Schneier on Security

Clive Robinson • March 19, 2021 10:48 PM

@ SpaceLifeForm, Bruce, ALL,

have found two undocumented x86 instructions in Intel CPUs which completely control microarchitectural state (yes, they can modify microcode)

Oh well it was to be expected eventually.

What was in theory, now anyone who can,

1, Get information on to your computer.
2, Can get the information run as code.
3, Owns the microcode interpreter in the CPU.
4, Can do anything the wiring between the registers alows.
5, Including those hidden test harnesses.

So those,”write only registers” you find tucked away in crypto co-pros etc now become read/write. I also hear that Intel are using RC4 to encrypt the info blocks not AES… In effect code signing on the very cheap.

So Major embarrassment for Intel, yet again.

What do you think is going to happen to their stock price this time 😉

But from a practical point of view I guess the cold hard reality is security of the average computer is not going to be realy effected by this.

And even those few computers Level III attackers are going to go after are not realy any worse off with respect to “being attacked”… But those writing ISA or above code to look for such microcode “root code” are going to have a much harder time of it.

I suspect the real winners if Intel do not close this problem out, are going to be games and specialised software writers who will be able to get algorithms one step closer to the hardware, but there will be all sorts of thermal managment issues involved.

Clive Robinson • March 20, 2021 6:18 PM

@ SpaceLifeForm,

RF. Signal. Noise.

Too little information given to say for certain, but… It’s almost certainly “Signal + Plan”.

To understand why you need to look at the broadcast spectrum of TV channels. Even modern digital ones have “holes in the spectrum” you can use.

The signal of choice to use oroginally is “Direct Sequence Spread Spectrum” (DSSS) due to “chipping/coding gain”.

In essence you create an artificial very broad band “noise signal” and “bi-phase” modulate it with your digital data[1]. The “coding gain is more or less the clock rate of the artificial noise signal divided by the data bit rate. Using modern codecs that are open source (see OGG) you can get very low data rates with inteligable audio so getting a coding gain of a thousand is not that difficult.

However what people forget is that the coding gain is reciprocal in nature. At the receiver the artificial noise signal would spread the TV signal by the same coding gain thus diminishing it and improving the wanted Signal to Noise ratio (Look up “Low Probability of Intercept” (LPI) systems).

The upshot is you can tuck a DSSS comms system under a TV channel and have it virtually undetectable by the majority of people watching TV broadcast in that channel. It’s a trick I used to use when sending “unauthorized” signals through satellite TV transponders back last century. You can still do this with little difficulty using US Navy geo-stationary satellites which is done by Brazilian and other South American criminals some of whom are very much into the drugs trade. Though they tend not to use DSSS systems for various reasons.

But using this spectrum this way in the TV broadcast bands is not of necesity “Gypsies, Tramps and Thieves” it’s actually approved and licenced in some parts of the world[2].

[1] You can do this at home with a few TTL gates. First ceate your noise signal with a Linear Feedback Shift Register”(LFSR) using clocked latches and a couple of XOR gates. Take this signal to one input of an XOR gate and apply your digital signal to the other gate. The output is a bi-phase modulated squarewave. You can clean it up and use it to drive a diode ring double balanced mixer with an RF carrier and that’s your DSSS transmitter…

[2] You might have heard of “TV White Space Broadcasting” which is the modern version of how to share Broadcast TV spectrum, much of which is empty. It’s been going on since the 1970’s one way or another. The original use of DSSS was to get around the problem of diferent parys of the spectrum being used or not in any given area. Back in the 1970’s and early 1980’s it was felt that using “guard channels” entirely would be to difficult. However with modern computer chips and Digital TV Broadcasting (DTB) the systems can just use the local “Matrix Channel” to find out what frequencies are and are not in use so using guard channels exclusively is now not a problem.

https://www.bbc.co.uk/rd/projects/tv-white-space-devices

Clive Robinson • March 21, 2021 9:44 AM

@ name.withheld…,

What gives?

