Posts by Electronics'R'Us

Ad blocking made Google throw its toys out of the pram – and now even more control is being taken from us

Re: Misses the point

Until then, how do you support the websites you rely on?

I have a 'premium' subscription to a national newspaper; this means I get the dead tree version and all the content of the web site.

The problem I have is that they still fling page upon page of ads at me even though I am paying hundreds of pounds a year for that subscription.

That website gets the Firefox + NoScript treatment (the only domains allowed are the ones that actually let me sign in).

I did an experiment to allow all the ads and scripts one day.

The result: the laptop (which is no slouch - dual core 2.4GHz 7th gen I7, 8GB RAM) got maxed out with almost 100% CPU utilisation- the fan started up almost immediately which is unusual unless I am running something such as a circuit simulation in LTSpice.

I took a look at all the domains running scripts and it was so long I had to scroll down three times to see all of them. It was also interesting to see that the scripts listed only on the home page is quite short (and virtually all blocked anyway now) so that means that a script from a third party domain is calling other scripts from other third party domains and so forth.

The other issue I have with this is quite simple: ads in the print edition are subject (quite reasonably) to some due diligence, but how do they do that with an ad flung from $DEITY knows where and scripts loaded on the fly from third party domains? My view is that it is impossible to effectively do.

An observation; running scripts directly from third party domains is stupid, in my view.

What does my neighbour's Tesla have in common with a stairlift?

Re: Battery EVs

...as long as the lithium is only dug up in places that nobody in the West cares about...

There is some exploration going on into the possibility of mining Lithium in Cornwall.

Re: Charging

There is no doubt that alternative fuel is getting a lot of investment and we will, no doubt, see other (perhaps much better) alternatives in the future. My engineering experience is that there is always a trade-off; that is as true at system level and power source level as it is for components.

Sometimes 'old' technology is actually the best fit for the job; a lot of satellites (especially communications) use a vacuum tube for the final power amplifier (a travelling wave tube usually).

As I like to remind those who think old ideas aren't valid; we have not yet repealed Ohm's law.

I am not opposed to EVs conceptually; I just cast my sceptical eye over the assertions without proof and wonder just what they are ingesting (must be expensive given some of the pronouncements we have seen). Certainly they have come a long way, but it has also taken a long time. EVs are evolutionary rather than revolutionary.

EVs can make a lot of sense for the correct use case; I just have doubts over the infrastructure. Governments rarely (ever?) actually operate for the long term vision (unless it is a political ego project like HS2 that is going to cost $Deity knows how much for a very limited return).

On that note, the proposed EV infrastructure is a large government project. What could possibly go wrong?

Re: Charging

I don't know what you're smoking but can I have some?

Indeed. There are a lot of issues with batteries. Here are just a few.

1. Self discharge. All batteries leak to a greater or lesser extent. Rechargeable batteries have pretty high rates (not incredibly high but it is still noticeable). That is just as true for the battery in my diesel vehicle but it has one primary duty - start the engine; it is not the primary source of motive power.

2. Charging efficiency. The closer you get to 100% charge, the less efficient the charging becomes. That is why most hybrids only charge up the battery to about 70% (this will vary depending on the specifics of the battery). Fast charging can also be detrimental due to excessive heat during the charging process.

3. Limited charging cycles. Industrial Li+ batteries certainly have more charging cycles than consumer kit but it is still limited. That means, somewhere down the line, the batteries will need to be replaced and the recycling chain for that doesn't appear to be in place yet.

The amount of extra fuel I use to recharge my vehicle battery still needs to be provided of course (but as long as I have fuel in the tank and a battery that can start it I can get going - not so with an all electric if the battery is severely depleted).

So how are these all electric vehicles going to get to this charging station when there may not be enough energy left in the battery to do so?

So called battery 'gas gauges' (there are such things - a better name would be a 'state of charge monitor') are notoriously inaccurate in these types of application until they have been 'trained' - that is because they are primarily doing coulomb counting and it needs a few cycles of the battery to adjust for a realistic measurement. When you see 50% charge many expect you can go the same distance again but this does not account for another inefficiency - power conversion. The batteries do not directly drive the power train but are run through converters which have their own inefficiencies which has a dependency on the input voltage from the batteries.

Power train power is an energy conversion issue, so as the battery voltage goes down, the battery output current goes up for a given power demand.

Electric vehicles do have (current ones anyway) regenerative braking to recover energy that would otherwise be lost to heat so that is a positive; this has been used for decades in some applications.

Lots to do yet, and I don't see it happening any time soon.

Court orders encrypted email biz Tutanota to build a backdoor in user's mailbox, founder says 'this is absurd'

Re: Dear Courts. No. Go away.

Let's do a perfectly possible scenario.

I encrypt a message (to paper) using a one time pad (the only provably secure form of encryption).

I now securely destroy that one time pad.

