Just because you failed doesn't mean you weren't right We take a break from busted backups in today's Who, Me? in favor of laser-guided missiles and a reminder to check every element in your test plan. Our story takes us back many decades, when laser-guided ordnance was new and our hero, who we will call "Herman", was working with both a well-known military organization and an …
Because the Italian Army was not well equipped, not very good at combat overall, and often not even much willingly to combat? Landing in Sicily was a little easier than landing in Normandy, wasn't it?
Mussolini may have dreamed about being a Roman general - but he no longer had the combat capabilities of the Roman Legions... when your highest-ranking general is someone called Badoglio, it's quite clear your Army is not going to be made of the brightest and most dedicated ones...
You can have pass/fail OR how close the bomb was because the shock wave could do quite a bit of damage.
Old saying applies and I've heard in the military and in engineering.... "Close only counts in horseshoes, hand grenades, and nuclear warheads". So may close would have been good.
When the NAZIs were busy building the concrete fortresses on their norther coast, the Allies let them finish and then bombed them with Barnes Wallis's 10-tonne bombs. The idea Wallis had was that a near miss would allow the bomb to undermine the foundations and wreck the entire building, whereas a direct hit would do far less damage.
One bomb achieved the 'very near miss' and the 'fortress' was abandoned.
During the War (don't mention the war!), my father was an His Majesty's Inspector at a large armaments company based on the north side of Clapham Common. Their Research Department was tasked with developing a sighting mechanism that would enable Barnes Wallis's Bouncing Bomb to be dropped accurately over the German dams. It ended up as three pieces of plywood and a length of broomstick, with four nails driven into the plywood to indicate the correct drop point when used to sight on the towers on the dams. In retaliation, Gerry came over and bombed the shit out of Clapham, so production was moved to a satellite factory near Barnet.
My father who was a WW2 RAF Bombing Leader had a photograph of a refined version. A triangular piece of plywood with a metal plate with a slit as an eyepiece near the apex, and two painted pins near each end of the base. The photo was marked SECRET. He showed it to me when I was working in a similar area of MoD in the 1970s. He was asked to stay on in 1946, but decided his skills were redundant because of the development of nuclear weapons.
What you do is, right, you nail the broomstick to the target. If you can still find the broomstick after the bomb's been dropped, then it missed."
"Yes thank you Mr. Dibbler. I can't help but feel that it's about as useful as your dragon detector. Which it appears to closely resemble, too."
As someone who spent most of my career in quality assurance (mostly in more traditional engineering, not IT, I hasten to add), the first step after a failed inspection is to verify the calibration status of the test equipment. That’s when, of course, you hadn’t checked calibration status *before* the inspection (which should have been the default).
Most times, when called in to investigate an unexpected inspection fail (i.e. something that all previous checks indicated no problems and, for me to be called in, something that was going to be expensive to fix), my checks on the test gear found the culprit. I won’t say that heads rolled, but expensive lessons were learned :)
I used to test machinery using a variety of transducers etc and 'precision' instruments. Whenever a machine failed to achieve the design performance, the first assumption from the affronted designer was always "instrument error". Re-calibrating the appropriate instruments or using a different set never improved the measured performance.
I was once at a well known testing lab, and our product failed their mains leakage current testing (which it shouldn't have). We asked them for the data on the test equipment used, and they produced the manufacturer and model number. I found one on eBay cheap, and purchased it.
Using it (as received, i.e.: uncalibrated) to test our gear, I came up with the same numbers the testing lab had. After scratching my head a bit, I went back to the IEC spec and read the paragraph on the test, which helpfully provides a test method using a simple test circuit and a meter. When tested using the recommended setup, we easily passed.
Moral of the story is, of course, "don't trust it just because it has an LED readout and a computer inside"
Another little gem to note: when testing/inspecting always verify the 'zeros' on instruments/readouts before the test starts. As customer, I have experienced the testing/witnessing having to be abandoned when I noted a 'calibrated' instrument displaying a negative or impossible value before starting.
As supplier, I made sure the customer was happy with the instrumentation before the test began. On a good day, the customer would take the readings/observations!
