January 2021 - cypherpunks - lists.cpunks.org

EXIT through Parlour
by professor rat 12 Jan '21

12 Jan '21

https://twitter.com/MatthewKeysLive/status/1348372956246732800 MatthewKeysLive VIDEO: Parler CEO admits his comment yesterday about "many competing for our business" was wrong, says no one wants to host website after being dropped by Amazon. Additionally, he says Parler's attorneys have dropped them as a client. [Fox News] https://bit.ly/3br0AmX

3 3

Exit Donald Dumbfuck
by professor rat 11 Jan '21

11 Jan '21

https://www.palmerreport.com/analysis/so-many-things-are-happening-right-no… Vulture's, sorry, helicopters are circling Vice President Mike Pence’s residence

1 0

Home-made tamper-evident security seals for kids and adults alike
by coderman 11 Jan '21

11 Jan '21

http://blog.ssokolow.com/archives/2017/04/08/home-made-tamper-evident-secur… Home-made tamper-evident security seals for kids and adults alike Posted on [2017-04-08](http://blog.ssokolow.com/archives/2017/04/08/home-made-tamper-e… by [Stephan Sokolow](http://blog.ssokolow.com/archives/author/ssokolow/) Suppose you need to keep siblings, roommates, children, or even friends with wandering hands out of something, but you can’t use a lock. Maybe you’re worried they’ll find the key, maybe you need something that has no metallic parts, or maybe you’re a kid and your parents are worried you will lose the key. This is the kind of thing numbered security seals are good for… but they’re expensive. (The cheapest I’ve found are roughly 50¢ each in packs of 100 or more) …so I decided to do a little experimentation and came up with a cheap-but-reliable solution for homemade tamper seals that’s so simple and safe that even children can do it. Instructions Step 1: Making the seals The most important thing about a security seal is that it’s unique, so it can’t be replaced, and either can’t be reclosed after being opened or you’ll notice if it is. To satisfy these requirements with paper, we need to print or write something on the paper which other people can’t reproduce and which will look wrong if someone cuts and re-glues the seal. If you’ve got a printer, you could print out a random image from the Internet, delete the file and empty your browser history, then cut strips from it to use as seals, but Inkjet ink tends to cost a fortune and most people don’t have laser printers. What I recommend is this: Cut strips from a piece of paper, then sign and doodle all over both sides of them so any cuts or complete replacement will be obvious. (You do it on both sides, so that you can easily see if someone mends a cut by glueing another piece of paper to the inside of the loop.) Don’t forget to use your scissors to make the ends of the strip round, rather than square, so it’s harder for someone to attempt to pick at the joint if they want to try to peel it apart. Step 2: Applying the seals To apply the seal, you just need to pass it through or around whatever you want to seal (eg. cabinet handles) and then glue it into a loop. (a tight loop, if you’re securing knob-shaped cabinet handles, so it can’t be lifted off without breaking it.) If you want to seal a box, try gluing several strips together into a shape similar to the ribbon on a Christmas present. It’s the details which make it tamper-evident: - Get a bottle of Polyvinyl Acetate glue (A.K.A. white glue, school glue, PVA). - Apply a enough glue to one end of the strip that, when squeezed, it won’t leave unglued corners. - Close the loop and squeeze the ends together for at least five seconds, as hard as you can. - Wait at least 5 minutes, but ideally 10 minutes. Step 3: Making the seals tougher Supposing you’re trying to keep really clever people from sneaking access, there are two more tricks you should get an adult to do: - Paint the glued part with clear nail polish to waterproof it, so nobody can try to invisibly open the seal by steaming the glue. (This also makes it harder to try to pick apart.) - Cut some diagonal stripes into the glued spot using a utility knife to absolutely guarantee that any attempt to peel apart the glue will result in obvious tearing. The “Why” I was very specific in my instructions, because I actually did a lot of testing. If you want to repeat my tests yourself, here’s what I did: Goals - A successful seal must not have the glue fail when pulling on both ends of the glued joint. - A successful seal must show obvious damage when someone tries to pick and peel at the glued joint. - A successful seal must have the paper visibly fail before the glue if a solvent is applied to the glued joint. Methodology - Cut a bunch of paper strips, approximately 3″ long. - Glue pairs of strips into longer strips, using various test glues. - On each glued strip, write the time the process was finished, so drying time can be considered. - For each drying time tested, grip the ends of a test strip and pull evenly outward as with a [Christmas cracker](https://en.wikipedia.org/wiki/Christmas_cracker). The