Schneier on Security

Clive Robinson • January 20, 2023 7:46 PM

@ Bruce, ALL,

This might amuse,

“What Not To Write On Your Security Clearance Form”

http://milk.com/wall-o-shame/security_clearance.html

(note the date of 1988 oh and take a pinch of salt 😉

Whilst I’ve never had the authorities chase me in this way, my mis spent youth has had more than sufficient scrapes one way or another.

My father once told me,

“If you get yourself into trouble you can not talk your way out of, then you deserve what you get!”

Admittedly it was mid 1970’s in times when the authorities were somewhat more honest in their approach[1].

I’ve had a few bits of fun, one involving a 55gallon drum full of fertilizer and a handfull or three of coal dust in the middle of a farmers field… The resulting “duck pond” is quite impressive… Doing that today would probably get you an orange jump suit and an indefinate stay holiday at the arse end of Cuba[2] or worse…

And that as they say is not the least of it… My involvment almost pre-teen in “Broadcast Radio” and being in effect a “Pirate” caused me to meet a lot of people before they became famous more legitimately was fun but that sort of “lime light” never was of interest, as they say “hup narf” I was “doin it frer t’ crac”[3]

Oddly great fun though it all was, when I went through the security clearence process, it never came up… The big problem was when getting put through the process again the issuers of clearences knowing what level I held were puzzled by the fact another organisation wanted me cleared at a much lower level[4]…

Which should tell you something about the way bureaucracy combined with securiry DOES NOT work…

Any way, as they say,

“Plenty more stories where those came from…”.

[1] I used to think it was at the end of the 1980’s when things turned dishonest… But in more recent times now the Old Witch is dead, and thirty years have gone by, I was informed after enquiry that Mrs Thatcher when UK Prime Minster and effective Head of State had called for me to be setup to be arrested and discredited thus neutralized. Why because I had the temerity to point out grave security faults in a system called “BT-Gold”. It had caused much embarrassement by being hacked on the BBC Micro-Live show estimated to have been seen by about half the UK viewers back then… Most embarrassing because Maggie was trying to flog of BT to the UK public and she wanted nothing to bring the share price down… So my writing up further bad news, could not be alowed to happen… Luckily it was my own bl@@dy mindedness that foiled the plan. I got a request to go and demonstrate, which I thought was “odd” what I’d written up was both accurate and easy to follow as well as clearly showing BT had been less than honest with the UK media and public. I said “NO” I’d better things to do. The request came back down another path, so I chatted to someone and they said if BT were that desperate ask for big money and noliability all signed and paid up front… They tried again with the “fame” aspect and I just said I’d get more from publishing so they did get back kind of offering money but not the noliability so by now I thought it was way too fishy, so said no. Unfortunately by then it had become “Yesterdays news” so the journalists I’d talked to were not interested (turns out in one case the editor had “warned off” the journo…). A very short while later the BT-Prestel and “Hack of HRH Philip’s” in box made the news and some friends of mine found out what Mrs Thatcher had in mind…)

[2] https://metro.co.uk/2017/11/03/where-and-what-is-gitmo-prison-7049500/

[3] https://en.m.wikipedia.org/wiki/Craic

[4] The first was in for a “daytime job” designing “green use” equipment, the second was for “wearing the green” part time, and needing to work on the same equipment… It also ment I had to go on a basic traing course on the equipment and got told off for “not paying attention”…

Clive Robinson • January 23, 2023 5:17 AM

@ Deanna, Winter,

As far as I’m aware no country is completely one set of measures, which is as it should be for the development of the mind.

“Customary measures” has several guises, and thus you will have a national measure for weight be it SI, Imperial, or derivation there of (see history of US Gallon[1] via wine measures and “Queen Anne’s Gallon”). But trade measures where the likes of the Troy for bullion and Carat for diamonds pearls and other gems. With the strange sounding apothecary measurments. With beneath our feet or atleast around them there is barlycorns still used in shoe sizes.