I put this message in an envelope and mail it to a recipient (who may not actually be the final recipient).

The authorities want to know what it says, but I never memorised the pad key (which is nigh on impossible anyway for any moderately large message and there are other things one can do to muddy the waters even if I could actually tell you the key I was working from - might be an intermediate key as is often done in electronic crypto).

The mail is intercepted and read, but it is, of course, gibberish without the recipient one time pad. There are ways of knowing when an envelope has been tampered with, incidentally.

Trying to get that key from a cutout is useless as they will not have the key but they may have been instructed to simply destroy the message if any sign of tampering is found.

So they have some options, but none that are attractive. In the UK they could demand the keys, of course, but let's put that aside for now.

Do tell what is actually illegal about that scenario.

Pure frustration: What happens when someone uses your email address to sign up for PayPal, car hire, doctors, security systems and more

Email addresses with numbers attached

When I was working for $BIG_PLC I had an internal email address with the suffix 9 (I have a relatively common name - at least it lets me hide in the phone book).

I lost count of the number of times vendors would send <item> to the person who did not have any suffix but the same name. We eventually set up an email group for everyone with the same name in the company so we could ask if such and such was meant for someone else with the same name.

Cops raid home of ousted data scientist who created her own Florida COVID-19 dashboard

Re: Probs just standard operating procedure

Many years ago (possibly before I actually went to the USA permanently so mid 70s) my then g/f worked in an area in Norfolk, VA well known for problems at night. (She was working in a bar opposite a 'massage parlour').

She told me that one of the local cops had turned down a promotion because he would have had to drive a desk and he 'enjoyed kicking ass'.

No change there, then.

Apple's M1: the fastest and bestest ever silicon = revolution? Nah, there's far more interesting stuff happening in tech that matters to everyone

Re: I fear that too much shiny is taking a toll on some people's attention span.

In really high performance computing, which is massively parallel, the speed of moving the data between nodes is a key performance metric, quite apart from the speed of the cores.

You can see the state of the art www.top500.org

There are a number of things involved (latency is a key issue) but in those environments there is a huge amount of data being shovelled around.

Running joke: That fitness gadget? It's, er, run out

Re: why?

We see you have been 'marking your territory' at least 6 times today and you were not in the vicinity of a proper facility.

Can we interest you in a real bathroom?

Calls for 'right to repair' electronics laws grow louder across Europe

Re: Just because...

I think the trend towards BGAs needs to be understood in the wider context and from a designer's perspective.

BGAs have a number of things going for them. They save physical space; for a given number of contacts, BGAs are the smallest solution. When in reflow (which is basically a multi-stage oven with different temperature zones) a BGA will actually pull itself onto the pads even if it was not precisely placed by the pick and place equipment. That cannot be said of non BGA packages which will be soldered where they are physically placed.

Contrary to what is widely believed, BGAs are really not that difficult to remove and replace; it requires slightly more specialised equipment but it does not cost thousands (unless you expect to be doing a lot of them when the cost can get quite high).

The use of security devices to ensure you can only buy the OEM parts is a blight on the world of electronics in many cases but absolutely necessary in others (you really wouldn't want anything safety critical to be using anything other than the part we know really works and that can include even a die revision which is something I have seen in a particular autopilot).

When it is used to force you into using the dealer network (especially for things we used to do ourselves) it is simply wrong and needs to be clamped down on.

When designing modern equipment, the form factor (size) is often a key for marketdroids so we are forced into very high density layouts which do require somewhat specialist equipment - try desoldering (or even seeing clearly!) a 01005 component which are now commonplace in smart phones and have been since around 2000 or so.

As to adding a mezzanine, that introduces a connector. In electronics, about 90% of faults are connectors.

Deliberately soldering down memory in a system that does not really require it is just taking the p*ss but there are occasions where that is a better solution.

There are companies out there that use lock in methods to extract as much money from people as possible and it is those I avoid.

Having a robust repairability index would be a very good start.

Who knew that hosing a table with copious amounts of cubic metres would trip adult filters?

Re: Funny placenames

As you take the A2 east bound from the Dartford Tunnel, you will see a sign for

CHALK

THONG

I was always amused at that one.

Near where I used to live (in Thanet) a particular sign was forever being stolen so it was eventually made into two separate markers.

It pointed to

HAM SANDWICH

It's always DNS, especially when a sysadmin makes a hash of their semicolons

Make

Some time ago (prior to the turn of the century) I was writing a lot of diagnostics for a product (using an SGI Indy - fun times)

One fine day, I checked out the latest branch to address a requested feature but when I tried to compile, I got many screens worth of errors.

Turns out that the previous checkout (by users unknown) had loaded the makefile into their editor which converted tabs to spaces.

In that version of make, tabs vs. spaces had meaning so I had to get an older version of the makefile and check it back in.

That made for an interesting morning.