test is a success if the paper breaks outside the glued patch. (If it fails far enough from the join, repeat until there’s nothing left to grip) - For each drying time tested, attempt to peel apart the two strips of paper. The test is a success if the paper is visibly damaged in a manner that wouldn’t be hidden by glueing the joint back together. - Repeat each test at least once, to account for variations in the process. - Prepare another set of test strips and allow the glued spot to soak in a drop of mineral or vegetable oil for 30 minutes. Repeat the tests. The glue passes if it wasn’t weakened by the oil. (In my tests, the breakage still consistently occurred in the dry portions of the paper, indicating that the oil had not weakened it significantly.) - Prepare another set of test strips and paint the glued patch thoroughly with clear nail polish. Repeat the tests. The glue passes if the solvent in the nail polish didn’t weaken it. (I have already confirmed in previous experiments that nail polish will waterproof paper without weakening it and, in these tests, the breakage still consistently occurred in the un-painted portions of the paper.) Results I performed these tests with the following dollar-store adhesives: Double-sided tapeI tested two different kinds of foamless double-sided tape. Both were too weak to be suitable, losing their grip on the paper during the pull test with barely any visible effect.Roll-on glue tapeThis contact adhesive passed the pull test, but was not tamper-evident in the face of picking and peeling at the joint and was so thoroughly weakened by the oil test that it was trivial to pull open without harming the paper.Non-frosted Scotch/SellotapeFrom personal experience, I know that the frosted variety is meant to be possible to peel off without harm if you’re careful. I tested the un-frosted kind and, while it just barely passed the pull test, it was too easy to peel off without evidence. Also, it was severely weakened by exposure to oil and I suspect this to be a trait common to all readily available contact adhesives.Roll-on glue tape, plus Scotch/SellotapeDidn’t perform significantly better Scotch/Sellotape on its own.Glue stickPassed the dry and oil tests, but whether it passed the peel test depended very heavily on exactly how I applied it, so I can’t recommend it. It also got severely degraded by the solvent in the clear nail polish.White/School/Elmer’s Glue (PVA/Polyvinyl Acetate)I was surprised how well this performed… though I probably shouldn’t have been, given that it’s used as a bookbinding glue. PVA dries quickly enough that, after 5 minutes, the still-damp paper next to the joint breaks under test and, after 10 minutes, the paper is back to normal. It is so resistant to peeling that the paper tends to tear, and neither the oil nor the nail polish solvent had a measurable effect on the joint’s strength. I did, however, notice that peeling at a corner produced damage that was easier to overlook… thus my recommendation to round off the ends of the strip before gluing. I would have moved on to testing other avenues, such as hot glue, super glue, contact cement, and so on, but I don’t need to test them to know that, at best, they’d match the performance of white glue for this application. (They’re harder to work with, more expensive, no faster to dry, and PVA already produces a bond stronger than and just as water-resistant as the paper around it.) (I did, however, test super glue on some scraps of projector transparencies meant for black-and-white photocopiers. It passed the pull test but was trivial to peel apart without leaving evidence of tampering.) UPDATE 2018-08-15: After I realized how effective lighter fluid is for safely getting tape and other adhesives off books and boxes without damaging them, I ran another round of tests. Lighter fluid does not appear to significantly weaken white glue. However, since my initial tests, I’ve switched to a new bottle of glue and it seems that the resistance of Polyvinyl Acetate (PVA) glues to being peeled off will vary. (I’m not sure whether it varies between batches or just between brands.) As such, since the quality of your white glue cannot be guaranteed, I now strongly recommend the “making the seals tougher” steps be always used (a coating of clear nail polish to seal the seam and some cuts to improve the chances of tearing when trying to peel it apart), with the use of chevron-shaped (< and >) cuts to ensure there’s no angle one can try to peel at which will be completely parallel to every cut.

1 0

HOT here yesterday
by professor rat 11 Jan '21

11 Jan '21

So hot I saw a marsupial blowing on its nuts. https://www.palmerreport.com/analysis/parler-is-over/35431/ These Zenaan, Batshit-crazy, Gramps type lunatics all quit (or got banned from) Twitter and went to Parler, only for it to be shut down. Now they’re presumably reduced to standing in their back yard and yelling their conspiracy theories at squirrels.

