Hacker News Reader

Comment by Valgrim | Hacker News Reader

The basis for the metric system was not defined by some arbitrary prototype at first, but by measurements everyone on the planet could reproduce using the sun and a bit of water, at least to a very acceptable approximation.

The kilometer was based on 1/10000 the distance between the poles and the equator. The liter was based on the meter, the kilogram was based off the weight of water in a liter, the celcius was based off the freezing and boiling point of water at sea level, etc.

The reference meter was put in place to fix the problem arising from the error margin in measurements, but for practical purposes, the definition was all that was needed for a country to join the metric system.

View on HN

> The kilometer was based on 1/10000 the distance between the poles and the equator.

The definition we should have used (and almost did) was that the meter was the length of a pendulum with 1 second swing (in other words, a pendulum with a 2 second period). https://maa.org/sites/default/files/images/upload_library/22...

If we had done that, the gravitational constant of Earth would be exactly pi^2 m/s^2: Neat, right? https://godplaysdice.blogspot.com/2007/09/why-g-2.html

I wish it had been defined in relation to the speed of light in a vacuum. 299,792,458 metres per second. Set it at 300,000,000 meters per second and derive the length of the meter from that.

Interesting. The meter was defined in 1791. James Bradley got the speed of light correct to about 1% in 1728, and I think that value stood in 1791. That value was less accurate than either the 10000km pole-to-equator or 1-second-pendulum, as measurable in 1791. And until special relativity, the idea of speed of light being constant wasn't really firmly established. Ultimately, I think in 1791 I think using the speed-of-light as the basis wasn't really practical.

P.S. Not really necessary but in writing this comment, I read these documents, each of which taught me something new:

https://interestingengineering.com/science/a-brief-history-o... https://spie.org/publications/tt82_25_speed_of_light https://en.wikipedia.org/wiki/Luminiferous_aether

> I wish it had been defined

Wish what had been defined? the second?

> and derive the length of the meter from that.

I would assume they meant meter since they wrote that.

The meter. The distance light travels in a second in a vacuum is fixed. Divide that distance by 300,000,000. That is the length of 1 meter.

So the current definition is:

> The meter (m) is defined by taking the fixed numerical value of the speed of light in vacuum c to be 299,792,458 when expressed in the unit m s−1, where the second is defined in terms of ∆νCs.[0]

Your proposed change is keeping this the exact same but changing 299,792,458 to 300,000,000?

You said: “I wish it had been defined in relation to the speed of light in a vacuum.”

I believe it is already, as written above.

0: https://www.nist.gov/pml/owm/si-units-length#:~:text=The%20m....

If you prefer for some universal constants to be round numbers you can choose a Natural Units systems according to your preference. For example in some of these systems c = 1, which turns Einstein's famous equation into just E = m. Convenient.

I wish we used a unit more similar to the foot as the base. Its much more useful for human-scale understanding - as seen by many civilisations (from the Japanese, Chinese, indians, ancient egyptians, romans, etc.) adopting something similar (something between about 10 and 18 inches). You can build off that as you need of course.

That's an education bias. I find meters way more intuitive than feets.

Same here.. especially feet can have a 100% difference in small and large feets, even in adults alone. Any human measure is prone to personal bias then, similarly obscure as if we had defined 2m to be the height of the average human?!

I think something similar to stride length would be more useful as base, about 70-80cm. Riugh measuring by strides is something everybody does from time to time. Meter kind of works for that, but you need to compensate by concsciously making longer strides.

I don't thik foot is particularly useful in day to day life - it's awkward to measure walking distance with that, and for measuring smaller objects it's awkwardly placed, an arm length unit might be more useful for measuring smallish objects.

You mean, imperial?

No you can still use a metric system, just use the foot (or similar) as a base and then kilofoot, megafoot, microfoot etc. rather than inches, yards, miles, links, chains, rods and perches etc.

We already kind of do that. I’ve seen things like 1.5 gallons, 2.5 miles, 10.7 feet. I’ve never seen the intermediate imperial units being used.

At least in the UK my experience is we don't really use decimals with Imperial units - it seems much more natural to use fractions.

For example, "10 feet 8 and a half inches" not "10.7 feet" (yes I know the rounding is slightly different). For a smaller distance you might say "three sixteenths of an inch" (since thats what the imperial side of a tape measure is typically divided into) though there are some exceptions eg "0.1 inch spacing".

Pints and fl oz the same (you see this quite often in recipes for example) - "1 pint 3 fl oz" or even "23 fl oz" not "1.15 pints". For gallons I think most Brits would likely use a whole number of pints rather than subdividing (ie "21 pints" not "2 gallons, 5 pints").