EU says Boeing 737 Max won't fly over the Continent just yet: The US can make its own choices over pilot training

Re: Two angle-of-attach sensors not enough

AF447 had a triplex system.

In such a system, there are 3 independent computers with their own sensor feed and they vote every few 10s of milliseconds on the parameters they read.

For this to work properly, 2 computing lanes must agree.

AF447 lost at least 2 pitot tubes (and therefore 2 computing elements).

The telemetry showed this as one of the first messages was 'airspeed disagree' followed by 'alternate laws'.

This is a well known term in avionics (and in particular flight control systems).

The flight control system can no longer trust some inputs and the control laws are relaxed. With all 3 pitot tubes giving different readings there was no way for the flight control computer to actually fly the aircraft.

This is where the fact that the crew in the cockpit did not know how to fly became the issue.

Not sunshine, moonlight or good times – blame it on the buggy

Compiler verbosity

I was working with a bare metal system using good old GCC with -wall as the default setting and as happens with bare metal, many pointers are loaded into the internal registers of the microcontroller*.

So I was happily writing said code and at compile time there were loads of warnings about assigning a uint32_t to a pointer.

I went back through, meticulously prefixing them with (uint32_t *) to do the typecast and get rid of the noise and added comments:

/* typecast to silence GCC's verbal diarrhoea */

My manager looked and said 'You can't leave that in there!' (I did).

* Modern microcontrollers have so many damn peripherals internally that the init code is an eye glaze inducing list of pointer assignments into the various peripheral registers.

UX / UI

At the place I currently work, it was decided to introduce a new system for evaluations (which, in all fairness, seems to be working quite well).

A new in-house set of web forms was duly produced and clearly not given to average users to test.

The web form for recording these conversations (which is really what they are rather than the 'PDR' crap) had a bright rectangular piece of text on a contrasting background that gave no indication it might actually do anything (didn't highlight with a mouse over, cursor didn't change, no url in the status bar), but when I decided to check things out and click on said apparently inert object (the first conversations were the beta test effectively), I was whisked to a completely different page.

I had a bit of a vent about that one.

Rule 1: If a button can do something then make it clear with a visual clue.

Linux Foundation, IBM, Cisco and others back ‘Inclusive Naming Initiative’ to change nasty tech terms

Puritanism

I agree with you.

Here is a definition I heard many years ago.

Puritanism: The nagging fear that someone, somewhere, is enjoying themselves.

Re: What about non-English offensive words ?

Are we going to insist that countries or communities change their names as well?

Montenegro comes to mind along with some others, such as Isla Negra.

Will there have to be an alias in a database of countries to avoid these terrible things?

</sarcasm>

Re: Lefties

You left out the term gauche

Have another pint :)

Re: > Also acceptable: * Parent/child

Deleting text in EMACS has the perversely satisfying response:

140* characters killed.

* your number may vary :)

Re: What are we going to do about the embedded devices?

The term master / slave describes very accurately a specific type of interface used by literally tens of thousands of parts (I may be a bit on the low side there).

Any terminology that still accurately describes this interface will still be offensive to those who are determined to take offence at anything they personally deem 'offensive'.

Re: So basically we're going to have to re-name everything.

Hilarious.

What about language keywords and perhaps some mnemonics and even some mathematics terms.

<sarcasm>

typedef - defining a type? that's offensive!

#define - defining anything without our consent is offensive!

union - we are not in a union, we were consciously coupled (Gwyneth Paltrow version)

class - the struggle goes on comrade!

bool - you cannot restrict anything to just two possibilities!

short - offensive to the vertically challenged!

do - commanding something? that's as bad as that master thing!

switch - could be potentially offensive in two independent ways!

pure virtual function - who are you to define 'pure'?!

operator overloading - exploitation of the workers!

multiple inheritance - could be considered offensive in multiple independent ways!

My favourite power architecture mnemonic of all time

EIEIO - offensive caricature of farmers!

Mathematics

irrational number - we'll be the judge of what is rational or not!

and it's corollary

rational number

</sarcasm>

I wish people would get a grip; the context of language (to convey meaning - no shit) counts far more than the actual terms themselves.

Watchdog signals Boeing 737 Max jets can return to US skies following software upgrade, pilot training

Stable?

From other comments long ago, I seem to recall that the 737-NG could not power out of an incipient stall, unlike the predecessors which could (i.e. applying power in the older versions could solve the problem).

The nose had to be lowered first (second nature to glider pilots).

The MAX has an instability (and a critical non conformance to the part 25 rules) in a small part of the flight envelope which MCAS was 'designed' to handle. Thus a computer that was bolted on to an otherwise dumb aircraft that had full trim authority (that was not even properly documented in the engineering documents).

At full extents, the elevators have insufficient authority to overcome it because the elevators are tiny compared to the trim area.

Aircraft automation is great, but without a proper design process it is asking for trouble.

