Comments

Peter CurryJuly 14, 2020 7:40 AM

As a security researcher it really is important for you to emphasize that the four-rotor design is not secure. Anyone with a Colossus Computer can break it. Please hold out for the seven-rotor at least, preferably the eight-rotor design.

Clive RobinsonJuly 14, 2020 8:33 AM

@ Peter Curry,

As a security researcher it really is important for you to emphasize that the four-rotor design is not secure

Nor will any rotor machine that uses an Enigma style "reflector", "predictable rotor stepping", or deficient rotor wiring...

Also the "plug board" and "hour box" whilst apparently adding many many more states actually do not do that much as they become a simple substitution cipher for the whole time they are in use at any given setting. In essence you can ignore them and break the rotor polyalphabetic cipher then do a fairly simple "fix-up" frequency analysis to get the plain text.

That's not to say that rotor machines can not be secure in certain circumstances, but you really need to know what you are doing with wiring the rotors and stepping them. All in all even in a software implementation they really are not worth the effort from a security perspective.

That's not to say I would not like to own one, I could add it to my collection of cold war spy-sets and similar bits such as the 807 based antenna amplifier from a Y-Station.

BillJuly 14, 2020 9:09 AM

It really doesn't matter how secure it is because it will be used rarely if ever, and only to demonstrate how it worked. And I'm sure only under gentle conditions. How many historic artifacts get taken out into the field and put to real use?

Jim KJuly 14, 2020 9:38 AM

Does the receiving party need one too?
Might be about as useful as a fax machine...

David (in Toronto)July 14, 2020 10:01 AM

The machine used to break the Enigma keys is the Bombe designed by Alan Turing and improved by Gordon Welchman. The Colossus machine was used to break the Lorenz cipher. It was designed by Tommy Flowers.

The Uhr box only stumped them for a few days and was in limited use. A case of too little, too late. Much more frightening and rare was the "D" reflecting rotor that was rewireable.

Welchman's book The Hut Six story has an excellent account of the details of the Bombe.

The basic design is a century old. It was intended for short messages only. It was broken because of systematic weaknesses and because the code breakers could accumulate a "depth" (i.e. volume) of messages enciphered with the same key.

While Enigma is insecure the M4 distributed cracking project a number of years ago still had some challenges cracking 3 late war naval messages.

rrdJuly 14, 2020 10:53 AM

@ Romulo Cholewa

>> I wonder how many died related to this piece of history.

I understand you to be talking about that particular machine, which is an interesting rhetorical question.

Looked at as a technology-of-its-day, I've seen it theorized that Bletchley Park's cracking of the various Enigma variants shortened the war by perhaps two years. From that perspective, the answer is, "Not as many as the Nazis wanted to, but more than they would have with a lesser piece of kit."

_______

@ ALL

There is a 4-part BBC documentary from the late-90s called "Station X" that documents Bletchley Park's establishment, evolution and utter destruction.

I find it excellent, but I'd be curious to know from my betters if it has any factual inaccuracies.

WayneJuly 14, 2020 12:12 PM

@David:

The Poles were breaking Enigma using an earlier version of the Bomba (variations of the spelling) for seven years before Turing got the device and did some refinements in 1940. They didn't share the information with their allies until just before they were invaded when they were able to sneak one of their Bombes to England and destroy the rest so the Germans remained clueless that the Enigma cipher had been broken years earlier.

I have/had a fascinating book that goes into the Polish history of the breaking of the Enigma, but I can't remember the title and have no idea where it is, if I still have it.

https://en.wikipedia.org/wiki/Bomba_(cryptography)

David RudlingJuly 14, 2020 12:30 PM

@rrd
The problem with "Station X" and anything else of that vintage is not so much inaccuracy as incompleteness. Swathes of additional material have been slowly released since the late-90s and continue to be released and await release e.g. item HW 43/82 in the UK National Archives, the overall official history of SIXTA - the traffic analysis organisation, was only released in September 2018 and a related document HW 43/63 which appears to be the more detailed report on the Traffic Analysis Party at Bletchley Park which was a key and early part of the overall traffic analysis structure remains withheld despite my repeated efforts to have it released. By no means has all wartime Bletchley Park material been released. But that is not in any way to belittle the "Station X" documentary series.

