Metrication

[kalvado]

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

Low 20s - mid-20s - upper 20s
It's really something that can be worked out as desired.
As for too crude... Once upon a time I had an industrial temperature controller doing bedroom temperature, with resolution pushed down down to 0.01C. It was a bit worse realistically, maybe 0.07-0.1C.
Try to beat that with your crude fahrenheit machine...

[1995hoo]

The bigger issue is that, as usual, Poiponen never responds to what someone says. I mean, look at his reply to JN Winkler. Regardless of whether you agree or disagree with what JN Winkler said, Poiponen's response is about as irrelevant as it could possibly be–arguably, he's just repeating what JN Winkler said.

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

It can be said that temperatures in e.g. a state or region are in 30s Celsius, when coolest is 31 C and hottest 39 C. And 0s would be pronounced "zeros" and 10s "tens".

[Scott5114]

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

Low 20s - mid-20s - upper 20s
It's really something that can be worked out as desired.

And in Fahrenheit you can also do low 70s - mid-70s - upper 70s and it's more precise there too.

Look, I love the metric system, but Celsius is just a stinker of a unit. I don't really see why it's considered part of the metric system, anyway; it's not like kilocelsiuses or centicelsiuses or whatever are a thing. The central unifying idea of the system doesn't apply to Celsius, so why is it included as one of its measures?

[Poiponen13]

The bigger issue is that, as usual, Poiponen never responds to what someone says. I mean, look at his reply to JN Winkler. Regardless of whether you agree or disagree with what JN Winkler said, Poiponen's response is about as irrelevant as it could possibly be–arguably, he's just repeating what JN Winkler said.

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

It can be said that temperatures in e.g. a state or region are in 30s Celsius, when coolest is 31 C and hottest 39 C. And 0s would be pronounced "zeros" and 10s "tens".

I just clarified that the "10s", "20s", "30s" can be used in Celsius too, in addition to Fahrenheit.

[Scott5114]

The bigger issue is that, as usual, Poiponen never responds to what someone says. I mean, look at his reply to JN Winkler. Regardless of whether you agree or disagree with what JN Winkler said, Poiponen's response is about as irrelevant as it could possibly be–arguably, he's just repeating what JN Winkler said.

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

It can be said that temperatures in e.g. a state or region are in 30s Celsius, when coolest is 31 C and hottest 39 C. And 0s would be pronounced "zeros" and 10s "tens".

I just clarified that the "10s", "20s", "30s" can be used in Celsius too, in addition to Fahrenheit.

So? Jonathan was saying that if you do that, you cannot speak with the same precision that you can in Fahrenheit. You didn't bother to respond to the main idea of his post.

[JayhawkCO]

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 — a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

Low 20s - mid-20s - upper 20s
It's really something that can be worked out as desired.

And in Fahrenheit you can also do low 70s - mid-70s - upper 70s and it's more precise there too.

Look, I love the metric system, but Celsius is just a stinker of a unit. I don't really see why it's considered part of the metric system, anyway; it's not like kilocelsiuses or centicelsiuses or whatever are a thing. The central unifying idea of the system doesn't apply to Celsius, so why is it included as one of its measures?

The central unifying idea of the system basically does apply to Celsius though -- being base 10. Water freezes at 0° and boils at 100°. Yeah, you can't use the prefixes to scale up or down, but there's no larger or smaller measures of temperature in Fahrenheit either.

[Scott5114]

I suppose, but I always considered the killer feature of the metric system to be the prefixes. If you're not going to have prefixes either way, I would rather use the smaller units.

(Just out of curiosity, how many US-based pro-Celsius people in this thread actually use it?)

[JayhawkCO]

I'm not necessarily pro-Celsius. From a physics perspective though, a lot of the formulas utilize temperature in units of Kelvin, which is basically just absolute-zero adjusted Celsius. (Fun fact #1: Fahrenheit has its own absolute-zero scaled version and it's called Rankine.) (Fun fact #2: Kelvin doesn't use °. It's just 273 K.) So, there are times in my life where I've used it, and plenty of others would now.