But I'll give you that we don't really use furlongs outside of horse racing :)

If feet is so universal and superior why is yards and miles used?

At the time the metre was defined in the late 1700s, it was already known that g is not a constant and varies from place to place, which is why they did not choose that proposal and picked the geodetic proposal instead.

That wouldn't really work though because Earth's mass over a longer period of time isn't a constant (and hopefully it's size is).

Gravity also isn't constant everywhere on the planet and it changes locally over time as well as globally. Precision measurements would be a nightmare/impossible.

How is 1 second measured in that definition? Is that not going to need a Swiss watch?

Wild guess: you don't measure a second. A second is defined as 1/86400 of a day, or 246060.

Lets say a pendulum can swing for about 2 minutes. You just take a 2 minute hourglass, make a bunch of pendulums of different lengths, and assign someone to count the number of swings between the moment you turn the hourglass and the moment the sand has passed trough. The one closest to 120 is the correct length.

Now this solution is mathematically elegant but comes with a lot of problems. First you must rely on a less-than-perfect time measurement device which may suffer from day to day variations due to temperature.

On the contrary, to calculate the distance from the poles to the equator, all you need is a good angle measurement device, from which you can triangulate a distance.

> On the contrary, to calculate the distance from the poles to the equator, all you need is a good angle measurement device, from which you can triangulate a distance.

Can you expand on how this can be done please?

Particularly the bit where you measure the semi-major axis and semi-minor axis of the Earth's ellipsoid and how you settle on a particular ellipsoid out of the many that predate WGS84.

https://en.wikipedia.org/wiki/Earth_ellipsoid#Historical_Ear...

"All you need" appears to bury a lot of detail in some hand waving.

It's unclear to me whether you intend to measure a distance along the Geoid (gravitational equipotential surface), along the Ellipsoid (the rotated ellipse of best fit), or the actual suurface (the path that would be walked, floated, traversed from equator to pole).

A complication with the Geoid path family and surface path family, of course, is that these vary by what line of longitude you might follow.

Eratosthenes did that in his time with spherical Earth model, and it was precise enough for ~2k years. A lot that can be hand waved with that route as opposed to relying on availability of a well defined and calibrated precise time measurement device, whether it's 1 second or 2 minutes.

Eratosthenes who now?

Sorry, couldn't help myself, unlike yourself.

Just recall we're in a thread about precise definitions of reference units, the title posed refers to tens of trillions of oscillations as the definition of a second.

When the kilometer was first coarsely defined by an Earth arc it was already a thousand+ years post shadow of a stick guy.

Once that kind of definition gets refined further the "arc of the earth" approach hits all of the hurdles I outlined .. for the benefit of the person I replied to who somehow believed it was a simple straightforward task to make it "better".

Hand waving a reference is all very well ... until you hit cases that require actual precision finer than the error bars on the hand wavy seat of the pants cowboy antics.

That bar was crossed some decades past.

I’ll never stop being fascinated by how neatly all of this fits together: take a hollow cube that has 10cm long edges. Fill it with water - now you have a liter of water. It weighs 1kg. Put it into the freezer, when it just starts freezing, it’s 0°C. Put it onto a stove, and when it starts boiling, it’s 100°C. Simple as that.

> Put it onto a stove, and when it starts boiling, it’s 100°C. Simple as that.

Not quite - don't forget your elevation | air pressure.

Of course not even elevation above mean sea level is straightforward .. what is the MSL of Australia, for example, and how does it relate to the Geoid, the long tidal run through the long narrow straits to the north of Australia, the WGS ellipsoid, etc.

Not to mention the dissolved solids as well, so it's not as simple as using tap water.

IMHO Celsius is the SI unit least deserving of praise. It's just bolted onto the system through the Boltzmann constant and isn't intricately linked like other SI/CGS units. It's also a little too coarse for modern use. Sure, I could use decicelsius myself, but that doesn't stop every other weather station, AC/thermostat manufacturer, etc. from still using whole degrees. You can buy $2000 Italian espresso machines with elaborate PID controllers that can only be tuned to whole degrees, which ironically means that using Fahrenheit is more precise!

I think there’s some confusion here - the Kelvin is the SI unit of temperature, not Celsius.

The isotopic content of the water matters too. Boiling point are not used to calibrate thermometers though. Freeze points and triple points are used for temperature calibration. And outside of a primary standards lab, it’s usually just a transfer standard (another, more accurate thermometer).