David RudlingJuly 14, 2020 12:44 PM

@Wayne
Possibly the book you are referring to is titled, at least in the English translation of the Polish original, "ENIGMA - How the German Machine Cipher Was Broken and How It Was Read by the Allies in World War Two" by Wladyslaw Kozaczuk. It was originally published in English by University Publications of America but my copy is a UK reprint by Greenwood Press of London. Highly recommended assuming the moderators don't delete this as a blatant advert.

David (in Toronto)July 14, 2020 1:16 PM

@Wayne You are referring to Rejewski and his two colleagues. I am aware. They figured out how to break the key from the key indicator procedure which used the machine. The Germans changed the process and the Poles didn't have the resources to proceed. Turing significantly improved the Bombe. Welchman came up with the diagonal board which cut down false stops. Sadly, Bletchley didn't use Rejewki and company.

Clive RobinsonJuly 14, 2020 2:16 PM

@ David (in Toronto),

Turing significantly improved the Bombe. Welchman came up with the diagonal board which cut down false stops. Sadly, Bletchley didn't use Rejewki and company.

The story is Turing took considerable time to improve the Polish bomba possibly due to having to spend quite a bit of time dealing with incompetent managment.

However Gordon Welchman was brought in compleatly uninitiated and came up with a design for his own bombe in issolation not knowing of Turings previous work. He did this in his spare time over about a three week period. On approaching his managment he was initially given the cold shoulder. So he carried on thinking about it. Eventually he was told about the Turing bombe after he had exolained his own method. Apparently what blew Turings sicks off was when he sketched out the diagonal board design infront of Turing.

However Welchman did not officially get credit, apparently that went to "Doc" Kean who made the engineering drawings up from Welchman's sketches.

It lead to a bitter row and Welchman leaving Britain for the US, where he wrote his book.

Unfortunately the British authorities played dirty yet again and had Welchman put into a legal limbo where his security clearance was revoked and they threatened to prosecute as at that time things were getting "political" in Britain.

He was still living under this cloud when he died of a serious illness.

However from his work in the US a military high resiliance network was built which is still in use in NATO countries, and a spin off in most other parts of the world as well, which most know as either the Internet or Web.

The UK still has a very bad attitude to creators and scientists. Honours are reserved for hierarchy baboons in both the military and civil service and those in industry who either push weapons of war into third world countries or give generous kick backs to political funds in return not just for honours but no look lucrative contracts...

Lawrence D’OliveiroJuly 14, 2020 8:02 PM

Four rotors means its a Naval Enigma, as I recall. The Army one only had three rotors.

AycupJuly 15, 2020 7:44 AM


Hi, The four rotors mean a Sea Enigma as far as I can remember. There were only three rotors in Ordu.

Who?July 15, 2020 7:53 AM

Who wants a "typewriter" with a german QWERTZ keyboard? I guess it would be worse a french AZERTY keyboard or a Dvorak one!

Seriously, it looks like new. This Enigma machine is in perfect conditions, at least it seems well preserved as shown in the pictures. No one would say it is an eighty years old encryption device.

rrdJuly 15, 2020 8:01 AM

@ ALL

Interesting. "Station X" hand-waved over the origin of the Bombe, just saying it was Turing's genius that allowed the machine to eliminate negative matches. The three Poles are only mentioned as having figured out the machine's internals and passing along that information to the Brits, but it was their actually sending a Bombe to Britain while ensuring the Nazis never learned of the compromise that may have been the pivotal factor in the war (after America's industrial might).

And, yes, it's still an *excellent* documentary.

Thanks, all.

ThunderbirdJuly 15, 2020 11:52 AM

I know that the comment about security was tongue-in-cheek, but I wonder how secure four-rotor Enigma traffic would actually be on the internet? If it were only used occasionally for short messages using an agreed-in-advance key schedule, it seems like it would be pretty secure. Mostly based on the idea that no one is going to imagine that that's what it is and try to crack it.

In fact, it would be fun to incorporate into a work of fiction (if no one already has).

Note: I am not actually recommending this to anyone. It is quite possiblethat anyone familiar with historical encryption techniques would look at a few messages and say "that looks like an Enigma!" My level of cryptographic experience is suited to dealing with the Jumble in the paper...


