week (and there are many), the one by PC World struct me as the one most worth reading. Ian Paul did a really thoughtful comprehensive review of miniLock, check out his article here.

Also, my miniLock talk at HOPE X in mid-July went quite well. You can watch it on Vimeo, but keep in mind that this talk was given before the public review period began, so some aspects of my describing miniLock's design are out of date.

Oh, and my miniLock ID is quBSaJLXKsRiaSrhgkPnswKocth711H29ZamMi1H9j4Mb.

This standard has a future. I'll keep improving it.

[nadimkobeissi]

Upcoming Cryptography Work

Fast Company has published an interview with me that covers recent advancements with Cryptocat, my design rationale for a lot of Cryptocat work, and also the current Cryptocat Kickstarter that's trying to raise $45,000 for the project.

I'm also traveling to New York City this weekend to attend HOPE X, where I plan to present my next encryption software project, miniLock. WIRED published a preview of miniLock last week.

miniLock will make use of some interesting and original file encryption and file sharing techniques using elliptic curve cryptography. I'll post the slides for my talk right after it's given on Saturday, and will also be releasing miniLock with a full, peer-reviewed design specification, unit test kit, third-party cryptographic audit, API, reference implementation and more.

Here's a good excerpt from the Fast Company interview:

Why the focus on the general public?

Cryptography is my main research interest and I've always had the opinion that you have to focus on practical, applied cryptography. I see a lot of research being done on really theoretical cryptography. But I don't like that approach because it's a very Ivory Tower, academic approach. I really want to focus on the kind of cryptography that has practical benefits to regular individuals in the world. Everything I've done related to cryptography has I think embodied the belief that if you want to do cryptography research it's much more valuable to do stuff that's related to practical or applied cryptography.

Why is it important for you that Cryptocat is free and open source software?

First this is an argument that I believe in at the engineering and programming level. I think that encryption software can't afford to not be open source. I think that in order to evaluate the research and security of any cryptographic software, you need to adhere to , which has been a staple for cryptography for many decades. The principle is simply that you cannot obtain security via obscuring your practices. You have to obtain your security via assuming that the adversary already knows all aspects of the system and nevertheless the system is still secure.

So how does that principle play out with Cryptocat?

The way I enforce this is not only by making Cryptocat open source software, but by making it free software under a permissive license. We document the software and the cryptographic specification and we open up our development process. We hire auditors to do biannual audits and release those audits for the public to examine. It's a very transparent approach to doing cryptography research. Unfortunately it's resulted in the misconception that Cryptocat is more broken than other tools because we keep talking about how many different bugs we find and fix. But the real reason for that is because no other software has this level of transparency.

[nadimkobeissi]

Visit Original Link

VHS Head's Persistence of Vision comes out today on SKAM records. I've listened to his entire discography and I can attest that Adrian Blacow pulls out many more stops in this iteration than in his previous ones. He continues to define his aesthetic, stemming from a mix of Boards of Canada style surrealism and the soul of 90s VHS culture. There's also an accentuated theme of "retro horror" to this release, which VHS Head plays well throughout the album.

Persistence of Vision seems stuck in limbo between danceable beats and IDM-style synths, and the result is surprisingly great. The album was delayed in production, and it shows: a full listen of this 15-track album flows seamlessly from track to track — VHS Head manages to paint his aesthetic cleanly throughout.

A criticism I have of the album is that some of the tracks lean back too comfortably on the Vaporwave scene that engulfed electronic music in 2012 and 2013, thereby allowing these tracks to meander through without contributing meaningfully in terms of melody or sound. Thankfully, such occurrences are rare.

I've chosen to post the track Body Magic since I feel it adequately showcases some variety present in this overall excellent album. But don't let it fool you — the whole album reaches out and covers far more ground. If you're an electronic music snob, Persistence of Vision is almost certainly going to mindfully contribute to your perspective.

I got my copy from iTunes.

[nadimkobeissi]

My Response to DMCA Protection Offers

As the developer of Bluenote, I sometimes get emails from agencies offering to "remove" the "thousands of pirated copies" of Bluenote off the Internet, in a "fast, effective and affordable fashion", using the DMCA.

This "removal" is discussed matter-of-factly, but in the context of their email, I feel like I'm being contacted by a hit man offering to "take care" of someone who's been causing me problems. To add to the climate of fear, they usually include links to Google searches revealing pirated copies of Bluenote. Gasp!

I've never replied to these emails, but the latest from "DMCA Force" annoyed me, and I wrote the following response:

Dear Jason,

Bluenote is especially made in order to be DRM-free, thereby making it not difficult to pirate.

I think that your work, in which you are effectively attempting to profit off Internet censorship enacted by the fears you attempt to instill in content creators, contributes to a less free Internet.

I also think that you should find a job that actually contributes to mankind’s greatest free flow of information instead of stymying it via the force of this kind of mediocre message. I think the fact that you effectively profit off Internet censorship is shameful.