And on computerised systems, the data being processed should be the figures you are witnessing!
My last foray into EMC testing was fun. EMC testing facility was basically a small wooden building in the middle of a field - with a rather long electrical supply cable.
Server had three ~1200W (I think) PSUs, and because it was "just testing" we had them all plugged into a single four way extension.
About to do all the zappy stuff to the box, which had been put in the back of my car the previous day, and so we plugged in the four bank. The inrush from those PSUs turned all the lights out for about half a second.
Fortunately the chassis and PSUs were already certified, so we just decided to ignore that...
I've been on both ends of this. Standard practice is to verify the test equipment before any important test, and again afterwards just to make sure. Occasionally this kept us from running a test with bad equipment, or discovering it had drifted out of tolerance at the wrong time.
On the other hand when product would call saying their readings were wrong and the instrument needed calibrating, 98%+ of the time the instrument was accurate!
Used to be fun when quartz oscillators needed to be temperature controlled. You needed to power up the kit and leave it to half hour or so to allow the thing to stabilise.
We have just one customer with a setup that required one of these testers and as they were just a few minutes walk away we lugged the tester round to their place and they had it on 'loan' so that it was always powered up and ready if we needed to visit.
This kit was from the early/mid 70s and I think it was because the cost of the chips needed to build the clock divider chain meant that the oscillator ran at 100s of kHz, which meant the crystal was relatively large and susceptible to changes in temp, where the digital watches of just a few years later ran at MHz and needed smaller crystals that were less affected
a large and empty Italian test range, a tower housing the German team and their laser projector (aimed precisely into the middle of the test range),
Let me guess, a clowder of cats, that suddenly came out of nowhere, tried to attack the target the laser was pointing at and at the same time as the test pilot fired the missiles.
The fur was displaced at around 1 mile radius and locals started talking about the "cat-astrophe" (from Greek astro-(stro)phe, the act of turning to the stars)
And hence why the word Catastrophe was born.
Same with coding as well.
I've had trainees who were told to produce a bit of code and then test it to make sure it worked. Cue a few hours spent with the trainee running their test, it failing, and them going back to the code to tweak it and try again. Eventually they asked for help, and it turned out that their initial code was fine. All the tweaks they'd added to it had changed how it worked, but not affected the (correct) outcome. They'd just omitted something critical in their test and hadn't thought to check that.
When you are chasing your tail trying to solve a bug - remember that Occam's Razor is only a probability. What you then find are two (or more) unconnected root causes that produce the same symptom.
It is not unusual to have fixed a bug - but unintentionally introduced another similar error in the process.
We once had a hard failure in a mainframe with parity errors on the console typewriter. The field engineer swapped a likely board with another parity board but this didn't solve the problem. Six hours later he discovered that the parity errors were no longer word position sensitive. Turns out that the board he swapped the suspect board with was also in the data path.
Sometimes it is best to forget all you have learned during trouble shooting and start at the beginning again.
My father used to teach in the Royal Navy. He would get a classroom full of people from the Commonwealth Navies and teach them. At the end of the first class, he asked if everyone understood. Most people nodded, and a few shook their heads. The people who nodded were allowed to go.
My father did the whole lesson again - much slower. At the end, he said "do you understand" and the class shook their heads.
Ever patient, my father prepared to do the course again, when one person said "We understand perfectly, when you ask use we all said yes".
In their country nodding means no - and shaking the head means yes!
As well as this little confusion they did not like to say no. The answer to a question "are you on target and ready to fire" was always yes. They did a lot of damage by hitting things outside of the target area.
Another problem was caused someone not using his right eye to line up the back sight and the fore sight, but using his left eye and only the fore sight, and shutting his right eye! This makes a big difference over a couple of miles.
in "Flying Freestyle: An RAF Fast Jet Pilot's Story" by Jerry Pook, there is a wonderful chapter on how the US Marines wanted to demonstrate laser targeting of smart bombs.
What the Marine designated to sit in a slit trench and illuminate the target did not understand was that the actual infrared laser beam was eminating from the targeting device located a little below the scope. Hence, while the scope was aimed perfectly on the target, the laser beam was illuminating the sandbags just in front of the trench he was sitting in.