MikeAJuly 15, 2020 1:16 PM

@Thunderbird: I have considered such a thing, and there was an episode of the US "Sherlock in or times" series "Elemental" which involved one of the villains using an Enigma. Why they didn't just use a software simulator and lower their CAPEX, I don't know, but maybe they had been following Clive and taken his (excellent) advice to have the security endpoints "outside" the communications endpoint.

Anyway, were i to do this ("historical" cipher over "modern" comm channel), I'd probably start with Snefru or Lucifer. Probably the former, as IIRC Lucifer was vulnerable to differential cryptanalysis (hence the "odd" changes made at NSA suggestion), but it can be implemented on a 8-bit computer from the 1970s, so there is that. OTOH, Snefru was designed (again IIRC) to be efficiently implemented in software on non-bleeding-edge machines.

Of course, this would be at a disadvantage against more modern ciphers, _if_ those modern ciphers were done on a trustworthy (not a "trusted") system. Good luck with that condition. OTOOH, if "they" want your comms, they will just use a $5 wrench.

Clive RobinsonJuly 15, 2020 5:06 PM

@ MikeA, Thunderbird, ALL,

Good luck with that condition. OTOOH, if "they" want your comms, they will just use a $5 wrench.

That's a threat I've been thinking about after a conversation with @Nick P some years ago and it turns out there are several ways to beat it.

What you need to do is play against their weaknesses by turning them into your strengths (yes it sounds trite but it is nevertheless true).

Their first weakness is that money and political power make you neither omnipotent or omnipresent. They are still as limited as you are by both time and location which levels the playing field quite a bit.

Their second weakness is no mater how much they threaten, tourture, or drug you, you can not tell them what you do not know.

Thus you need to come up with a system where,

1, You do not know a valid secret.
2, They run rapidly out of time.
3, They need you to be outside of their jurisdiction in a free state.

When you think on these it's actually not that hard to see several possibilities.

For instance assume that the hardware device it's self has a TRNG and say an 8kbit PKI private key. If you want to communicate with it you need to know it's public key and some shared secret as a method of authentication, but that secret is in effect encrypted within the device (M of N shares).

Breaking the device open does not give you very much. Because you can build up a system whereby even having the traffic as well does not give them sufficient information.

Provided the device is "filled" from outside the adversary's juresdiction not only do you have no idea what the secrets are the adversary does not have a way to get at them.

Thus those you wish to communicate with fill your device as part of the communications, and the device then wipes it.

To make the time interval as short as possible assume an evolving crypto key. To see how this might be done you can "whiten" it with an agreed secret and then hash it or encrypt it in an appropriate cipher and mode (say AES-CBC) and thus use a new key with each block of the message. Thus the key for each message block exists for only a few micro or milliseconds.

Whilst it is a bit more complicated than that you should get the basic feel for what could be done.

echoJuly 15, 2020 6:51 PM

@Clive

I'm somewhat over my funk other than to say I don't think people handle the topic of human rights and equality in a security context very well. It's really quiet an easy topic but office politics and individual perspectives get in the way including and especially among academics and mainstream media. I'd like to have another go at this topic but not now and not in the current form. It may just be a process of time and it's not a hill I'm prepared to die on.

As of the current discussion:

I presume good quality intelligence agencies not only know of this concept but also may use its equivalent.

I cannot comment on GCHQ as they are secretive and lay and rarely pusblish anything interesting or readable which might betray said lethargy or heaven forbid stop freeloading or tilt the focus of attention away from the profilic US. However, the NSAdo haveunplublished encryption protocols so a thought experiment.

1. Security by obscurity is a real and quite useful thing. I have a secret which I'm not going to tell you or even let you know thereis a secret. This works quite well as the laws of phsyics would attest by being a byztantine pile of unknown unknowns.

2. A security scheme can be polymorphic.

3. Constantly evolving R&D creates a "bleeding edge" which can itself evolve and be calibrated.

4. Monitoring curent internal and external research provides a measure of useful usable time allowance.

5. Don't use it unless you have to. See also #1.

I'm also fairly sure that US military doctrine encompases this philosophy which will also include the subsystems of networked stealth systems such as the F-35.