When I'm overseas, I've gotten relatively adept at identifying what temperature it is outside in Celsius, basically just remembering that 37° C is body temperature and adjusting from there.

[Scott5114]

I was about to say I would agree it makes sense to use Celsius/kelvin when working in a context with lots of units derived from kelvin...but I just looked it up and there aren't any, other than Celsius. But I do agree that in scientific contexts where you're using formulas and such that depend on kelvin/Celsius, it makes sense to use kelvin/Celsius, because converting between units is silly.

But that's not the context I'm in when I'm setting my thermostat or deciding what to wear.

[kalvado]

I've mentioned this twice before in this thread, but 1°C is about the smallest meaningful difference. People set their thermostats to 68°F or 70 or 72, not 69 or 71 – a difference of 2°F ≈ 1°C.

There are plenty of people who don't set their thermostats in integer multiples of 2° F--65° F and 75° F are popular values, for example.

Using Celsius for ambient temperatures also sacrifices the convenience of talking about a comfort range as a decade group.  We generally recognize that the eighties and nineties are different in comfort level, but in Celsius they are all in the thirties, which forces us to speak with false precision (e.g. "about 30° C" rather than "in the eighties") to make the distinction.

Low 20s - mid-20s - upper 20s
It's really something that can be worked out as desired.

And in Fahrenheit you can also do low 70s - mid-70s - upper 70s and it's more precise there too.

Look, I love the metric system, but Celsius is just a stinker of a unit. I don't really see why it's considered part of the metric system, anyway; it's not like kilocelsiuses or centicelsiuses or whatever are a thing. The central unifying idea of the system doesn't apply to Celsius, so why is it included as one of its measures?

Look, imperial, or whatever it's called - isn't that bad. People can live with feet and pounds, use SAE wrenches - but Fahrenheit is just a stinker of a unit.
And well, Kelvin is the metric unit - and millikelvin and microkelvin are in fact used.

As for unifying ideas, primary one is not suffixes - I definitely saw kilopounds used, for example. Key idea is minimizing number of random constants used. As a result... For example, electric power  and mechanical power come out in same units without any conversions.  That's not really the case for US units. 
Kelvin - and celsius, as one derived from Kelvin - is a stepchild here as a random constant is added for no good reason (ok, not a very good one). It's mostly to scale things down into digestible range (I am more or less comfortable saying that room temperature is 26 meV - but I am a minority, and even then I don't have natural scale in my head. Not to mention that electron-volt is a non-standard unit as well..

[JayhawkCO]

I was about to say I would agree it makes sense to use Celsius/kelvin when working in a context with lots of units derived from kelvin...but I just looked it up and there aren't any, other than Celsius. But I do agree that in scientific contexts where you're using formulas and such that depend on kelvin/Celsius, it makes sense to use kelvin/Celsius, because converting between units is silly.

But that's not the context I'm in when I'm setting my thermostat or deciding what to wear.

I think you'd adjust just fine though. You're a smart dude. Yeah, it's not quite as granular as Fahrenheit, but a little practice with it, and you'd be a-okay. Also, if you ever get the hell out of Oklahoma and get up to Colorado, you need to bring every kind of layer with you anyway. So whatever the temp is is somewhat irrelevant.

[1995hoo]

I suppose, but I always considered the killer feature of the metric system to be the prefixes. If you're not going to have prefixes either way, I would rather use the smaller units.

(Just out of curiosity, how many US-based pro-Celsius people in this thread actually use it?)

I'm not sure I fall within the camp listed in the last paragraph, but I do use Celsius for two things, aside from the pre-travel thermostat display thing I noted earlier:

(1) My wife has an Instant Pot fast-boil teakettle that boils up to 1.5 litres of water far more quickly than setting it in a pot on the stove, so I use the teakettle to accelerate the process of boiling water when we're making any sort of pasta. Its display happens to be set to Celsius and neither of us remembers how to change it to Fahrenheit. It doesn't much matter because all we need is for it to hit 100°, but even if we didn't know that was the relevant number, the thing beeps when the water is boiling.