> doesn't stop every other weather station, AC/thermostat manufacturer, etc. from still using whole degrees

I am European and while I agree on weather stations, I have never seen an AC or thermostat that doesn't at least use 0.5°C increments. (And I'd argue a whole °C is perfectly appropriate for weather.)

Besides Celsius not being the SI unit.

I'd be shocked if weather forecasting is ever precise to less than 1-2 °C

> I'd be shocked if weather forecasting is ever precise to less than 1-2 °C

That's not a forecast .. actual forecasts for a region are often more accurate than that wrt "mean tempreture" for that region.

Where I live the BoM (Bureau of Meteorology) cannot "accurately" predict the upcoming daily tempreture and rainfall for the town that I'm in.

They do very accurately predict the mean temp. and rainfall for the wider region.

Doppler RADAR gives a very accurate measure of the amount of moisture held by rain bearing clouds, stable wrt broad area winds give a good path for that moisture's travel in the next 12-24 hours, experience + data give solid predictions for the amount of water to be shed, along with bulk cooling from shade and water fall + evaporation.

The difficulty rarely lies in the bulk parameters for a broad area, the devil is in the small details which are inherently turbulent.

To measure the accuracy of forecasting it first has to be ring fenced by what area you intend to predict the mean attributes for, the timescale you intend to project for, etc.

Yes, but all of that is unrelated to the original point

The commenter I replied to was complaining about weather stations not using decimals.

As you point out we can predict averages with a better accuracy, but services which have that accuracy do use decimals, where decimals aren't used is in consumer weather forecasts where an accuracy of less than 1 degree is anyway useless because

> the devil is in the small details which are inherently turbulent

Also temperature is variable. It fluctuates through the day and then things like clouds or rains can drop it quite a lot or keep it higher during winters in parts of the world.

You can get pretty good hour level forecast in some scenarios. But how do you effectively communicate those outside showing some type of graph...

The only use case I encountered where Celsius is too coarse is measuring body temperature.

If you do this at Sea Level when Air Pressure is the barometric mean pressure with nothing but H2O, yes. Otherwise, not exactly, in ways that require formulas to quantify. Metrology is the dirtiest of all sciences, it does not grant simplicity.

Anyways.. the counter example was the old definition of the Ampere:

"the constant current which, if maintained in two straight parallel conductors of infinite length of negligible circular cross section and placed one metre apart in a vacuum, would produce between these conductors a force equal to 2 × 10−7 newton per metre of length."

Perfectly unrealizable within this Universe.

It gets even better: take that 1kg of water at 0°C, add 100 kilocalories of heat energy, and it's now 100°C and just about to boil.

(That's assuming you're at sea level, as another commenter noted - if you're at 5,000 feet, the boiling point of water is something like 95°C.)

It has some warts.

A mole of electrons has a charge of about 96487 coulombs. 100000 would have been nicer.

> the celcius was based off the freezing and boiling point of water at sea level

These are not particularly great choices. Water transitions to a solid when it freezes, and distance from sea level has a massive impact on the boiling point of water which Celsius did promptly characterize. Also, somewhat weirdly, Celsius used to be inverted, with 100 being freezing and 0 being boiling.

> The basis for the metric system was not defined by some arbitrary prototype at first, but by measurements everyone on the planet could reproduce using the sun and a bit of water, at least to a very acceptable approximation.

I've got sun here and a well that will give me almost 50 liters/minute of water indefinitely. What measurements do I need to do to get a good approximation of the meter?

Problem is "the planet" is part of the requirements in that sentence too, since the original meter definition was 1/40000 of the circumference of the earth.

You don't need to sail around the world for that, you can get it by shadow measurements, but you need measurements from two places on earth at the same time of year/position of the sun. cf. [ https://en.wikipedia.org/wiki/Eratosthenes#Measurement_of_Ea... ]. The result you get from that is knowing what fraction of the earth's circumference is between the two places, which with the 1/40000 definition you can convert to meters. It's certainly doable, but you need the places to be reasonably far apart to get any accuracy, so… more of a "summer road trip" project, but could be fun with kids? ;)

To be fair, the entire idea of metrology is that a few people do the work to get primary references and you distribute from there. There's a reason entities like NIST and PTB gets tax dollars.

Kids? I would probably have the time of my life organizing an expedition with 5-6 full grown adults where the goal would be to go out there with nothing but our brains and hands, no device or ruler or anything like that, and figure out the length of a meter.

>The kilometer was based on 1/10000 the distance between the poles and the equator.