I have a nasty habit of making missile designers and fastjet service engineers clam up fast by knowing more about the topics than they assumed. I get marginally more information out of Mountain and Arctic trained Royal Marines by sounding more dumb than I am. Honestly, you don't know how much cognitive load it takes to appear dumb which is why stupid is its own form of genius I suppose. Then again Chris Grayling, who the government tried to manage into the position of running the Intelligence Committee, just got punked by one of the Tories own,Julian Lewis, who then had the whip removed for out-Torying the Tories for pulling a dodge and getting himself elected to the role.

Oh, and while we're on this topic experts do matter. I'm not going to disclose sources or what the technical domain is but if it wasn't for strong QA telling booksmart managers trying to cut corners to take a hike a few European aircraft would be falling out of the sky. The cultural and economic pressures from the likes of Boeing, Intel, and Microsoft et al must be resisted as must the culture of bullying and lying which rewards those who cut corners and backscratching incestous remuneration committees.

I wonder what neutered and watered down version of the Russia Report will get published? See #1.

David (in Toronto)July 16, 2020 12:04 AM

@Randie - I see you saw the announcement of the cyclometer rebuild (pre-pre-bombe)
@MikeA - I think you will find that a "mechanics hammer" or a "rock hammer" is more effective. And there is also the big envelope of cash technique which recently was used at a Postbank in South Africa.
@Clive - a lot of traditional key management works this way. You have access to a 256 bit key part, I have access to another. Maybe there are others too. The crypto hardware XOR's them together for the key. Even if we all have photographic memories, they need to nab us all.

Clive RobinsonJuly 16, 2020 3:08 AM

@ echo,

I presume good quality intelligence agencies not only know of this concept but also may use its equivalent.

If I can think it up so can others, it's just a question of asking the questions in the right order to see the path through the forrest to one of a number of solutions.

So I assume that somewhere even in bad quality intelligence agencies this concept is known. However as we saw with the CIA and China they still appear to not be thinking things through to working solutions, and people died needlessly because of it.

Security by obscurity is a real and quite useful thing.

Yes the same as an OTP, and they both suffer from similar problems, primarily you can realy only use them once and as a consequence they do not scale well. Which means you have to "use with care" and as some kind of "bootstrap" / "leaver" to get a more resilient system up and running, and I'm one of those who likes the reliability that resilience gives.

Honestly, you don't know how much cognitive load it takes to appear dumb...

It depends on how you look at it, most look at it as being "simple" because "you are being simple". Even good actors find it hard to stay in character for a two hour play... Thus doing it all the time for extended periods is actuall very hard. Because in effect it's like living under cover any mistake and it's blown and with it what you hoped to achieve. Thus you have to not just double think but triple think everything you say and do, and then do as little as you can. The result being you have to escape from it just to "decompress" and that in of it's self presents dangers as well. So yes for those that have tried to live what is a double life it is hard very hard and can age you faster than you would think.

The cultural and economic pressures...

Ever since Victorian artisans started killing people with boilers, legislation, regulation and independent oversight was seen as the way to prevent needless injuries and deaths. Unfortunatly these cost money, and those that have got into power see this as their money to do with as they wish... Thus since the 1980's of Thatcher and Reagan, I've been seeing regulation dying the death of a thousand cuts. Unfortunatly as you note Boeing recently demonstrated why this should not be done, and they are just but one of very many making calculations on profit by death... For instance the current crop of politicians are very much doing this with COVID-19 which is not new as a policy, it's just that this time it's more brutal in your face and in large numbers. So much easier for people to see with their own eyes even though they might not wish to. The question thus remains, "Can we do anything to stop it? and if we can will we or not? and if not why not?".

Clive RobinsonJuly 16, 2020 3:13 AM

@ David (in Toronto),

Even if we all have photographic memories, they need to nab us all.

Nabbing us all, is a standard policy these days along with "collect it all". The problem is they are not realy interested in if you know something or not, it's just part of the "empire's" process "on the off chance" that's what "gitmo" is. We also saw this go public with a dead terrorist and his Apple phone as the DoJ and FBI tried to get "caselaw" in their favour... Apple fought back and the DoJ / FBI dropped the case rather than get a non favourable outcome.

Which is why my intent is to remove that "on the off chance" excuse for their behaviours and ever increasing empires they build on it. Normal people do not need lunitics running the asylum of intel agencies and guard labour at the behest of others whims or crazed power grabs.

echoJuly 16, 2020 4:09 AM

@Clive

I'm glad you mentioned Covid-19 as an example as this saved me explaining it and largely for the reasons you stated.