Let's be clear; MCAS is (was) an autonomous flight control system that had crap engineering all over it (I do not blame the engineers, I blame the management as they could override engineering concerns based on business issues, not engineering risk analysis).

Where an aircraft has been designed to be fly by wire from the outset (the B777 comes to mind as I am familiar with the system) and has been properly engineered the risks are incredibly low. (I am not referring to the latest version of the 777 here incidentally).

That was clearly never done for the MAX.

UK Prime Minister Boris Johnson pledges £12bn green economy package

Re: Not a chance !

I was assuming that's why they have a parallel plan to get people cycling and walking.

I assume that was a bit of sarcasm :)

On the fairly rare occasions I actually go into the company offices, neither of those are really options for me. If you have ever driven on the A38 in south east Cornwall you would understand why to say nothing of navigating the Tamar over to Plymouth - I would have to use the Torpoint Ferry and the road to Torpoint is 'interesting'.

I am pretty sure that an electric vehicle would get far less mileage on that road than on the (benign) test track too.

Re: Not a chance !

Another thing that comes to mind is population density.

Building out this infrastructure in sparsely populated areas (and there are a lot of them in the UK) is going to be far more expensive than it would be for a city. Interestingly, more and more people are moving away from cities but cities and rural areas alike have many challenges that seem to have escaped the politicians.

Where the city was laid out properly in a planned manner (Welwyn Garden City comes to mind) it may be possible to build out the base infrastructure (but that still leaves the detail of tiny parking areas as you note).

For older cities this will be a nightmare. I know many of those places quite well; Edinburgh would be really interesting, as would the older parts of London and as for Newcastle, let's not even go there.

Where I live (in the middle of a farm) the population density is really low, but being rural the need would be high for charging points at almost every dwelling and the cost of hooking that up to the (supposedly fully upgraded - yeah right) grid would be exorbitant.

When I say sparsely populated, it means that unlike in a fairly densely populated area, trying to use my postcode to find the house simply gives you a roughly 1 square mile zone of probability rather than to find the house number you need with all of them visible which can be interesting when we get new delivery drivers. Depending on which map service is used, my location comes up as either about 1/4 mile to the west or 1/2 mile to the east of the actual location.

I do not have gas to the house, but I do have an oil fired Rayburn (a type of Aga - it was here when I moved in as every installation of that type is bespoke); the fuel it uses is kerosene 28, and the primary users of kerosene are jet aircraft (which is good for me right now as total demand is way down and my oil is much cheaper). That also heats my hot water tank (there is an electrical heating element within it, but that is only turned on when we turn off the Rayburn and that has been for a grand total of 4 days over the last 2 years).

Trying to phase out that fuel would raise quite a ruckus (a lot of places in the southwest have this sort of arrangement as do many other places around the country).

The electric meter housing is not suitable for a smart meter (thankfully!). Would there be an incentive (discounts) on electricity to charge these vehicles? How would, or even could, that be implemented? (separate meters perhaps but I can see that being easy to subvert). If there are no incentives then will power theft suddenly go through the roof? Possibly.

So there are even more issues than you list (that is not a criticism in any way) and no easy answers.

I am taking all these announcements with a pretty large grain of salt right now as I cannot see how all this could be achieved in 20 years, let alone 9.

Max Schrems is back... and he's challenging Apple's 'secret iPhone advertising tracking cookies' in Europe

Re: They just don't get it.

"they should have the option to be able to OPT-IN"

Precisely.

I look at this from the perspective of a real set of shops and articles.

I can go to a library and read a few articles; now I head to a newsagent. How would you feel if they said to you "We know you have been reading <list of books at library> and we might suggest <x>".

Now the only way they could know this is if either the library told them or someone was watching my every move. User profiling is, by far, the latter.

If someone did that in real life (followed me around shops watching my every move) it would be classified as stalking.

On that subject, El Reg, it is really none of your business that I occasionally browse electronics component vendor sites (see name) so quit serving ads that show their products (or quit using the ad spamming service that does so). I deliberately keep those specific cookies for ease of login, not so an ad spamming service can sling an ad at me that is a total waste of money because I am going to ignore the ad as I get the newsletters anyway.

I have no problem with a site serving ads that have products that pertain to the site main contents; I have a real problem with a site serving ads based on what other sites I visit.

Of course, if sites were forced to adhere to that rule, a lot of people might actually have to get a real job.

Re: They just don't get it.

Firefox has a really great add-on.

Containers (with one specifically for facebook). The facebook container prevents scripts from facebook running and dropping cookies on you unless you are specifically on facebook so all those facebook share icons (which have scripts behind them) are effectively neutered.

The newer containers is a more general extension of that.

Worn-out NAND flash blamed for Tesla vehicle gremlins, such as rearview cam failures and silenced audio alerts

Re: Flash wear

One of the projects I was asked to take a look at in another division of $BigCompany was very early failure of compact flash cards (they use NAND flash).