And that measurement they got was ultimately off, which is why the meter became (until its later redefinition), for all intents and purposes, "the length of some arbitrary metal bar".

Clearly, "some fragment of roughly how large the Earth is" is a base unit designed first and foremost for human-scale measurements, and definitely not by self-fellating academics that would never seriously have to use the thing (and whose work does not, by and large, depend on the base unit used; Americans use thousandths-of-an-inch for everything under 1/64th). Fortunately, metric time was too weird even for the French.

>but by measurements everyone on the planet could reproduce using the sun and a bit of water, at least to a very acceptable approximation.

By contrast, the "foot", which is literally named after the device used to approximate it. The median human being has two of these. In fairness, you usually have to be taught which part of one's body approximates the inch and is thus harder to do.

>join the metric system

The metric system was initially imposed at gunpoint by the French upon most European nations in the first place due to the French going batshit insane just before 1800 and conquering most of the Continent. The Russians, being a nation that rather famously was not conquered by the French during that time, would retain Arshins until Communism (at which point the Soviets would impose the metric system, also at gunpoint- kind of sensing a theme here); the English would retain the Imperial system even longer than that.

It's also why people living in the only nation that really has a choice all measure distance and height in (square) feet, inches, and kilometers (per hour), human weight in pounds (and very rarely ounces), most other weight in grams and kilograms, non-cooking-related volume in liters, and non-cooking-related temperature in celsius.

It would have been objectively better from a usability to have just set '1 meter = 1 foot' and call it a day- could have even been (one of) the French foot, too. But no, just had to fuck it up for everyone who actually has to use it because "we know better".

> And that measurement they got was ultimately off, which is why the meter became (until its later redefinition),

> By contrast, the "foot", which is literally named after the device used to approximate it. The median human being has two of these.

It seems odd to criticize the metre for a lack of accuracy against historical basis when the modern definition is accurate to within 0.02% of its historical basis, and yet praise a unit that, if measured by its historical basis, would be off on its mean by around 20% and vary by 40-50% amongst the population.

I never thought I would see an American Exceptionalist comment about a system of measurements.

Metric measurements are as intuitive to me as American customary units are to you. I can just as easily use my handspan or the lengths of my fingers to estimate lengths in centimetres as you can in inches. I know without thinking what clothes I should wear when I see the weather forecast in Celsius. I know roughly what temperature I should set the oven to in Celsius based on what I’m cooking.

Having grown up in a metric country, American or imperial units mean nothing to me. If I’d grown up in America I’m sure American units would be intuitive for me. The only difference, as you say, is metric is far easier and more intuitive as soon as you want to do anything complicated with those measurements.

> I never thought I would see an American Exceptionalist comment about a system of measurements.

You haven't been on the internet long, have you?

Welcome, and I'm sorry

>American Exceptionalist

Ah yes, because all criticisms of SI's inherently flawed design must be from an American. Lol, lmao even.

>I can just as easily use my handspan or the lengths of my fingers to estimate lengths in centimetres as you can in inches

But you can't do that for meters, whereas you can for feet- and I only need to remember one body part for one inch (and one for one foot, naturally) whereas you have to remember several because the centimeter is too small, and the meter too big, to use in this way (you cannot even really generate 1 cm- so you have to do multiplication every time you want to do this).

The SI base unit's design handicaps its usability here- having to use fractions or having to multiply by 2 or 3 afterwards to compute what would normally be counting by 1s is, obviously, not an improvement.

See, the average Joe gains zero benefit from the knowledge that he is using a measurement that has something to do with how large the Earth is. Sure, navel-gazing academics like the symmetry, and math problems in school are a bit easier when you're just staring at an abstract number rather than having to sanity-check a measurement of an object you have sitting in front of you- but a system of measurement designed for human beings (and not states or vague modernist principles) should have picked the latter to work on, not the former.

It is not a coincidence that every country in the world, pre-SI, set their measurement's "1" (or 1/10th, 1/12th, or 1/16th) at what we now consider to be 25-30mm: good design, intended for humans. The fact that number in SI is not a round one: bad design, intended for... something else.

>is metric is far easier and more intuitive as soon as you want to do anything complicated with those measurements.

Until you want to divide them into thirds, fourths, sixths, or eighths, then you have to remember the 3 numbers of the fraction anyway. Doing this makes sense for abstract things like money and (to an extent) volume, since I hope you're not measuring liquids with your hands, but "visualize how much water this is in relation to some length" is generally not a problem anyone who isn't writing things down already (and can thus deal with arbitrary conversions) has.