I'm actually glad you mentioned perspective. Yes, that's a good one. I find observational comment like this far more useful than therapy which I'm pretty resistant to however its managed.

echoJuly 16, 2020 4:48 PM

Less glamorous handheld and much less secure encryption devices are available during the same auction. I'm still interested in what the theoretically possible smallest mechanical encryption device capable of hand assembly and naked eye verification is. Sadly my knowledge of the underlying maths and mechanical logic requirements are not good enough to have an opinion.

echoJuly 16, 2020 5:19 PM

@Singapore Noodles

I suppose this might be an approach to consider. Thank you! Following your suggestion I did a quick search. Someone has made a mechanical Turing machine out of wood!

My personal starting point is from the point of view of bespoke watch design. I'm equally interested not just in the functionality but also aesthetics and wearability. There is the issue of mechnical logic but I've also wondered about OTPs and whether it is best stored in a wire or on a disc. I don't know enough about materials to know if there can be an optical-mechanical component including a read and/or write mechanism and amplifier. For the purposes of this exercise I exclude nanotechnology as it is too small to be seen with the naked eye or simple lens.

If the device looked and functioned like a watch in ordinary use this would be a bonus.

JonJuly 17, 2020 2:42 AM

@ Clive Robinson

What you need to do is play against their weaknesses by turning them into your strengths (yes it sounds trite but it is nevertheless true).

Their first weakness is that money and political power make you neither omnipotent or omnipresent. They are still as limited as you are by both time and location which levels the playing field quite a bit.

Their second weakness is no mater how much they threaten, tourture, or drug you, you can not tell them what you do not know.

Thus you need to come up with a system where,

1, You do not know a valid secret.
2, They run rapidly out of time.

This is why I proposed "The Random Data Email Exchange".

The idea being that a group of people, loosely connected, on an irregular but fairly frequent basis (daily or so?), email each other a load of totally random data.

1) Any well-encrypted data is indistinguishable from random.

2) Given a specially-created one-time pad, any random data can be "decrypted" into anything of less than or equal length.

3) In some places, having encrypted data and refusing to give the key is a criminal offense(!)

4) Unless you have a plausible argument that there is no key...

5) And, given a back channel, you can "suggest to George that Henry might want to try XORing together 'Random Data 19898345' with 'Random Data 2341256' and see what happens".

The authorities can just XOR together all of them - but their problem goes exponential while yours doesn't.

Jon

#3's a bit harder... ;-)

JonJuly 17, 2020 3:02 AM

@ Clive Robinson

3, They need you to be outside of their jurisdiction in a free state.

Given that the Mossad and the CIA (among others) have been known to engage in kidnapping and assassination (among other things) in countries in which they have no legal authority at all, I'd say that your #3 requirement is impossible.

Perhaps that's the point.

Jon

echoJuly 17, 2020 4:31 AM

@Jon

Acting against point 3 (jurisdiction) seems to be covered by point 2 (running out of time). Now the basic scheme may protect a mission and breaks may form a mitigation against jurisdiction attacks. That said the operation to capture Bin Laden did eventually work through everything to capture him.

I'm a blabbermouth so never likely to need a scheme like this which is, oddly, a security scheme in its own right.

Clive RobinsonJuly 17, 2020 9:53 AM

@ Jon,

Given that the Mossad and the CIA (among others) have been known to engage in kidnapping

Perhaps I paraphrased it to much, look at it as,

3a, They need you to be outside of their jurisdiction,

3b, They need you to also be in a free uncoerced or otherwise compelled state.

Let me explain why.

To define a circle you need three points on it's circumference. If you have only two all you get is a straight line or in the case of one some coordinate point.

So if you use the radius or focus point of that circle for a secret key, you need all three points on the circumference to the required accuracy.

Thus if you give three unrelated people you trust say 10 coordinate points only one of which is on the circles circumference locked up in a crypto vault (think password vault). When you turn up they decide if you are under duress or some other influence. If they think you they give you one of the nine wrong coordinates.

Thus neither you nor your three trusted people know what the key is nore can they individually work it out. If these people do not know who the others are and they are all in different jurisdictions things become problematical for the agency (especially if you add in a time based component).

Thus any agency has a problem in that if one of those three circumference points are wrong the device becomes locked in some way where other knowledge then becomes required.