Living in a different time zone at the moment…

Somethings it makes easier, others tiring.

As for AMSAT and Co, yes they’ve done a fair bit over the years, the trouble is that in certain parts “politics with a capital P” have taken over. Both the ARRL and RSGB saw declining membership prior to COVID and for some involved, a few moments of ego food was the price for letting those they represented down and doing things that are going to cause lasting harm, to the benifit of others in various undesirable industries who would rather “slash and burn” in others territory rather than clean up the mess they have created in their own backyards.

Any way the less said the lower my blood preasure 😉

Clive Robinson • March 22, 2021 1:57 AM

@ name.withheld…, ALL,

What gives?

Courtesy of the not so “minnie me” that is my son, a You-Tube link on Amateur / Ham “transponder” satellites

https://m.youtube.com/watch?v=sjKcID_R8bk

It has some history and science as well as a bit of communications theory, but nothing you can not do easily with a calculator or pencil and paper…

If folks want to know the more indepth knowledge of how frequency mixers work well…

All you are doing is using a nonlinear curve function such as y = x^2 or y = K(e^x -1) to multiply two sinewaves together to get four output frequencies. Most of the readers here I suspect remember trig identities from school without much fondness 😉

So… Rather than have me go through the derivation of sinA.sinB = (cos(A–B)-cos(A+B))/2

I’ll let a Wiki page do the lifting,

https://en.m.wikipedia.org/wiki/Frequency_mixer

(but also note the current in a diode is (e^VD -1) and that e and logs are strongly related and adding the logs of two numbers is the equivalent of multiplying which is why slide rules and log tables work and why diodes make reasonable “mixers”).

Clive Robinson • March 22, 2021 3:47 PM

@ JonKnowsNothing,

Hmm it’s worse than I thought, thanks for the update.

What makes me very nervous is thr fact the “blue one” gets to collect, keep and do as they see fit with every Cal Citizens medical data, I can see oh so many ways that is going to get abused by them.

With regsrds the “buyers regret” yes I’ve seen the news, the clips I’ve seen are shall we say “feral at best”, and that is going to “go home USA” North, South, West, and East taking four of the VoC’s right along with them. Not so much Spring Break, but Easter Wake. It was bad enough with that motorbike rally last year, this “Fun in Florida” weekend has all the potential of “Super Spreader Event of the Decade” if not century.

Clive Robinson • March 23, 2021 11:46 AM

@ name.withheld…,

I’d thought you’d have a variac in your power stream, precision variable control with a bit of power filtering (at least inductive)

Old school reliability but not enough inductance to close the bandwidth down enough. But also there is the IL EMP problem, in that variacs get slagged by lightning strikes. So they are not going to survive Solar Weather events, or god forbid some idiot with a modern Otto Marx generator. Or sometimes worse a HERF gun using either microwave generation (think oven valve/tube pulsed at a MW) or a dialectric transmission line and fast switch (remember those gas discharge avalanche switches that fire at upwards of twenty thousand volts, or even the now easy to make UV triggered spark gaps).

A little protection device to make is a foot long “line” made by two one inch wide copper bus bars 1/5th of an inch apart mounted in a two inch steel pipe or aluminium box section with lightning discharge tubes along the length and “transorbs” (semiconductor clamps) at the far end. You use it as a fast rise time clamp, as long as the immediate “load” is low impedence (drum of RG213 coax is ~1pf/cm at 2.5kV) you can get away with dumping very high power short duration (sub 1nS) pulses and time stretching the energy via reflections from load mismatch at much longer time periods. The trick is to eventually end up with a horribly over damped LPF that gives time for other much slower protection to kick in.

Whilst doing this for a “cage” with a few hundred Watts of computers or communications gear is not overly difficult or expensive, a whole house is a rather more interesting problem especially if the feeder voltage is low thus current high.