1 0

Coderman - a cloudflare-NSA salesman.
by Punk-BatSoup-Stasi 2.0 11 Jan '21

11 Jan '21

coderman has promoted the use of shitflare DNS servers repeatedly, laughably 'arguing' that a centralized NSA server (which is of course just what shitflare is), is 'better' than the more decentralized ISPs' servers. FURTHERMORE, he is advocating the use of 'dns over https' which means that the connection between the shitflare-NSA global server is 'encrypted' preventing any NSA competitors from spying as well. Which is of course the most basic plan of the US military of which coderman is an agent. So again, the guy wants all DNS queries to go straight to the NSA.

1 0

Re: Where does this ' Coderman ' Jackass get off?
by professor rat 11 Jan '21

11 Jan '21

" Its those who make peaceful change impossible that make violent revolution inevitable " Oliver Tambo I have peacefully requested the USA surrender to the anarchic internet in general and a revived PAM-FutureMaps in particular, every year since 2009. And you have a moral problem? Thank fuck its not a REAL problem! Fuck off C-man.

1 0

Tactical support for anti-trust
by professor rat 11 Jan '21

11 Jan '21

I think this is warranted on the basis that SV companies are the new Robber Barons. Bill Gates is a long time target of ours - continuously deadpooled since 97 ( CJ? ) And he's since been joined by Ellison, Thiel, Bezos, Zuckface, Dorsey, etc The usual suspects. Even Tim Cook, whose crypto annoys the FBI enough for them to embarrass themselves, again, needs killing on the basis that a benign despot can easily be worse than a malign one. Of all tyrannies, a tyranny sincerely exercised for the good of its victims may be the most oppressive. It would be better to live under robber barons than under omnipotent moral busybodies. ..." etc... CS Lewis

1 0

Where does this ' Coderman ' Jackass get off?
by professor rat 11 Jan '21

11 Jan '21

Every revolutionary cypherpunk is in a legit state of violent revolt against every nation-state in general the USSA in particular. " National borders are not even road humps on the information superhighway " And here comes this cuntface " Coderman " is again with the detailed technical bumf that is so boring Perry probably wouldn't print it. Mongo would barf. Fuck Telegram. Cypherpunks is a list for anarchists to discuss how strong-crypto helps ( or hurts in some cases) global anarchism - its treason to this list to do what this ' Coderman ' jackass keeps doing. Statist bores are not welcome here - kindly fuck off and let us get on with the social revolution. Tia.

2 1

Cryptography Dispatches: The Most Backdoor-Looking Bug I’ve Ever Seen
by coderman 11 Jan '21