The application was using (IIRC) MS SQL for transactions and the cards were failing within a month; turned out there were millions of transactions per week as the particular version of Windows was not particularly flash memory friendly (such as buffering up transactions for later commit).

Re: Flash wear

Interesting, but what is meant by data retention?

Data retention time is the amount of time that the device will read back the same data as you put in and yes, this is a form of erase / errors risk.

This is the reason that in all the critical projects I have been involved with two (sometimes 3) copies of critical data are programmed; the chance of all of them failing at the same time is tiny (infeasible as we say in avionics).

For flash, the problem is that an entire sector can suddenly go pop (even though there may only be a single bit in error).

The problem is worse in consumer grade MLC (multi-level cell) which encode more than one bit per cell with a range of voltages; what happens is that the sense voltage drifts over time and may not read back the same data because a particular cell set of thresholds have moved enough to change the interpreted read result.

SLC (single level cell) devices are common in critical applications but it is getting more difficult to source.

Re: Flash wear

Heavy writing to a flash device is not the standard use model, agreed.

Although that is one wear mechanism, the data retention time is heavily influenced by temperature.

Most devices state the data retention time for a device held at a constant 25C; increase that temperature and you get a dramatic drop in that (typically 1/2 for every 10C rise). The test for this is usually a 1000 hour test at an elevated temperature.

If these devices are close to a heat source (such as all the processors that are in these vehicles doing some pretty heavy grunt work) then the data retention time will not be anywhere close to the datasheet specification (the datasheet only guarantees data retention at a given temperature).

Panic in the mailroom: The perils of an operating system too smart for its own good

Impromptu tours

I was about 12 when I was coming home from school.

Just a short distance up the main road from where we lived was a Plessey plant (long since gone, sadly).

I was always interested in electronics so I walked in and asked questions. Had quite the tour including being mesmerised by the wave soldering machine (what safety covers?).

Probably couldn't happen today without it being pre-arranged with a full health and safety risk assessment. A shame really.

The revolution will not be televised because my television has been radicalised

Iam looking at a new TV

Possibly.

The one we have is getting a bit old and there is space for a larger one as when we moved to the new place a while ago there is more space and as I am not getting any younger a larger screen might be good.

You are correct that all the new ones want to connect, as does my sky box which has never been connected to anything (except to the phone line at first installation).

It whines that I will get a better experience* (yeah, right - Sky will get to sling ads depending on what I watch).

I see no reason to actually connect any of my appliances (apart from the computers when I need to use that functionality) to the internet. I am not a modern day Luddite (I have been designing electronics for decades) but as with so many things in technology, a connected TV (or fridge for that matter) it becomes a matter of "I know you can, but why?"

* Given that my so-called broadband struggles to get to 1Mb (usually half of that) I am not sure just how much use that would be to it anyway.

Python swallows Java to become second-most popular programming language... according to this index

Re: "Why not? K&R is ~250 pages long"

I started out doing some BASIC (horror!) and then assembly for Z80 (on a ZX-81 - that pretty much dates me) and followed that up with stuff on 6502, 68xx and more. The move to C was natural and relatively painless (there are many gotchas but the language is small enough that you tend to find those out quite quickly). That was even despite getting a rather terrible book (which perversely led me to dig further in search of answers and in a way actually made me better at the language).

Since then I have used a number of different languages, each with its pros and cons. Poor software is not about the language but more about not choosing the right language or not considering the problem space properly. There are others such as totally unnecessary levels of abstraction which is also a programmer problem, not a language problem. I tend to use C more than anything else simply because it fits the problem space I most commonly encounter.

My view is you need to understand one level below what you are doing. Much of my code has been hardware diagnostics and as such I needed (and still need) to have a detailed understanding of all the underlying hardware and hence the use of C.

The major problem I see is the use of inappropriate features of languages; inheritance and abstraction (as provided by C++) can be incredibly powerful and useful in a particular problem set, but bare metal diagnostics is not (usually) in that category.

Python has a well deserved place where it fits the problem space.

Tax working from home, says Deutsche Bank, because the economy needs that lunch money you’re not spending

Tax on stupidity?

If there was a graded tax on stupidity (the more stupid, the higher the tax), then Luke Templeman would very probably be broke.

As noted earlier, he works for a company that invested in lots of high density offices in cities and are now looking at the prospect of that income being significantly impacted.

From each...

Maybe we shouldn't get paid directly and all income goes to the government. They can then give us an allowance from that based on what they think we should have.

That is actually a cornerstone of Marxism.

From each according to their ability, to each according to their need

I do realise that you were engaging in a bit of sarcasm :)

Bad software crashed Boeings. Now it appears the company lacked a singular software supremo

Fly by Wire

Aren't most modern planes only flyable by wire?

Many modern aircraft are indeed fly by wire with a reversionary mode (particularly Boeing aircraft that were designed specifically for fly by wire systems).