Such "M of N" secret sharing is a well established process and N within reason can be as large as you like with M being some subset of those N shares.

Without going into details you can extend the system such that you do not even need to know where the M individuals are or how big a group M is.

Thus the only reason to imprison or tourture you is as a lesson to other people. Which generally is extreamly counter productive.

Singapore Noodles July 17, 2020 12:23 PM

@echo

The big grand complication watch has ~1000 parts. How many parts does an Enigma have ? Maybe your idea is quite feasible.

JonJuly 17, 2020 2:56 PM

@ Clive Robinson:

Not quite...

To define a circle you need three points on it's circumference. If you have only two all you get is a straight line or in the case of one some coordinate point.

Two points can define many circles (just not any circles with diameter less than the distance between), and if the law enforcement personnel decide that your two points are on the circle they have in mind, you're, well, you'll want a paddle. And a lawyer.

Still, I think you're missing the point (ha ha). The idea here isn't anti-$5 wrench tactics, rather dealing with #1 and #2; and flooding tactics work okay there.

Jon

55962

echoJuly 17, 2020 5:11 PM

@Singapore Noodles

I usually wear ladies watches myself. They're almost always smaller. At the moment I wouldn't be too concerned about the appearance or haptics as it would contrain initial design ideas. Something of the order of what you are suggesting size and complexity wise would technically fit the problem. I'm also open to mechanical shortcuts if novel forms of compressing the logic into fewer or smaller parts is possible.

I've wondered about whether a fine wire could be used to store a OTP and be read like it was a record. It's a lot smaller than punched tape! "The first wire recorder was invented in 1898 by Danish engineer Valdemar Poulsen." "One of the world's first stored-program computers, SEAC, built in 1950 at the U.S. National Bureau of Standards, used wire recorders to store digital data." (Source: Wikipedia.)

echoJuly 17, 2020 5:42 PM

@clive

I'm a bit puzzled by what problem you are trying to solve. Is your scheme about protecting a "dumb courier" or something else? I'm also not sure securing the endpoint is scaleable for all cases. There's also the issue that detention in extreme cases may be about locking up all the unknown unknowns. I think what I'm asking for is some clarification and restating what the exact problem is because there's too many variables in play I'm not sure about.

myliitJuly 18, 2020 4:29 AM

@MK

“Price Realised: GBP 347,250”

Thanks, I was wondering what the Enigma sold for.

Clive RobinsonJuly 18, 2020 6:02 AM

@ echo,

I'm a bit puzzled by what problem you are trying to solve

I suspect you kind of know but have not thought about it in the right way.

Any security system from a whole in the ground upwards fundamentally has two parts,

1, Technical agency.
2, Human agency.

The main purpose of security though can be sumed up by,

    "Three can keep a secret when the other two are dead"

Whilst an extream view it makes it makes the point that the main purpose of security is to protect the interests of one party against all other parties, even the second party required for communications.

In essence it's the control of information. Communication is a necesity to give information value, that is information has no value when not in use.

At a very basic level information has no physical component, however you can modulate matter/energy with it to give it a form where physical enterties can do one of three things with it,

1, Store information.
2, Communicate information.
3, Process information.

You can only have full control over information is you are neither communicating or processing the information.

From early times we have regarded wealth in terms of the possession of unique physical objects. Thus from before the time mankind could effectively communicate puting objects of value out of sight of others was a way to protect them. Thus from simple "out of sight" to vaults within vaults in secure fascilities is a logical progression. Which is why people talk about cryptographic systems like they are safes with keys that we lock ibformation away in. That is we have projected protection mechanisms from tangible physical objects to intangible information objects. However the projection only holds true for stored objects that are "unseen" by second or more parties. This is because we can not copy unique physical objects perfectly, but we can granular information objects. Worse copying information objects has now become so inexpensive and easy there is realy no way the first party can know if other parties have done so.

However to process information and make it of use/value, requires you to comnunicate it from where it is stored to where it is processed.

There are two basic aspects to communication, firstly is you communicate information from one point in space to another point in space, as this requires matter/energy the movment is constrained by the speed of light which in turn imposes a temporal asspect to communications, that is it always has a time component no matter how small related to the distance between the points in space (this has other seamingly wierd aspects when you start looking at "time cones" and the like).