11 Jan '21

https://buttondown.email/cryptography-dispatches/archive/cryptography-dispa… Cryptography Dispatches: The Most Backdoor-Looking Bug I’ve Ever Seen 2021-01-10 Welcome back to [Cryptography Dispatches](https://buttondown.email/cryptography-dispatches?tag=header), my lightly edited newsletter on cryptography engineering. PSA: I've been doing some [livecoding on my Twitch channel](https://www.twitch.tv/filosottile), working on [age](https://age-encryption.org), check it out! — Filippo This is the story of a bug that was discovered and fixed in Telegram’s self-rolled cryptographic protocol about seven years ago. The bug didn’t get any press, and no one seems to know about it, probably because it was only published in Russian. To this day, it’s the most backdoor-looking bug I’ve ever seen. Google Translate does a good enough job on the [original article](https://habrahabr.ru/post/206900/), which is still available on Habr, but I’m going to walk you through it along with some context. Telegram is a popular chat app that uses its own… bizarre protocol to encrypt chats, called MTProto. The protocol is used both to encrypt all messages to the Telegram server, and to encrypt opt-in 1:1 end-to-end “Secret Chats”. In text I can’t do justice to the facial expressions of cryptographers when you mention Telegram’s protocol, so just believe me that it’sweird. The current consensus seems to be that the latest version is not broken in known ways that are severe or relevant enough to affect end users, assuming the implementation is correct. That is about as safe as leaving exposed wires around your house because they are either not live or placed high enough that no one should touch them. The original version was, however, completely broken, in the most puzzling of ways. End-to-end Telegram chat sessions use [finite-field Diffie-Hellman](https://en.wikipedia.org/wiki/Diffie%E2%80%93Hellman_key_ex… to establish a shared key between the two participants. The negotiation happens through messages relayed by the Telegram server. Diffie-Hellman is a fundamental building block of many cryptosystems, and it allows two parties to establish a shared secret that any eavesdroppers can’t derive. It is however only one part of a secure key exchange, because an attacker capable of intercepting the messages could simply establish two separate sessions with the two parties, carrying out a[Person-in-the-Middle](https://en.wikipedia.org/wiki/Person-in-the-middle_… attack. The parties need some way to verify they derived the same secret. In TLS, they use a signature from a certificate. In most secure chat apps, there is a fingerprint (“Safety Numbers” in Signal) that the two parties can compare out-of-band. What’s important is that if the two sides derived the same secret, they can be sure no one else has access to it. The Telegram key exchange is described in [the “Key Generation” section of Telegram’s end-to-end API docs](https://web.archive.org/web/20131220000537/https://core.telegram.org/…. Concretely, Alice requests the DH parameters (p, g) from Telegram, painstakingly verifies them, computes a random a value, and sends g^a mod p to Telegram. Bob receives (p, g, g^a mod p), similarly computes b and g^b mod p, and sends the latter back (along with a truncated hash of the derived key, for some reason). Now, normally the two sides would compute the shared key as (g^a)^b mod p and (g^b)^a mod p. Instead, the original version of MTProto computed it as (g^a)^b mod p XOR nonce where nonce was an arbitrary, supposedly random value sent by the server along with the peer’s public contribution. This was a completely non-standard and useless addition, and all it did was let the server perform an undetected Person-in-the-Middle attack. Let’s see how. In a normal PitM, the server negotiates two separate Diffie-Hellman sessions with Alice and Bob, who end up with different shared keys, which they could detect by comparing fingerprints. Alice Telegram Bob a = random() A = g^a mod p -> t = random() T = g^t mod p -> b = random() <- B = g^b mod p key = T^b mod p <- T key = T^a mod p T^a mod p != T^b mod p With the nonce addition, however, the server could “fix” Alice’s key to match Bob’s by manipulating Alice’s nonce. The two parties would end up with the same fingerprint, and couldn’t tell that an attack happened, but the server (and no one else) would know the shared key, allowing it to decrypt all messages. nonce_bob = random() key_bob = T^b mod p XOR nonce_bob nonce_alice = A^t mod p XOR B^t mod p XOR nonce_bob key_alice = T^a mod p XOR nonce_alice = T^a mod p XOR (A^t mod p XOR B^t mod p XOR nonce_bob) = B^t mod p XOR nonce_bob = key_bob Why do I say this addition was useless? Because it literally had no purpose! Indeed, the vulnerability was [fixed by silently removing the nonce step from the docs](https://web.archive.org/web/diff/20131220000537/20131225140924/http:/…. [A later API revision](https://core.telegram.org/constructor/encryptedChatRequested?laye… removed the nonce parameter with the caption “Improve secret chats”. All [the original API reference](https://web.archive.org/web/20131028041748/http://core.telegram.… said about the nonce is “Random server sequence for calculation of key”. I never heard a plausible explanation for why the designers of MTProto went out of their way to add useless complexity to their protocol, with the only outcome of making undetectable interception possible. Edit (2021-01-11): [@asdofindia linked me on Twitter](https://twitter.com/asdofindia/status/1348491279798128641) to [an official statement by Telegram about this](https://telegram.org/blog/crowdsourcing-a-more-secure-future) that I couldn’t find anymore. It claims the nonce was there to protect clients with weak random number generators. Here’s what I had buried into a footnote when I couldn’t find a citation to attribute that explanation to Telegram: > This doesn’t make sense for a number of reasons: 1) clients with weak randomness are likely to be toast anyway, because Telegram’s bizarro not-a-MAC relies on randomness in the payload to avoid an offline decryption oracle (there is a plaintext hash of the payload on the wire, I told you this was weird!); 2) the API also allows clients to request random bytes from the server to XOR with their secret share; and 3) defending against weak randomness by relying on a server contribution defends against everything but the server, which is the relevant attacker in the end-to-end setting. (Said another way, anyone that can intercept client-server messages can see the extra randomness, making it moot.) Non-practitioners might think this is a reasonable defense in depth, belts and suspenders kind of thing, but in cryptography engineering adding complexity to defend against scenarios that lead to compromise anyway is simply pointless. Anyway, it’s been a while, the world is a different place now, and maybe [Hanlon’s razor](https://en.wikipedia.org/wiki/Hanlon%27s_razor) cuts deeper than I thought. I think there are better reasons not to use Telegram today than this old bug , but it’s still what I think about every time people talk about far-fetched “bugdoors”. The bar is high!

1 0

Welcome to Belmarsh
by professor rat 11 Jan '21

11 Jan '21

https://www.hitc.com/en-gb/2020/01/16/inside-belmarsh-prison-how-many-episo… A likable journalist! We ought to have him stuffed.

1 0