Regards, NK

If you can't afford Bluenote, by all means pirate it. If you feel like I deserve the money, then it would be nice if you paid the five bucks. But I would much rather see myself losing money than have someone play my fears in order to use a misguided law to further censor the free flow of information on the Internet.

[nadimkobeissi]

Visit Original Link

Format & Journey North is a really underrated electronic music track by Oneohtrix Point Never. I say underrated because it's barely available on YouTube and no one seems to have heard of it, even though it's an exceptionally expressive and creative ten-minute piece that's been out there for years.

The track is split into two parts. The first part is calm and reflective (even including a rainforest background). But as the track reaches its halfway point, it transitions into an intense melody with countless overlapping samples that manage to still produce a coherently powerful message.

This is one of my favourite electronic music tracks of all time. As usual, listening only with good headphones or a solid speaker system is recommended.

[nadimkobeissi]

Hitler's Soft Side

Reading accounts of Hitler's behaviour at his Berghof mountain house is fascinating and distressing in equal measure. Here are some quotes from a recent Washington Post interview with one of Hitler's twenty two maids at his mountain retreat (the last surviving one):

"Late at night, Hitler liked to steal away to the kitchen for a bite of “Fuhrer cake,” a specially prepared sheet cake with apples, nuts and raisins that the kitchen was expected to always have on hand."

Another interview from 2008 with one of Hitler's former maids yields the following information:

Recalling her first direct request from [Hitler], she said she was drying some porcelain cups when he came down the stairs. 'Hello,' he said softly. 'Sorry to trouble you, but could you make me some coffee and bring some gingerbread biscuits to my study?'

These tidbits are distressing to me because I've always imagined that these small things, things like sneaking into the kitchen late at night for some of your favourite, special cake, or softly asking someone for a gingerbread biscuit to eat while you work in your study — were the things that unquestioningly granted you some allowance of honesty and humanity. No matter how small these acts are, they reflect a facet that I've always prioritized when evaluating people. But here they are, reflected in modern history's worst human being.

It's easy to imagine that Hitler snuck down for some Fuhrer Cake and then went back upstairs to lock himself in his study and draft an order for another shipment of Zyklon B.

[nadimkobeissi]

How do we know when decryption is successful?

I recently answered a nice question on StackExchange's cryptography forum:

Judging by the algorithm on the Blowfish Wikipedia article, there is no way for the process to fail with an error. How then does GnuPG know when to tell you your password is correct when decrypting a file, rather than proceeding to decrypt meaningless data?

An important property of a ciphertext is that it has to be indistinguishable from truly random data. This allows the encryption cipher to produce ciphertext that reveals no information about the plaintext (other than size) or the encryption key. In fact, this property even allows encryption algorithms to act as pseudorandom byte generators by simply making them generate a stream of random ciphertext.

In the case of stream ciphers, decryption is the result of XORing each ciphertext byte with its corresponding keystream byte in sequence. If the keystream byte corresponding to the ciphertext byte is correct, the result of the XOR will yield the original plaintext byte. Since ciphertext cannot be made to provide any information regarding the plaintext, it follows that it also cannot be made to reveal whether the XOR of a certain keystream byte yielded a correct plaintext for the byte or not. Otherwise, the cipher would be clearly broken and allow us to basically query it for information regarding the plaintext.

A similar principle applies to block ciphers such as Blowfish, except those ciphers operate on entire blocks of bytes instead of individual bytes, and also employ other operations such as substitution in tandem with XORs, organized in structures such as Feistel networks.

Therefore, when it's important to be able to inform the user whether a decryption operation has succeeded in yielding the expected data, cryptography engineers use Message Authentication Codes. Hash-based MACs allow a candidate plaintext to be compared against an authenticated hash value. If the check passes, then we know that it is the correct plaintext and are then able to notify the user that the decryption function was successful as intended.

Generally, if you want to verify the integrity of encrypted data as it goes through the wire to reach someone, you would generate a MAC for the ciphertext. But if you also want to verify that the decrypted plaintext is correct (as seems to be your case here,) then you would generate another MAC for the plaintext and send it along with the ciphertext.

While hash-based MACs are probably the preferred way to do this, they aren't the only way to verify successful decryption. For example, TrueCrypt will check the first four bytes of the volume header in order to see if decryption is successful. This approach is likely to be more error-prone than HMACs in most applications, however.

Finally, there are block cipher modes of operation (such as Galois Counter Mode) that grant block ciphers such as Blowfish or AES the ability to self-authenticate. This pretty much allows a ciphertext to verify its own integrity without the need for an external check. Depending on what you're trying to accomplish, you may want to investigate both HMACs and Galois Counter Mode and decide which offers the verification properties you're looking for.

[nadimkobeissi]

Daniel Rehn is a great guy who takes the time to salvage beautiful art from the computer age. Today he introduced me to these gems, all book covers or magazine art from two or three decades ago.