And for those who do work in odd fractional measurements... they're already working in decimal anyway and have been ever since the caliper was invented, so what difference does it make for them?

>But you can't do that for meters, whereas you can for feet- and I only need to remember one body part for one inch (and one for one foot, naturally) whereas you have to remember several because the centimeter is too small, and the meter too big, to use in this way (you cannot even really generate 1 cm- so you have to do multiplication every time you want to do this).

I never had to measure anything with my fingers or other body parts. I don't know what the point of that is. Being off by an accumulated error of 5 mm is often inaccurate enough that you failed and have to redo things.

I don't have to remember "several", because kilometers, meters, centimeters, millimeters are just meters with zeroes to the left or the right of the comma.

The thing is, Americans use their own nonstandard metric system. They use metrics like thousandths of an inch, which is just 0.0254 mm or as big as 25.4 micrometers. Thousandths makes things sound smaller than they really are. Then you realize it's just 1/40 of a millimetre and it sounds really big at that point and the engineering profession shows it, by having another unit. Ten thousandths of an inch. Wow, amazing. We have almost reached parity with the micrometer!

The average Joe massively benefits from the fact that the measurement relates to the planet, because the planet is a shelling point common to all humans and therefore a good way to reach an agreement. Meanwhile your ideas are great if you want to divide humanity into multiple disagreeing camps.

Also at the same time possibly using various fractions of inch... Like can't you really decide on doing one way or an other?

> It's also why people living in the only nation that really has a choice all measure distance and height in (square) feet, inches, and kilometers (per hour), human weight in pounds (and very rarely ounces), most other weight in grams and kilograms, non-cooking-related volume in liters, and non-cooking-related temperature in celsius.

If you're talking about Canada, none of that mess comes from a plethora of choice, it's just the result of American cultural and economic influence, in addition to history. Yes Canadians are well versed in both systems but it's not like we particularly like having to deal with US units in our lives.

This can get quite frustrating where the two worlds meet, for instance in construction.

Reducing SI adoption to “gunpoint” is just silly.

It was objectively better system and that’s why it ultimately prevailed.

>It was objectively better system

No, it isn't, and I've already outlined the reasons why.

>Reducing SI adoption to “gunpoint”

Is what actually happens. The French pushed it through their Revolution, the Soviets did it in '25, the PRC in '59.

Interestingly, a good chunk of the existing Chinese measurements were metric themselves for a while: simply imposing a new order was the objective. Thus always to modernists.

> No, it isn't, and I've already outlined the reasons why.

The only argument in favor of imperial measurements you've given is that humans have two feet, which isn't much of an argument given the variation in sizes of those physical feet.

I still have no idea what an ounce is supposed to measure. Is it mass? Is it volume? Is the ounce for measuring gold different to the ounce for measuring food?

> people living in the only nation that really has a choice all measure distance and height in (square) feet

The U.S. automotive industry certainly disagrees. They've largely (80~90%) gone metric quite a while ago.

> It would have been objectively better from a usability to have just set '1 meter = 1 foot'

Which foot: the international foot (0.3048 meters) or the US survey foot (1200/3937 meters)?

* https://oceanservice.noaa.gov/geodesy/international-foot.htm...

* https://www.nist.gov/pml/us-surveyfoot

Here's another one: which weighs more, a pound of gold or a pound of feathers? It's a pound found of feathers that weighs more.

This is because gold (and other precious metals) are measured in the Troy system, where one pound is the equivalent of 373g:

* https://en.wikipedia.org/wiki/Troy_weight

Whereas most other things use the Avoirdupois system, where one pound is the equivalent of 454g:

* https://en.wikipedia.org/wiki/Avoirdupois

Now, which weighs more: one ounce of gold or one ounce of feathers? In this case the gold, because the Troy system has 12 ounces per pound, so you'd have 31g, while the Avoirdupois has 16 ounce per pound, so you'd have 28g of feathers.

And remember that there are British/Imperial units and US customary units, so things like cups are different between countries:

* https://en.wikipedia.org/wiki/United_States_customary_units

So when you say "We should have used unit X as the basis of Y", whose unit of X should have been used?

Also worth noting that the US has used the metric system as the foundation of the customary system since 1893:

> Under the Mendenhall Order of 1893, metric standards, developed through international cooperation under the auspices of BIPM, were officially adopted as the fundamental standards for length and mass in the United States, though some metric standards were used in practice before then. The definitions of United States customary units, such as the foot and pound, have been based on metric units since then.

* https://en.wikipedia.org/wiki/Metrication_in_the_United_Stat...

Reply on news.ycombinator.com