Clive Robinson • March 23, 2021 6:57 PM

@ vas pup,

Yeah, liberal arts are more preferable unfortunately in US in particular

The reason is twofold,

1, For most jobs a “degree” is a HR requirment for hiring. Mostly the subject studied is irrelevant.

2, For many students it’s a question of risk of debt and chance of qualification.

So the aim is not “a degree in XXX” just “a degree” to get a job. The cost of a degree is not uniform in all places but even where it is you incure the debt regardless of if you qualify or not. So a sensible person goes for the lowest cost degree that is easy to pass.

So “liberal arts” is going to be high on the “low risk” chart.

One of the hardest degrees to get, harder even than straight hard science, is “engineering” and other “proffession” degrees. The risk of debt is actually higher with engineering and worse unlike the traditional proffessions the fiscal reward for engineers is generally a lot lot lower. Oh and proffessional indemnity insurance can be higher. Because generally a Surgeon can only kill one patient at a time, an engineer can kill several hundred if a plane drops out of the sky. More if it plows ibto a sky scraper. Then what about an off shore drilling rig turns turtle with a whole heap ot environmental damage to follow. Oh and just how many people get on cruise ships these days? One of those turning turtle or breaking it’s back mid ocean in a storm is a high risk no matter how many life boats and rafts.

If you want people in “technology and engineering” you need not just change the risk/reward on the qualifications, you also need to make the work have the same soet of rewards in professional remuneration.

Otherwise those doing STEM will become “quants” or equivalent in the finance industry where they can earn ten times as much as they can doing what they trained to do. Or almost as usless chase work in Silicon Valley at the likes of Google, Facebook, etc writing more or less meaningless software to grab more PII and package it up. A good mind with algorithms and some managment skills can pull, +$1,000,000 rewards. Look up what an engineer designing medical electronics or FMCE gets as a comparison.

Unfortunately quite a few engineers see their jobs as more than a vocation but an entire reason to exist. Which unfortunately makes them easy targets for those that are unscrupulous or see them as being of less value than the “blue collar” or “oily rag” type “pump jockeys” that put gas in the tank and give the windscreen a rub down.

Similar is now happening in academia as well where almost the least important part of their job is doing research. As for teaching increasingly it’s being done by TA’s that in reality are other students a year or two ahead trying to earn money to pay the bills.

Those who work in STEM related employment are increasingly seen as “unimportant” in the neo-liberal view point. Even though they are the people that make it possible for there to be a society above third world subsistance farming.

Clive Robinson • March 24, 2021 4:29 AM

@ SpaceLifeForm, Bruce, ALL,

Evergreen performance problem.

Has potentially turned into a major supply chain failure, that could cause hundreds of ships to be delayed and potentially some cargos lost. Though it’s mainly oil that moves through the Suez canal currently in tankers almost as big.

Whilst there is a some amount of slack in international shipping to alow for bad weather, it can quickly be used up. So after a relatively short time of a week or less the delays caused by this incident will like the ripples on a pond from a stone thrown in the middle, spread out and continue for quite some time after the blockage is cleared. But, as cargos will stack up on docksides and backwards into road, rail, and other infrastructure in many countries there is one heck of a lot of money at risk in what is a fairly cut throat business (as I’ve mentioned before, some things like oil pipe lines can not just have the tap turned off as they will cool and start to become quite viscous).

On the plus side COVID has reduced shipping volumes of late as countries in lock down are not oroducing trade goods (but they are still importing).

However when you look at the problem, it’s one of those things that looks easy on paper (or in this case electronic chart). But that’s rarely the case with groundings.

As to why it happened, my first guess would be around the minor problem that to maintain stearage way such a large vessel needs to move above a certain minimum water speed (not ground track speed). Get below that speed and things become unstable, and a fully laden ship that size has quite a bit of sail area and inertia to contend with[1]. It’s why in ports large vessels are moved around by tugs that pull from the bows and hold back from the stern etc, as even minor changes in wind direction or force can have big issues.