But there is also a hidden measures, that few realise, and because of avian flu one is becomming a rarity currently, which is the “hens egg”.

Whilst recipes are given in grams or ounces these days the “standard” by which they were developed was the weight and volume of a hens egg[2].

Many other measures were based on the likes of plant seeds etc, because they were relatively uniform where ever they grew, so became “standards”. See the apothecary measures for more on this.

However as recipies are the first “science” the hens egg holds an important but not well known place in our progress to a modern world.

[1] The US “sport” of “milk chugging” appart from being a significant danger to life (it’s based on a method of tourture known as the “water cure”) is a bit of a cheat. The ratio between US and Imperial gallons is often given as ~5/6 so it’s easier with the more deminutive US gallon. But… There is 128 US ounce in a US gallon and 160 ounce in the Imperial so you would expect the ratio to be 1.041666.. But is actually 1.0408423, the reasons are attempts at “standardisation” at various points in time.

https://en.m.wikipedia.org/wiki/Water_cure_(torture)

[2] Which as eggs vary in size means modern recipies that say a number of eggs then a weight of flour, fat, or water do not come out right… Re-calculate by adjusting a large egg to be equivalent of two ounces weigh your eggs and scale the other ingredients appropriately and your cooking will be more consistent. Similar but more complicated has to be done for milk, where fat, protien, and water have to be compensated for because recipies used to be based on “full cream milk” which you can not even buy these days as “raw milk”. Contrary to what they imply, fully skimed milk is not very good for you, and it’s actually a “waste product” that used to be thrown away or mixed in with other food waste to make an animal feed known as “slopps”…

Clive Robinson • January 24, 2023 3:15 AM

@ Winter,

Re : Exams for memory not knowledge.

“Back to multiple-choice exams?”

That would not help very much as that can be viewed as “just a change in user interface”.

Exams are actually not very usefull for most things in life as they do not test very much more than your ability to remember a whole bunch of “facts” or other information. And to a lesser extent rules to manipulate some of that information[1].

If you are a little better than exams and have the ability to think forward you will realise that in essence most of our exams are testing

Information + Rules

Which is the same as,

Data + algorithms

Which famously is more known as,

Programs = Data + algorithms

Need I say more on that?

It begs the question of what we are trying to achieve with exams, as they certainly are not indicative of being a “usefull employee” –what ever that might mean– according to employers confederations like the UK CBI (who’s main feature appears to be they “moan but don’t propose”).

It’s why as a prospective employer I was realy only interested in their “projects” rather than their grades. Most of all I wanted to know what had gone wrong –an indicator of the difficulty they had set themselves– and how they had gone about solving things –an indicator of investigatory ability– and what they would do differently and how –a good indicator of being able to design– which is what,

“We tool making primates are supposed to be best at.”

[1] It’s why “open book” exams are more interesting than “closed book” exams, because, the questions are broader. But arguably less fair as the number of books you are alowed to take in is limited. Which in this day and age of On-line repositories of near “full domain knowledge” size available through a smart device is archaic. It’s repeating all the same arguments about calculators, and nodoubt the side argument about programable devices that arose from that.

Clive Robinson • January 24, 2023 7:57 PM

@ SpaceLifeForm, ALL,

Re : AWS again.

Lets just say,

The cloud is proved insecure again.

Or,

“There is no on-line security for consumer or commetcial setups.”

Both are true, and the “insecurity” is not just from those labled as “hackers”. Or to put it another way,

“Are people daft enough to think the likes of Alphabet/Google, Amazon, Apple, Microsoft, and all other Cloud service providers are not looking at your data, for their benifit?[1]”

The simple fact is that even if you use so called “Secure Apps” or “Privacy Enhancing Technology”(PET) you are going to get “data raped” by your various “Service Providers” at the very least and everybody and his dog, at your “gateway” if you let them.

Encrypted packets are thanks to Google not secure they leak meta-data which is often more usefull than access to message content[2] its self.