Fly by wire systems are clearly flight safety critical and therefore get a great deal of scrutiny from a safety perspective at multiple levels (*). Boeing and Airbus (as well as Bombardier and Embrear) do not design or manufacture the fly by wire electronics or actuators; those are specified by the airframer to a suitable subcontractor as these are rather specialised designs.

The B777 (the original) is a fly by wire design and the electronics has a triplex architecture - 3 independent computers (and that means electrically isolated as well so a major defect in any one cannot bring the others down with it) along with 3 independent sets of sensors.

The problem with the 737MAX is that the aircraft has very little automation and MCAS was a bolt-on driven by certain rules (Part 25 stick force, primarily) but it was clearly never given the scrutiny that a fly by wire system should get.

* Boeing as an engineering company does not seem to exist any more, particularly since the McDonnell Douglas 'merger' (more like a reverse takeover) so in a very sad way it is not really surprising to see a total lack of ethics and scrutiny in the search for ever more profit regardless of the safety of the aircraft.

The fly by wire systems I am familiar with (which includes the B777) have very strict failure requirements, and those same failure requirements should have been applied to MCAS. Provided it can still control the flying surface then even a duplex system (single redundancy sensor and computers) cannot be considered safe unless it gives back complete control to the pilot in the event of a disagreement from the sensors and simply stays dormant in that situation.

The pilot who would need to be trained for such an event, clearly.

Re: Interesting spin

Ultimately, the root cause of the problem was terrible management overriding engineering which yielded, quite predictably, a terribly engineered aircraft.

A computer that has control of a moveable flying surface without pilot intervention that relies on a single sensor known for its unreliability and that is using that sensor information only and that was not hardened against single event upsets.

Really - what about true air speed, altitude above ground, rate of climb and dive (independent of the AoA sensor) which could all feed into this but even then with everything relying on a single computer the design is provably unsafe.

Safety Engineering

Boeing needs a complete safety team. That means you need not only software people, but safety engineers; there is such a field and it is common in avionics companies that take safety seriously.

Then systems engineers, ITAs (Independent Technical Authorities) who can be Boeing employees but have a veto over something going out of the door (cannot be overridden - the ITA is the ultimate signatory) and various design authorities. The list is quite long for a proper safety organisation.

Unless and until the Boeing attitude changes (so they become engineering focused, not MBA focused) this is simply putting lipstick on a pig.

Tech support scammer dialed random number and Australian Police’s cybercrime squad answered

Oh the narrow range

I might be amused if the scammers could come up with something original. Some of the ones I have had:

Your internet will be cut off due to IP abuse. There was silence at the other end when I asked if they knew what DHCP was.

This is Microsoft Support and there is something wrong with your computer. My better half likes to respond that 'my husband designs computers for a living and he says it is fine'. Works a charm.

Your BT line will be disabled unless you install this package <reason> . My line is not from BT.

Robocall that sounded like someone in a bunker. Left the handset off the hook for 20 minutes. At least I can cost them some minutes.

Other favourite actions:

Replay certain recorded parts of Home Alone (particularly the machine gun scene).

In a very deep voice "Let me get my husband". Used to have more shock value than it does today.

"You will need to wait until I stop recording the couple next door".

Occasionally I will use my best Edinburgh accent when answering the phone - "Hellooo!"

An older but good alternate "Big Joe's bar, grill and firing range"

One fine day when I was a single dad (and therefore my day is accounted for from 5:30 AM to 8 PM) I was cooking up my son's dinner (3 pots on the go) when the phone range and they asked to speak to the lady of the house. In my deepest voice "Speaking!"

With sales types who simply won't take the hint, I keep them on the line and critique their sales pitch (they have to make a certain number of calls per hour in some of the call farms).

I sometimes use the more amusing answering machine recordings one can get.

Suspended sentence for bank IT worker who broke into his boss's webcam because he didn't get a payrise

Re: I agree it seems like very light punishment

He would be questioned about the circumstances and exactly what happened

Perhaps. I recently had to do a background check (just the basic one although I have had SC in the past and will probably have to get it again) and you need to be able to account for the last 5 (10 for SC) years of employment (or lack thereof) and home address(es).

If you claim to have worked for a company that is now out of business, you will be asked for corroborating evidence (payslips, offer letter, resignation letter and so forth). One company I worked for was simply not contactable (they were still in business but no-one in the offices due to lockdown) and I had to provide said evidence.

Fortunately, I had kept some of the paper evidence and some electronic records that survived the last move a couple of years ago.

On the SC front, I had a (long) spent conviction from the mid 70s (young and foolish) that I declared and nothing was said about it (I am sure they looked for it though).

For those clearances, I agree that trying to hide something will most definitely be met with raised eyebrows and probable rejection.

Heck yeah, we should have access to our own cars' repair data: Voters in US state approve a landmark right-to-repair ballot measure

Safe and Secure Data

I too was highly amused by the name; their perspective, no doubt, is that it is safe and secure if they have exclusive access to it for their own (telemetry? marketing? ad-slinging?) purposes.