If I recall correctly, the author was impressed with the speed that even older high ranking marines could achieve when properly motivated...
I quite liked the way the Italians neatly skewered the typically German obsession with precision, detail, and procedure by quickly hiding the theodolite in a cupboard ("Milano", hah) and pretending to use a broomstick. And having a pile of comics on the table as well was just pure class.
.
:-)
HF PA circuits could exhibit a problem of unwanted parasitic oscillations. The cure was to wrap a few turns of enamelled wire round a resistor and make that the anode series connection. On the other hand if there was already such a "parasitic stopper" - then the cure was to remove it.
To detect if you had such parasitic oscillations you held a small neon bulb near the valve anode. The RF field would make the neon bulb glow red - but if there was a violet tinge then you had the problem.
We once had a lecturer describing adding ferrite cores to cables to stop ringing and similar. He said the best method came from a far eastern* "Dr Suckitan", who had three methods A, B and C.
Apparently the Suckitan C method was best - well done to those of you ahead of the wave.....
*Somewhere far away and foreign and not the UK where they have more interesting sounding names. Maybe Smith, Clark, Brown, Longbottom etc are all hilarious in some other languages.
A neighbour's German girlfriend was admiring some flowers in my garden. She couldn't understand why people started to giggle. The reason was that English has bowdlerised the pronunciation of "Fuchsia" - whereas the Germans recognise it was named after the 16th-century German botanist Leonhart Fuchs.
The world is warming, as we know. However, we (i.e. meteorologists) now have satellites to measure temperatures across the globe. But some satellites were recording temperatures which denied global warming. (https://www.theguardian.com/environment/climate-consensus-97-per-cent/2014/nov/07/new-study-disputes-satellite-temperature-estimates).
It turned out that the satellites had been lofted into orbits which were subject to atmospheric braking, so although they were meaning the temperatures at ground level on the same spots on Earth, they were measuring it through more of the atmosphere, and therefore getting incorrect measurements. They simply had not bothered to calibrate the results with ground based observations using mercury thermometers.
You’d expect those satellites measuring temperature to also be able to measure how far from the ground they are, also space telemetry and radar would show how far up in the atmosphere it is.
Satellites don’t stay in exactly the same spot, they drift in an imagined 3d box with the drift controlled by the onboard thrusters.
The data returned must be constantly calibrated against things like distance etc, else the error margin will be larger than needed.
Maybe, but I seem to have a 'Register Stalker' who downvotes even my factual posts without explanation, and happens to have read (and downvoted) my previous post on this thread as well.
One should not be too upset at such minor issues when there is armed conflict in Ukraine, Syria, Yemen, Covid-19 is still a major pandemic, global warming, plastics pollution of the seas and loads of other things. I shall as a true brit, 'keep a stiff upper lip and carry on regardless'.
@ Dave314159ggggdffsdds,
Well, umm, thanks for the honesty of your post, and your integrity in not going Anonymous Coward either. (So I shall not been downvoting it or reporting it for abuse for suggesting I shove something up an inappropriate place on my anatomy.)
It is the first time I recall any of my posts being described as 'far right rants', and I'm not sure what 'denier nonsense' of mine you refer to. Maybe the next time you come across either you could post a reply explaining why you think what I've posted is 'far right' or 'denier nonsense' please.
Curious. I just read both the Guardian article you link to, and the Weng etal paper it cites, and cannot see any particular reason to reach the conclusions you state in your second paragraph. Weng et al did seem quite interested in the effect of clouds, however, which I suppose matches "more of the atmosphere". Also, any remarks about calibration would seem to be more along the lines of which approach was best, rather than a claim no calibration was done in some case.
Further, the Guardian article ends with "So, the verdict is still out ..."; which also seems to suggest your second paragraph was perhaps over confidently stated.
Still, perhaps I missed something. Can you point out specifically the part of the Weng et al paper which I should look at most closely?
Apologies, the article I linked to was an example. I should have stated clearly in my post that the deduction that the scientists had not used ground based observations of land temperature to calibrate their readings was my own, and not in the paper.
My mistake. Sorry for the confusion.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