The effort required to maintain on-line privacy is difficult because of the point-to-point nature of visably sending a packet of data. Having an ordinary “mid point” service such as some file server or “Bulletin Board System”(BBS) acting as an Asynchronous “write, store and read” service just moves the problem to monitoring the mid point in some way. The solution which nobody has done yet is to overlay, a secure network over the top of the Internet system of UDP/TCP/etc that I’ve mentioned in the past. And sorry for the Fanbois, Tor is just not secure never was and never can be in the way it was designed and still used (and no I’m not going to go and explain it all over yet again, go look back on previous blog posts).

But even if we do get a basic private network, there are other issues to consider, you can look up the information leakage involved with the DNS protocols. Now consider how that plays out for two mobile devices that always use a different IP address. It needs a secure and private “rendezvous protocol” which means no static or unique mid point at the very least. So we do not even have more than a rough idea how to make what is required.

The Internet was designed for “reliability” and to a certain extent “efficiency” to have high “availability” the flip side is these make security thus privacy very hard to get.

[1] By benifit I mean for “business gain” I don’t mean just income, or profit from repackaging and selling your data, I include all “benifit” including economic espionage that Microsoft appear to be doing with git-hub, kick-back to legislators and “big energy” which they all do. Then there is the Inteligence Services, Law Enforcment and anyone else who they might give access to in return for “consideration” of some kind.

[2] Consider a simple message of,

“Hi hope you are well, I’m OK”

Let’s take it at face value –ie there is no “code within”– it tells you just one bit of data which is that the sender is OK. However the meta-data of source and destination addresses and port numbers and times gives a lot more information including quite often geo-location and who is talking to who and how often. Over time this enables a very detailed behavioural pattern to not just be, stablished but raise alarms almost instantly this changes in any way, such as message frequency, tine of day, message size, and more. That is “Traffic Analysis” gives a lot of information without having to crack the encryption or “end run attack” a users PC or Smart phone/device.

Clive Robinson • January 24, 2023 9:38 PM

@ SpaceLifeForm, ALL,

Re : I missed this.

“Pretty sure you should stay from AWS or any org that uses AWS.”

And how about,

Microsoft and it downloading “Malware”[1] to your PC for marketing reasons?

Microsoft is going to download and run spyware on your computer without notification or asking your permission[1],

https://www.theregister.com/2023/01/21/microsoft_office_count_update/?td=keepreading

This is not just unlawful in many places it’s decidedly immoral, as their reason for gathering this and potentially other private data is not of tangible benefit to the operator / user of the computer, whilst having significant benifit to Microsoft.

[1] In the UK and other places, what Microsoft are planning to do is contrary to the law (CMA90[2]) and constitutes unlawful access without explicit authorisation. Even giving a “click through” dialog, would not comply with the access requirments as the user has to make an “informed decision” which Microsoft is quite deliberately denying the operator / user.

[2] UK “Computer Misuse Act 1990″(CMA90) see §1 Offenses,

https://www.legislation.gov.uk/ukpga/1990/18

Clive Robinson • January 25, 2023 7:51 AM

@ name.withheld…,

“If you remember my recalling of a little “accident” in a facility where testing of a HERF gun proved”

It’s a few years back, so lets just say my memory of the details is probably not as sharp as yours.

But I seem to remember you saying it had a near instant detrimental effect on the test equipment thus people were unaware of the field intensities they were standing in.

Clive Robinson • January 25, 2023 9:51 PM

@ name.withheld,

Re : It’s the rise that counts.

“At the time of testing, a target (d-dot probe mounted on a stand and test trailer) sat forward of the gun about 20 meters,”

Curious, I’m assuming the D-dot was differential thus passive, but they usually only have a connecting transmission line of 3-6ft max due to capacitance issues. Normally that would connect to a storage scope…

How were you protecting that if you were getting 1ft arcs on what I’m assuming was, atleas the -30db side of the “motor”…

A back of a napkin calculation suggests the “light” your eyes were seeing was down the radiation transport by one or three steps so there could have been a few “hard sun tans” or worse year or fives X-Ray dose limit…

“I am assuming a field of megawatts needed to make a remote spark gap device through the air.”