It is not safe for them if people suddenly figure out just what data these vehicles are actually collecting.

As for uploading and GPS co-ordinates, it is quite simple to make the appear to come from somewhere else as spoofing a GPS receiver is pretty much a doddle these days. Might be fun to change the co-ordinates by a hundred miles or so every couple of minutes :)

As for the data link, it is not that difficult to emulate 'poor signal conditions' with an appropriate RF sink (which basically attracts a large percentage of the energy coming from the transmitter - coupled with a random number generator it would look like a noisy link).

US govt ups minimum H-1B tech salaries to $208,000 a year, more than startups can hope to afford, say VCs

Re: But wasn't that the point of the H1B?

The nature of tech has changed dramatically since H1-B was introduced; there were far fewer IT bods as corporate networks were a rarity at the time.

The programme was for people highly skilled in arts and sciences and in the early days that is precisely who the visas went to.

I lived and worked in the USA from 1982 to 2004 although not on a H1-B - I got a green card which was simpler in those days but carried a lot of caveats such as no benefits until you had 5 (10?) years worth of paying into the system, you had to have a skill that would be beneficial and some other things that I don't recall right now.

As the world of technology changed and more and more people ended up in quite highly paid IT and programming, the nature of corporate needs also changed (our wage bill is too high!) and some of those saw a way to game the system with 'some' moving more towards 'many'.

The H1-B programme has not been properly used for at least the last 15 years (and probably longer) and the blatant gaming of the system made it farcical.

Those who now complain that the minimum pay is now too high should be complaining about the companies that gamed the system so blatantly as that is really what precipitated this change.

The we can't find people with the right skills is really code for we can't find people with the minimum necessary skills who will work for much less than the locals expect.

Given the truly blatant way in which companies gamed the system a backlash was almost inevitable at some point.

So don't complain these rates are too expensive; many corporations in the USA have been paying sub-par wages for decades and had an effectively indentured workforce.

I also seem to recall that these visas are not now getting automatic renewal (which effectively was happening for a long time) and have to be re-applied for in the same way as a new one.

So the overall effect of this is to move the goalposts back to at least close to where they originally were; to bring in people who really do have those rare and difficult to find skills.

Google's plan to make User-Agent string even less useful breaks our device detection tech, says NetMarketShare

Re: Original Sin

Deep in the mists of time, there were a number of websites that would complain to Firefox and Netscape that their site 'only works with Internet Explorer' <some version>.

Changing the UA string (pretty simple to do in Firefox) got rid of the complaint and in the vast majority of cases the site worked just fine.

One company I worked for from 1998 had some developers in California (I was sitting in my office in Princeton NJ running Netscape on an SGI Indy) where an internal set of pages written by said developers gave me an error that 'your browser does not support frames'. Quite amusing really as those were invented at Netscape IIRC.

Software engineer leaked UK missile system secrets and refused to hand cops his passwords, Old Bailey told

Re: BAE SYSTEMS' SECURITY

As with any other company that handles classified information, (see List X) there are controls on just what can go on machines (and servers) that connect to the external internet and secret (and above) is not permitted. Even on those servers, user access is pretty fine grained.

From what I can see in the article, this chap wrote the emails with the details from memory presumably using his own email account (which is, no doubt, why they wanted the passcodes).

His downhill spiral would very probably have been noticed and SC (possibly DV but that is both expensive and takes a long time to get) clearance likely revoked - if that happened then he would not have been permitted continued access to any of that information after that and could easily lead to him getting fired (clearance required and now not held by him).

I have seen some rather odd things; even renewing SC costs money and sometimes (for cost reasons) it is not renewed if the person involved has no reason to have continued access to secret material. I know one person who could not read a report he had written as he no longer held an SC.

BAE

thus imperilling BAe's profits

First, it is BAE (upper case E) Systems </pedant>

It is more likely to get them more in counter-counter measures. Most contracts are FFP these days (firm fixed price) unless the scope changes which happens a lot with MoD contracts.

That aside, the secret bits (or possibly TS) are more likely to have been dreamed up in QinetiQ and possibly BAE doing the implementation as part of the overall system as it is rare that a complete weapon is actually completely designed and manufactured by a single company now (BAE are likely a sub contractor in this case but there is nothing to say that BAE were actually involved - the article simply states the chap used to work for them).

Developer survey: C# losing ground to JavaScript, PHP and Java for cloud apps, still big in gaming

Language complexity

To solve a complex problem does not require a complex language; it requires a well designed expressive language that is appropriate for the application space.

I would not try to use JS in a bare metal project with a microcontroller (where C truly excels and yes, C has its issues but they are manageable) and I would not try to use a purely interpreted language where performance is a major issue.