Also the bottom of the canal is not exactly going to be flat so as the bow and stern aproach the sides they are going to ride up, over, and dig in. Which will make it almost like a latching motion. If the bottom is silt/mud, as the vessel settles water gets squeased out and as anyone who has had to contend with two sheets of glass that have got water in between the level of suction can quickly get to the point where breaking things is likely to happen.

If memory serves correctly the navigation is a pilot way, which means that a pilot who is very experienced in navigating the canal from the “port authority” is brought on board and is effectively in charge and any orders they give should be obayed by the helmsman…

Unless a clear cause as to what has happened is shown such as mechanical failure or unexpected adverse weather or similar, then “things will get legal real fast”… And probably quite political as well…

In a normal “grounding” which this is not, you generally wait for a change of tide, and move or remove cargo over the part of the vessel that is grounded. However movment of any cargo needs to be done with considerable care. As big though the vessel is, it’s not realy that strong and the wrong strain could permanently damage the vessel if not cause it to buckle and “spring plates” and start to take on water.

All though the canal is tidal I can not remember what the range is but I don’t think it’s that large. We are just past the first quater (neep tides) and approaching full moon (spring tides). Which means the high tide hights are increasing and will be highest around Tue next week.

Lets put it this way, I’m glad I’m not the one who has had this land on their door step it’s going to take a lot more than “shovels and brooms” to clean the mess up.

[1] There are other interesting effects to contend with as well, caused by hydrodynamics. One of which is when a vessel is in a flow of water as it gets close to another ship or dockside it can be “sucked” towards it, which makes obtaining stearage way harder.

Clive Robinson • March 24, 2021 5:24 AM

@ ALL,

I’m hearing news that several Danish health agencies think we are giving the COVID injections wrong, and that this may be the cause of the extreamly rare blood clots.

So whilst they investigate further they are recommending a change of procedure, which will slow things down a bit.

As a general rule vaccines are injected onto muscle not blood vessels. In the past when volumes of injections were larger, the recomnended proceadure was to “aspirate”.

Put simply when you’ve drawn up the vaccine, you push the needle into a suitable muscle (which used to be your back side for those a little older than others). Importantly even though the odds of having the needle in a blood vessel is low, you pull back on the syringe plunger. If blood appears in the syring you resite the needle to ensure it’s in muscle.

However for various reasons the pull back on the plunger appears to have stopped being taught some time after the 1980’s (when I was taught how to give people the needle when wearing the green 😉

So yes there are reasons why vaccines should not go into blood vessels and thrombotic events are one of them.

If that is the cause or not nobody knows, but it would appear to be a wise precaution anyway.

Other concerns have already been voiced by others about people being injected to low on the deltoid muscle thus hitting it where it is thin or not hitting it at all. A part of this is down to the clothes people wear. Rolling up a longish or tight sleeve is going to make the job a lot harder for those giving the injections to do it right…

Clive Robinson • March 24, 2021 12:01 PM

@ Weather, Winter, ALL,

I’m sorry certain “not suitable for work” behaviours are still going on.

The intent appears not to be at individuals as such, but at anyone, as it gives the attacker the excuse for their apparently childish and idiotic behaviour.

I suspect the real target as such, is by making this blog unusable to many, getting at the blog owner our host @Bruce.

What the motivations for that are could be many / any / none but mostly not comprehensable to anyone of normal sensabilities or reason. So look on the behaviours as in effect the childish scrawl of bad graffiti without purpose.

In effect behaving as a scavergibg mange ridden mutt[1] widdling up against every lamp post if for no better reason than to advertise “An unhouse broken idiot has been this way and left it’s malodorous outpourings”.