The most peak power I’ve pushed was into a spark gap air “UV laser” using twenty foot long coaxial “water capacitors”. With enough energy to vaporize a chunk out of a four inch nail (not the worlds largest fuse but certainly fun to watch from a safe distance).

There is fun to be had in high energy physics, till either the power bill or second head arives 😉

Clive Robinson • January 26, 2023 7:36 AM

@ ResercherZero,

Re : Putting your neck out.

“because you can’t shut the door with your neck in the way.”

Ever notice we say “head and shoulders above…” etc not head, neck, and shoulders. Also how do you define a door to microwaves in a cavity.

The neck and shoulders would be absorbers of the microwaves not too disimilar to the head on a first appoximation. For safety reasons most microwaves fire “across the cavity” not “along the cavity” anyone who has played with waveguide knows that a “cavity” does not need all four sides of the box.

So the head would still get most of the power from the magnetron, probably in the right ear canal region.

But also consider your reasoning to domestic gas ovens, that back in “town gas”(Carbon monoxide) days people all to frequently did kill them selves with by “putting their head in the oven”.

But back to the fun stuff, that US Naval Institute”(USNI) article by Admiral James A. Winnefeld, U.S. Navy (Retired). I suspect he either does not know what he is talking about, or does not technically proot read. Take,

“With all eight DF-21 reentry vehicles inside the HPM beamwidth, a single, short- duration, extremely high-power electromagnetic pulse is fired in their direction.”

As an example do you spot what’s badly wrong?

If might make bad 1950’s SiFi reading but it does not pass muster in the real world. Ironically he then goes on to talk about the deficiences of lasers, perhaps not realising that those deficiences also effect other Radient EM systems like microwave weapons… Even at very low EM frequencies you will hear the expression “cloud warmers” the fact is water vapour/dropplets of which you get a lot of at sea both reflects and absorbs EM energy depending on frequency and the water vapour and dropplet size and density. Whilst there are low absorbtion areas that’s where the radar and similar “threat detection” systems of the ships defence capabilities are… Anyone who has blinked after a photo flash or seen car headlight beams on a foggy or rainy night would know what the problem with that is, if they had cause to think about it.

But he quite rightly states from reading it elsewhere that,

“And lasers require significant dwell time on a target to achieve a kill and can only engage one target at a time.”

But he does not appear to understand the “why” of it and that it applies to microwave weapons every bit as much to lasers. For instance a laser can near instantly kill any sensitive optics in an inbound threat but that would need the threat to firstly have optics, and secondly be pointing towards the laser source and thirdly be sensitive in the frequency range of the laser. The same three issues apply to the “easy kill” idea people have been pushing about microwave weapons… The thing is it’s actually easier to protect against microwave radiation than it is a laser as the basic laws of physics and geometry indicate.

I could go on, but look at it another way, civilians are easy to disable with any old projectile weapon because they don’t take defensive measures. Back in the 1970’s the US army worked out it was taking them 10,000 rounds on average to kill just one viet-cong adversary, even though they did not wear any kind of body armour how many more rounds on average would that have multipled by if they had?

The reason consumer and commercial drones are somewhat susceptable to microwave weapons is they have no defences built in. But just remember two thinks,

1, You don’t need your eyes to walk toward a light, blind people can do it reliably when they know where the light is.
2, Whilst EM radiation can not go through a metal box unless it melts/disrupts it, gravity has no problem and inertia can be boxed in and work just as well.

The latest submarine navigation systems work that way very well, so it will work with drones and other air born threats just as well, all they have to do is jig-abit to not have a predictable path for a kinetic response.