As has been seen on the Internet if a third party reroutes a communication between two parties they can increase the communications time sufficiently that they can get a "reply" into the first party prior to the real reply from the second party. As the protocol has the deficiency that it accepts the first reply and drops all others without inspection --as they are assumed to be delayed copies-- the first party gets no warning they are being interfered with. The Great Firewall of China used this trick to stop people inside seeing information from outside[1]. The NSA, GCHQ and presuably other SigInt agencies have used it to send false information.

The important point to note is that by an inherent temporal delay in a technical system a third party was able to

1, Copy information from a communication.
2, Process information from the communication.
3, Take action prejudicial to the communicating parties.

This temporal issue exists with all technical systems as it's fundemental to their functioning and can not be solved, only mittigated against in some way.

It's one of the reasons that "code signing" serves a purpose against malicious third party, who's location can not be factualy determind by either the first or second party.

That is the fundemental nature of the Internet and all multi-node networks limits what you can know as factual to the first router upstream of your organisational perimeter (hence the Five-Eyes prediliction of taking over routers away from the leaf nodes in the Internet).

Once you realise that this temporal issue exists it's a bit of an eye opener because you start to see it where ever you look[2].

The thing is it's a masive exploit vector not just within technical systems but also the human systems around then, where "Human agency" can work it's evils.

As I've noted with technical systems you likewise can not stop temporal vectors in human systems, thus all you can do is mitigate them. One way is to make the use of temporal vectors as a method of excusing human agency evil is to make it "factualy obvious" that is if you can prove it's not possible for you or somebody else to know something then any third party using some coercive method is not going to get what they excuse such behaviour with. Thus their real reason to tourture or denyvliberty becomes more obvious to all for what it mostly is "The excercise of abusive power because they can for show or pleasure" which as it becomes more widely known reduces their ability to exercise such abusive power.

[1] In essence all you do is muck up the syn/ack TCP open protocol by sending a fake "close". Many mobile phone operators in places like the UK use the same trick to stop you getting at their "naughty list of sites" which in one case at one point included schneier.com... You can actually spot this black listing because of the speed you get a "site not available" notification. If a site is blacklisted it's very fast, if the site it's self sends it as it's undergoing maintenance etc it's slower, and if the site were genuinely "not there" then you would get an appreciable wait for a time out to occur. So in theory you could get an approximate range to the node doing the blocking and a tool like "ping" or "traceroute" give you the hop distance abd IP information. This is such usefull information to your online situational awareness it makes you wonder just why it's not built into mobile phone OS's and their web browsers...

[2] Back in the first half of the 1990's I was at a "works Christmas Party" and it had what looked like a lavish buffet enough to sait the appetite of any worker five times over. Then somebody said "theres a lot of scotch eggs" and suddenly all you could see was scotch eggs... I'm told this is like the "Pink Rhinoceros" trick, where once some one has said it you can not stop thinking about them, unless you don't know what a rhinoceros (mind you I still can not realy tell the difference between black and white rhinos they both look grey to me ;-)

AndersJuly 18, 2020 2:27 PM

@Peter Curry

"As a security researcher it really is important for you to emphasize that the four-rotor design is not secure"

Rotor count doesn't count here. Enigma had one fundamental flaw -
a letter could never be encoded into itself.

echoJuly 18, 2020 5:41 PM

@Clive

I suspect we're arguing from different points of view. The problem you're trying to solve simply doesn't exist for me. I drafted a long reply explaining all the none technological issues why but this falls very firmly into TMI. It's probably one of those things I'd have to show you before the penny drops and possibly not even then which would rather prove the point.

Clive RobinsonJuly 19, 2020 2:59 AM

@ echo,

The problem you're trying to solve simply doesn't exist for me.

But it provably does for others, and it can be changed.

We've seen the start of this with police being videoed with mobile phones. It's got to the point where even police officers are saying publically the behaviour of their colleagues is criminal, thus can nolonger be ignored or tolerated.

The old "out of sight out of mind" behaviour has been pushed out by technology. With it the excuses such as "the witness has an agenda" implying they were lying.

Look up what went on in "The battle of Lewisham" and subsequently. Whilst some of "the canteen culture" still exists (eg Steven Lawrence) it is slowly going little by little with each retirment of an "old lag".