The sad thing though when you think about it, is that they need some “automated assistance” to appear to be anything other than a “spoiled brat” having an endless “terrible twos hissy fit”…

[1] https://en.m.wikipedia.org/wiki/File:Street_dog-Gianyar_Bali-2009.jpeg

Clive Robinson • March 24, 2021 6:50 PM

@ SpaceLifeForm, ALL,

Should be a rule that length of ship can not be greater than two thirds of width of canal.

Desirable as that might be, I can tell you it’s unlikely to happen.

The resason is the upgrade to the canal which apparently cost around 8billion dollars equivalent, needs to be paid back from earnings.

That might have started happening pre-COVOD, but the demand for oil which is one of the primary cargos has dropped significantly, as have other cargos.

It’s hard to put figures together from disparate reports but as I inderstand it they need between a hundred and fifty and two hundredd large vessels a day each way and they are barely managing fifty before this event.

It’s been reported in the press that the cause was a strong wind… Yes the vessel has a very large sail area on the beam as I’ve mentioned, but the vessel would also have to have been moving dangerously slowly or slower not to have had sufficient stearage way to correct for it.

The other thing we need to know is at what state of the tide the accident happened. As far as I can see for Suez current tide tables it would have happened just after the neep tides but I don’t have sufficient information to apply the twelfths rule and multiply it out to find the grounding hight when it happened. Hopefully it happened shortly after the low tide point thus the stress on the ships hull will be a lot less.

But every day that ships are not moving through the canal the more effective debt is being built up…

Clive Robinson • March 25, 2021 11:35 AM

@ Winter,

You need the S plane and Laplace transform to calculate how digital filters result in real signals that you can hear, see and measure.

Back in the 1960’s and through the 80’s a very large chunk of an engineering degree, was the use of the S-Plane and the Laplace transforms. Because they were such usefull short cuts.

But the truth is just like log tables you do not need them to design electronic circuits, machines or systems. They belong in the “Shutup and calculate” sub group with almost as much powere as the Teamsters Union used to have. It’s something the quantum physics community has fallen into. The basic fallacy is that somehow they will improve “insight”… Realy they do not, I challenge anyone to get real insight by staring at an S-plane plot or a set of tables derived from them.

In fact one of the biggest use of them, to design filters and check circuit stability can be done by the use of tables that I’ve several books of.

If you want insight these days especially with digital filters, you’d be better off writing a program in Python and pushing the numbers out to something like GNU Plot or even Excell. You can then fiddle with the weightings in real time and see what effects you get.

Clive Robinson • March 26, 2021 4:20 AM

@ SpaceLifeForm, Weather, ALL,

Just say NO to NFS

The article is an interesting start but, incomplete, so a degree of guessing has to be used.

The design sounds like one I’ve used in various forms in the past, the big problem was the designer did not apparently understand the concept of “use immutable memory” for the boot-ROM.

Unlike USB thumb/card drives which use Flash-ROM, CD/DVD disks can be immutable, as can EPROMs as NetROMs on network cards.

As for NFS it’s more secure than TFTP used in BootP but less secure than other more modern protocols like the proverbial Stunnel etc through to TLS.

Many do not realise that Intel’s “Preboot eXecution Environment”(PXE) that still uses the older insecure protocols was back in 2015 added to and as far as I’m aware still forms part of the “Unified Extensible Firmware Interface”(UEFI) standard, found on most motherboards these days.

Thus I suspect the use of the USB memory device as a boot-ROM was used by someone with a little more knowledge to get past the use of PXE in UEFI. Though potentially they did not go far enough and stopped at NFS instead of going to TLS.

In short they may have followed an old recipie found in one or two books on how to build Linux “Clusters”, “Thin-Clients”, “Kiosks”, “Net-Appliances” or Embedded devices (which have become very popular again with the Raspberry Pi being so cost effective and available).

But what is far from clear is how the “root of trust” was established on each of the net-appliances. Mention is made of using “signed code” which implies that some kind of PKI was implemented but potentially at a point to late in the chain as the author only mentions it with respect to the final application code.