Oh there is one lesson to think about from 1950’s SiFi, “how do you stop a rock from space?”… Space craft on reentry survive way more power than any current EM weapon can generate and deliver…

This “EM Weapons” game is just another way to get your hands on Government Money by the bucket load. And if it does “get going” it will turn into another ECM, ECCM, ECCCM game with each progressive step being some significant power in cost increase.

The way to win the game is not to have a weakness to an opponents strength, and be novel / original, that way you get the same advantage terrorists do.

Clive Robinson • January 26, 2023 9:33 AM

@ SpaceLifeForm, ALL,

Re : Chasing crypto Ghosts

With regards the ARS artical, this actually made me laugh rather than just chuckle (which got me a couple of looks),

“That would certainly be catastrophic since someone might be able to use the capability to cryptographically sign malicious updates with a Microsoft or Apple key and distribute them to millions of people.”

As you know Microsoft are having a bit of a flap “because not everyone got the MD5 memo”

So we already know “catastrophic” that can be…

But one thing is very right and I’ve said it before on this blog several times in several ways. But now you can have a “Pro Stamp it” with,

“But maybe we shouldn’t be moving to just post-quantum algorithms. Maybe we should be using the post-quantum algorithms and RSA in parallel because there might be a problem with the post-quantum algorithms.”

But also both @Bruce and myself have commented about what academic and now comercial cryptographers in the “Open Community” are doing or rather not doing,

“Expert says the focus on quantum attacks may distract us from more immediate threats.”

It’s something we don’t talk about or take action on. But in all honesty do people realy think the SigInt agencies are only doing “Quantum Computing” on “Key agreement algorithms”?

No of course not, they have a much broader, more practical outlook when it comes to privacy invasion. It’s something the Open Community realy should take onboard.

But as I pointed out a few days ago these NIST competitions are about “fame” and “making a name” rather more so than just cranking out papers by the dozen. Which researchers hope will get translated into good money, job security and getting well paid speaking gigs etc.

But also there is “startup&VC” potential, as Simson Garfinkel put it,

“The second thing they are reasonably good at, but we don’t know for how much longer, is they’re reasonably good at getting funding.”

So all the researchers are busy jumping on the same wild bull rides down in El Passo, rather than covering the territory more widely.

But lets be honest, Quantum Computing is just one of a whole slew of potential attacks to “oblivious root of security sharing” that RSA etc do. What we realy need to be doing is “getting rid of the problem” as it’s a very fragile bone, rather than just covering up a symptom with yet another sticking plaster.

Clive Robinson • January 26, 2023 4:43 PM

@ ALL,

Re : Hive Ransomware Network Takedown,

This is from the UK “National Crime Agency”(NCA),

https://www.nationalcrimeagency.gov.uk/news/hive-takedown-nca-in-international-operation-to-shut-down-100m-ransomware-threat

Not very informative.

A little more just in from Auz,

https://www.abc.net.au/news/2023-01-27/live-news-blog-the-loop-ransomware-hottest-200-succession/101898164#live-blog-post-19508

Apparently an error in Hive’s control panel alowed access to LEO’s last july. Servers have been grabbed in Germany and Holland, but no arrests.

Not clear how or if this will stop Hive, or stop Hive members reforming. Apparently Hive were still active just a few days ago,

https://www.securityweek.com/hive-ransomware-operation-apparently-shut-down-by-law-enforcement/

Clive Robinson • January 27, 2023 12:48 AM

@ name.withheld…,

Re : Cat watching.

The UV-laser was part of a “pulse generator” for an even higher energy “physics experiment”.

It’s no secret that voltage multiplier ladder circuits using caps and diodes only do a few hundred thousand volts, and you have a choice, high voltage or high current the diodes just don’t do both.

So as you probably know the next ladder type uses “spark gaps” and can if they are designed right give you very high voltage and current. The problem is the wooly trigger threshold translates into not very good time.

Not as well known as it might be is that another trouble with spark gaps is they can trigger each other. That is one fires dumps out a UV flash that will trigger adjacent spark gaps.