For better or worse technology is "an agent of change in society" politicians come and go as do their policies (Cameron pro China Johnson anti) and even legislation. So it's a question of how we the "directing minds" use technology as banning it is pointless, history shows us that over and over. Even a millennia ago Anglo-Saxon King Canute knew that power had limits, something his advisors apparently did not.

But a more contemporary observation was made in 2014 over the Ukrain crisis,

    "Political power or office often gives those who possess it the illusion that they control events. That, after all, is the reason why the story of King Canute retains, and will always retain, its relevance to the current political situation."

You can no more stop the development and use of technology than you can stop the moon pulling the sea and if you could the results would be catastrophic in both cases.

At the end of the day technology is agnostic to use, it's existance and development is based on the utility of force multipliers, something mankind is not going to ever give up. Which is a point many miss with the begining scenes of "2001".

Clive RobinsonJuly 19, 2020 3:58 AM

@ Anders,

Enigma had one fundamental flaw

As I noted towards the top of the thread it had more than "one" fundamental flaw...

The Enigma machine and how it was used is a good example of why "Complexity is not Security" for it's time it was a very complex mechanical device and it's rules of use likewise, but it was never secure.

Hugo Koch's original 1919 rotor machine from which the German Enigma was developed, was used in the Spanish Civil War and was regularly read by many other countries inteligence agencies. As far as we can tell it was broken early on by the French using the "Method of batons" which is very basically using strips of paper and moving them around (most non mathmatical ciphers can be carried out with strips of paper, bits of card, paper clips, and elastic bands).

The Germans made several changes but... were they of any use? The addition of the plug board and later hour box was realy pointless complexity as all it realy did was add the equivalent of a simple substitution cipher. At best we can say it got added as a result of mistaken thinking based on then common ways of cryptanalysis.

cliveJuly 19, 2020 4:44 AM

@Clive

Yes that's all fair comment and something I'm mindful of. I still think you're leaning a bit too heavily on the thought experiment and not enough on mitigations and countermeasures and necessity. Also for an extreme problem there exist a number of extreme solutions I haven't noticed you actively consider.

I don't think anyone would be in a hurry to stop a bumbling dufus like me carrying anything electronic which is a house of cards made out of swiss cheese at best. So what are they going to do? Feel me up? Laugh at me? WOW. Lock me up while I joke it's like a three star hotel without a television? And if I landed in a basement with meathooks and a hosepipe I'm probably dead anyway. The thing is like you say there is the bigger politics and agents of despair are human too. In my experience they can't wait to get rid of me.

AndersJuly 19, 2020 4:13 PM

@Clive

YES.

I wanted to point out that reflector is tied to that
"not encoding into itself" problem, it's not easy to
comprehend to those who don't know Enigma's inner working
down to bits.

One problem was German punctuality and prediction.
Brits bombed out their light buoys and Germans right away
sended out encrypted messages containing "ERLOSCHENISTLEUCHTTONNE".

When war reached to the end Germans stopped to use encryption
at all - code sheets were not distributed, trained persons
were killed, inexperienced didn't got yet training etc...
Radio traffic was mainly unencrypted.

ps. in soviet army i repaired Fialkas among other things
when serving my mandatory time :)

AndersJuly 20, 2020 2:00 PM

@Arduino Enigma

If you want to see some nice inner pictures:

www.nf6x.net/2010/04/soviet-m-125-3xx-fialka-cipher-machine/

Adam LauraJuly 25, 2020 1:15 AM

Dr Ajayi have be a great doctor even know in the world, am here today tell the world my testimony about how i was cured from herpes i was having this deadly disease called herpes in my body for the past 4 years i but now i know longer have it again. i never known that this disease have a cure not until i meet someone know to the world called Dr Ajayi he cured me from herpes he have been great to me and will also be great to you too. i have work with other doctor but nothing come out. one day i did a research and came across the testimony of a lady that also have same disease and got cured by Dr Ajayi. then i contacted his email and told him my problem he told me not to worry that he have the cure i didn't believe him when he said so because my doctor told me there is no cure he told me i need to get a herbal medicine for the cure which i did and now am totally cured you can also be cure too if only you contact him via email ajayiherbalhome@gmail.com. you can also call or whatsapp his mobile number +234 815 472 4883.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Sidebar photo of Bruce Schneier by Joe MacInnis.