But as @Weather has noted,

3 months to mount a attack? They really wanted to get in.

Not just that, the article also mentions both alarms and deadman’s switches. This is a level III+ attacker and it’s fairly certain from the alarm that they have their tenticles in in other parts of the infreastructure in or around that of the authors client.

As I’ve noted in the past on this blog the ICTsec industry tends to think of attacker profiles as,

1, Opportunistic Criminals.
2, APT SigInt or other IC.

That is a sound bite “managment view” for the C suits. With the follow on argument of “The Criminals are low hanging fruit attackers and it’s just a matter of time before they get in, and APT you can do nothing about because of zero-days”.

The point every one appears to ignore is that the attackers form points in a spectrum and by focusing on Crims-v-IC is leaving most of that spectrum not looked at…

Potentially this may be a case of something on the far side of APT, that is as thoughtfully put to gether an attack as you would expect from the likes of GCHQ and other High end European SigInt agencies, who we know are upto things but just don’t get caught doing it, unlike other nations SigInt agencies.

So very definately worth a bowl of popcorn or two and the comfy chair to watch this one.

Clive Robinson • March 26, 2021 4:47 AM

@ SpaceLifeForm, JonKnowsNothing, MarkH, ALL,

Looking at dredging up to 20,000 Cubic Meters of sand (and rock).

Remember Egypt taught the Israelis a thing or two about “hydro-mining” when they used very powerfull fire hose systems to wash the Israeli troops out of their dug in positions.

Also that modern “dredging” is not “digging” where harbour and shipping channel “silt up” is concerned, they use water pumps.

As some may remember from the news not far back Egypt has spent a considerable sum of money widening parts of the canal, thus probably still have silt removing dredging platforms still around doing tidy up and stabalisation.

If they have moving 20,000 cubic meters of canal bed is not going to take very long.

I guess the real question is where they are going to remove it from and how to do it safely. Whilst the vessel is not broadside on to the canal there is still one heck of a lot of “wetted area” of the hull presented to any movment of water in the canal. Which as the canal is tidal means a fair old rate of water movment more than twice a day. So only about six hours of sufficiently “slack water” to work in unlike ordinary dredging that can go on 24×7.

Clive Robinson • March 26, 2021 10:29 AM

@ MarkH,

I wonder how long it would take the Authority to create a mile or two of bypass canal.

I’m not sure. But one thing that is most likely is that it will not be on the West bank, as there is a large village / small town there, and they won’t want that becoming an island as the bridges would be a significant issue, as would ferries.

But from memory the canal is a bit over a tenth of a mile wide at that point with a usable width of 120meters or so. It does not give an indication of the required navigation angles for ships 450m or more in length and 70m or more in the beam (which are what people are talking about moving up through (if the canals become wider).

But if you assume 5 degrees or less you can see just how long a bypass stretch would be. You would be looking at millions of cubic meters of soil to move.

As for getting dredgong equipment into the vessel, I kind of assumed they would do a main cut followed by steadily smaller cuts, thus using equipment that goes down in size as you get closer to the vessel.

But what ever they do to the canal it’s self, I think they will still want to “lighten the load” at some point if possible, as it potebtially gives not just the greatest gain for effort, if done properly it’s also the least risky up to a point.

But… Unloading is kind of “moving the problem” rather than solving it, as at some point people are going to want their cargo. So anything “off loaded” will have to be “loaded up” again somewhere. Which is why the fuel and balast water would prove more attractive, even though it will increase the risk.

Clive Robinson • March 26, 2021 5:01 PM

@ JonKnowsNothing,

The fun part is the ramp. Egypt’s Pharaohs knew all about ramps and loading heavy goods on wood barges. The blueprints are painted and etched on the walls.