Thus the idea was rather than try to stop the problem was turn it into a virtue.

So line the spark gaps up diagonally so there was a narrow line running down between them all and fire a UV laser down as the trigger.

That way you can get some really fast rise times with some even bigger water capaciters.

I was never told the “purpose” but we know lightning does not have anything like that sort of rise time.

So my guess 😉 would be something like an EMP generator or Z-Pinch device to generate lots of coherant X-Rays or even neutrons.

But a visit from “the end customer” was a face from the past. I’d worked with them on producing a plasma rocket engine (basically take the electron gun from a cathode ray tube scale it up and leak a little nobal gas in, and you get a thrust that is still good at a sizable fraction of the speed of light). I now know NASA had been playing around using Z-Pinch for rocket engines to get considerably greater thrust using some more exotic gases (which if pulsed in a cavity…). Imagine just what sort of acceleration dumping out a ton of gas at 90%C over say a calander month would give you as a resultant speed on a capsule not much bigger than used to sit ontop of the Apollo rocket…

Clive Robinson • January 28, 2023 11:30 AM

@ Nick Levinson,

Re : Bulldog the man.

“but is not head, neck, and shoulders above it has a serious medical problem.”

+1 😉

Mind you, in films and such like, a man[1] such as that would be a body guard called Knuckles, Masher or similar…

Oh and “serious medical problem” but only with “Head and Shoulders” that might please other shampoo manufacturers 😉

[1] Let’s not go there with jokes about female Russian weight lifters, as Billy Connelly has done a couple of times. As far as I’m aware I’ve only known one and that was back in 98, and she was also a stunt actress / coordinator and spent a lot of time flying around the world. We actually met at a party thrown for her by her half sister who was doing a PhD, and I’was kind of going out with at the time, and they were both quite cute as well as being realy smart. It was one heck of a party and I got into a drinking contest with a Russian actor which only ended because we drank the bottle dry. The following day was somewhat painful to put it mildly and I realy could have done without a head with the aches it had…

Clive Robinson • January 28, 2023 7:04 PM

@ name.withheld,

“Yes, have to appear inductive in a capacitive frame is always the first order problem.”

Without giving away any secrets,

If it stays a first order problem with fixed load impedence the simple trick is just turn it into a transmission line.

But with voltage multiplier ladders that is not sufficient because the load changes dynamically and as the voltage rises the transmission line can easily break down. So the next best thing is say stuff-it for the charge cycle and go for a matched transmission line on discharge, because that is where you want the fast and time stable edge.

Only you cann’t do it because the transmission line won’t support the voltage at low load impedences.

One trick is to use Wilkinson dividers to impedence match, and after a little thought you realise you can make a balanced binary tree of capacitors with Wilkinson lines and capacitors at the nodes. The two problems are, as you work back from load to source you realise you can not in anyway use stub matching as that is a time delay and does no matching on a rising edge. So what is normally “the best tool in the box” is found wanting. And the Wilkinson circuit is normally fairly narrow in frequency[1]. Thus you have to jump a little, as you know a linearly tapered line can also power match and is more broadband. But linear tapering whilst easy to model is actually not as good as a logrithmic taper

I suspect at this point you can see where it’s heading[2]…

A little over half a decade ago, people started getting interested in “RF Power Harvesting” and hit up against similar problems (if you think about it they are trying to use the same circuit but in reverse).

Thus if you look around you will find papers like,

https://pure.hw.ac.uk/ws/portalfiles/portal/15450742/uwb_rectifier_tmtt.pdf

But you will see they are still missing a trick or two.

[1] Another problem is RF does fly off at corners the usual 45degree chamfer used on microstrip does not always “travel well” to other transmission line types.

[2] For those wanting more up todate tricks with WPD’s and some interesting refrences about broadbabding them,

https://repository.londonmet.ac.uk/4582/1/manuscript.pdf