The appropriate language for the appropriate problem is far more important. The problem may be complex, but what we are trying to do is describe the solution in an elegant manner. There is a reason we speak of the elegance of simplicity.

Adding complexity to a language invites problems (or sending it to the land of rm in a couple of cases) as it becomes harder to actually describe the solution.

So give me an elegant language, not a complex one (and one where I can understand a function without having to dig through multiple files).

Proposed US fix for Boeing 737 Max software woes does not address Ethiopian crash scenario, UK pilot union warns

Re: Fundamentals

The 737 MAX requires MCAS (or something that does the job albeit in a less lethal way) because without it, it cannot be certified under part 25 rules (Passenger aircraft). The aircraft still has to pass these tests even if it is a derivative design.

The rule in question states that for a given amount of pull force on the column, the rate of change of AoA must not increase.

In a particular part of the flight envelope, the rate of AoA increases without additional pull force and the aircraft is therefore not airworthy according to those rules.

So MCAS was not only to pitch the aircraft (no pun intended) to the airlines as being 'just the same' as a 737NG but also to pass the airworthiness rules.

Linux kernel coders propose inclusive terminology coding guidelines, note: 'Arguments about why people should not be offended do not scale'

Some more interesting possibilities

Many terms used in electronics (particularly interface protocols) and software have a very specific meaning.

Before I get to that, it is interesting that over 20 years ago on Irix 5.3 (IIRC) we had hosts.allow and hosts.deny.

Master slave interfaces (such as SPI) have been around for decades and the term is accurate (which engineers tend to approve of as the meaning is quite clear). This is also used for internal clocks which must be synchronised and phase locked (perhaps at different frequencies) among many other things. In SPI, the I/O is actually named MOSI and MISO (Master Out Slave In, Master In Slave Out) in every such controller datasheet and reference manual I can recall over the last few decades

Some other terms that might make heads explode:

Multi-master Bus - commonly used in PCI as an example (where multiple devices can be the master at different times).

Race condition - a specific issue in combinational logic where signals are changing at the input of a gate at the same time so for a short period of time the output is unknown but this can also occur in multi-threaded software.

We could use so many more here: DMA - Are you peering into my mind for control? Eek.

Peripherals - they are no longer at the centre of things!

Let's not forget the JBOD - they are being marginalised! (I note there is also the MAID acronym - not hard to see how that might get the ultra PC crowd fuming).

Madness. In the context in which these terms are used, the meanings are well known and have specific meanings. There are producer - consumer architectures; I bet someone could get annoyed at that although it is an accurately descriptive name.

Many of the posts I see online are from the 'professionally annoyed' (they appear to try and find things to be upset and annoyed about).

What would they think of a declaration I once used many years ago (it was completely accurate too):

short volatile boss;

Vertically challenged, not short!

Hey, Boeing. Don't celebrate your first post-grounding 737 Max test flight too hard. You just lost another big contract

Re: It worked once, ship it!

Electronics for safety critical systems on aircraft have been around for a long time. The B777 has a triple redundant system with manual reversion (the pilots can take control).

Some modern fly by wire actually drives electrical actuators.

The mantra of the old (engineering led) Boeing that the pilots ultimately knew best. That was before the bean counters were put in charge.

The avionics that make up flight control systems are rigorously tested against the airframer requirements for verification (verification can cost a lot of money).

Boeing did the system integration and did things that should not have passed even a cursory check of what MCAS was capable of (the final system was nowhere close to what had been declared to the FAA).

The MCAS computing element would have been designed, built and tested by a third party to Boeing's requirements and I would not be surprised if those vendors know where the bodies are buried as I am sure that if the vendor know this could control flying surfaces from a single sensor there would have been significant questions raised.

This'll make you feel old: Uni compsci favourite Pascal hits the big five-oh this year

Re: pascal was simply useless.

A company I was with for the first half of the 90s had a pretty major application written in Turbo Pascal.

When we got this newfangled thing of a network (Netware 3 IIRC), that team was trying to access network drives but there was no way to identify them as such from within TP (at least as far as that team was concerned).

I was doing a lot of C at the time (still am for that matter) and so I wrote a linkable function that accessed the list of lists that could determine the type of storage and the drive letter (local, removable, network). I was the toast of the team (well, for a short time - after all, that was (sniff) just C).

I will note that a pretty major ECAD package (electronic design) was originally built with TP and is now built using Delphi.

There were pitfalls, of course; I once wrote a fairly small piece to parse credit card details from smartphone data which worked perfectly - one of the Pascal people did the same, but it choked on occasion with an incorrect record (too short as I recall).

The record lengths were fixed and the bank involved insisted that the records had no delimiter to re-sync should a record be wrong. Easy to enforce in C [using fwrite()] but quite difficult in Pascal.

ICANN finally halts $1.1bn sale of .org registry, says it's 'the right thing to do' after months of controversy

Picked up by mainstream press

But they clearly don't know the whole story.

Telegraph UK article