Fun historical note, somewhere in NASA’s archives is an informal paper I wrote about the fact that the Egyptians did not have the capstan or windlass nor do they appear to have ever used the pully as a way to gain lifting advantage. Oh and the mathmatics says they could not have built ramps beyond a certain size because sand behaves like a liquid under pressure (as far as I’m aware nobody has yet solved the issue of “rocks rising through sand” when you agitate the mixture).

It was back in the mid 90’s a UK Prof was asked to look into how the Egyptians “project managed” the building of the great pyramids.

As part of his talk he mentioned the lack of ability to get the stone blocks up due to not having what we recognise as lifting gear. He and I ended up having a chat over a coffee and I casually mentioned that they did not need a series of pullies as a block and tackle or a windlass or capstan to get a ten to one or more mechanical advantage. He looked some what dubious especially when I mentioned that Pythagoras would have realised how to do it.

So he asked me how… To which I replied four lift ropes four binding posts and a set of pull ropes on each lift rope. The speed with which I answered made him realise I had actuall knowledge even though he still looked puzzeled. So he asked me how it would work…

The answer is actually simple and “bleeding obvious” once you have been told 😉

The four lift ropes are used as two pairs of ropes. When you pull a pair appart by a short distance as they can not shorten the weight suspended has to go up a small amount. This makes the other pair slack and you wrap them around the binding posts. Now you pull that pair appart and the weight rises a little more and you wrap the now slack first pair around their binding posts. You just continue inching the weight up that way.

You can work out the mechanical advantage by assuming two back to back 90 degree triangles with the rope forming the two hypotenues. As H^2 = A^2 + B^2 where A is the horizontal distance you pull out to form the triangle thus B must get shorter B = sqr(H^2 – A^2) you can see that if H is 50 and A is 2 then B goes from 50 to 49.96 or shortens by 0.04. So for a movment of 2 you’ve lifted 0.04 or a mechanical advantage of 50 on each triangle or 25 overall.

He asked me to write it up and I stuffed in some stuff that my mother who was a historian had done on Egyption models of sailing vessels found in a tomb and he forwarded it on with his stuff to NASA.

Obviously for some medium weight items using ropes in such a way is going to be a lot faster than using a capstan or windlass and not far off of that of a block and tackle pully system.

Now I have no idea if the Egyptions did this or not, but that is not the point. What it proved is that there were other ways to get mechanical advantage that had not been considered by modern historians.

As it turns out, you get an interesting mechanical advantage profile, which I’ll let others consider for other things seeing as I will probably get a “yellow card” from the Moderator if I say 😉

Clive Robinson • March 27, 2021 4:08 AM

@ JonKnowsNothing,

… a common problem in California with fencing. We use T-Posts for fencing, metal posts driven into the ground and the wire tied to the post.

Funny that you should mention that, a friend of mine owns a company that makes those “fencing robots” that automate the putting up process.

He asked me to look into solving a particular problem (not the one you mention) that happens with all such systems… Let’s just say watch this space for an improved system.

It kind of brings new meaning to “field trials” 😉

Clive Robinson • March 27, 2021 8:05 PM

@ markky,

the Saipem 3000 is one of the ships stuck waiting to enter the Suez canal from the south and not at all that far from the site.

Looks like the old Sun God still shines down on the Kingdom of the Pharaohs 😉

Assuming of course, the crane can work across the stricken vessel and more importantly some one is prepared to draw up a contract that some one else is prepared to sign.

As for the 7000 I saw it years ago, it must have been the late 1980’s back when it was new and doing pipe laying. It is enormous, and about 100m at the narrowest. I also remember it had a deepish draft and could only be moved at some ridiculously slow speed like 5cm/sec or 180m/h, not sure if that was the actual top speed it could be moved at or the speed it laid pipe at. The last I heard of it, it had stoped doing pipeline work and was now an off shore factory floor for assembling wind turbines for off shore wind farms.

In size comparison from memory it’s about four times the area of the SpaceX barges and many times taller, lets just say I’ve sailed around smaller islands, that had people living on them.