(2) We have a recipe for doing corn on the cob in the oven and the recipe is written in metric units. It calls for wrapping it in foil and baking it at 200° Celsius. So I simply hit the "°F/°C" button on the oven to switch the display to Celsius and then set the temperature accordingly. Before anyone says "just memorize the conversion," that doesn't work–it converts to 392°F and the oven doesn't let you set it to that temperature if the display is in Fahrenheit. (I doubt it would matter much if I just set it to 400°, but since I can set it to the prescribed number, I see no reason not to do so.)

[kalvado]

I'm not necessarily pro-Celsius. From a physics perspective though, a lot of the formulas utilize temperature in units of Kelvin, which is basically just absolute-zero adjusted Celsius. (Fun fact #1: Fahrenheit has its own absolute-zero scaled version and it's called Rankine.) (Fun fact #2: Kelvin doesn't use °. It's just 273 K.) So, there are times in my life where I've used it, and plenty of others would now.

When I'm overseas, I've gotten relatively adept at identifying what temperature it is outside in Celsius, basically just remembering that 37° C is body temperature and adjusting from there.

From physics perspective, it's either some flavor of T1/T2, which would work equally in Rankine and in Kelvin; or k_BT - instantly converting out of Kelvin.
What really matters for unit sake is almost infinite datasets of various properties - heat capacity, resistance vs temperature, melting/boiling points, expansion coefficients etc etc etc.   

[TXtoNJ]

As someone who goes back and forth on a regular basis, Celsius is much easier and more intuitive once you're used to it. Nobody's switching their thermostat either, if it's not some ancient model from the '70s.

Everyone just thinks of the decadal Celsius values in terms of "upper" and "lower". The 20s are nice, but the upper 20s are warmish-nice, and the lower 20s are coolish-nice.

[NWI_Irish96]

As someone who goes back and forth on a regular basis, Celsius is much easier and more intuitive once you're used to it. Nobody's switching their thermostat either, if it's not some ancient model from the '70s.

Everyone just thinks of the decadal Celsius values in terms of "upper" and "lower". The 20s are nice, but the upper 20s are warmish-nice, and the lower 20s are coolish-nice.

I like the idea of round numbers being the freezing and boiling points of water, but not only having 100 degrees between them.

Maybe a modified Celsius with 0 as the freezing point and 200 as the boiling point.

[kalvado]

As someone who goes back and forth on a regular basis, Celsius is much easier and more intuitive once you're used to it. Nobody's switching their thermostat either, if it's not some ancient model from the '70s.

Everyone just thinks of the decadal Celsius values in terms of "upper" and "lower". The 20s are nice, but the upper 20s are warmish-nice, and the lower 20s are coolish-nice.

I like the idea of round numbers being the freezing and boiling points of water, but not only having 100 degrees between them.

Maybe a modified Celsius with 0 as the freezing point and 200 as the boiling point.

you're 150 years late for the party

[TXtoNJ]

As someone who goes back and forth on a regular basis, Celsius is much easier and more intuitive once you're used to it. Nobody's switching their thermostat either, if it's not some ancient model from the '70s.

Everyone just thinks of the decadal Celsius values in terms of "upper" and "lower". The 20s are nice, but the upper 20s are warmish-nice, and the lower 20s are coolish-nice.

I like the idea of round numbers being the freezing and boiling points of water, but not only having 100 degrees between them.

Maybe a modified Celsius with 0 as the freezing point and 200 as the boiling point.

When you stop thinking of Celsius as a scale that has 0 at freezing and 100 and boiling, and instead start thinking of it as a scale that measures the effective range of temperatures on Earth from -50 to 50, it really does illuminate things about how temperature works. Having 25 be the division between generally cooler temperatures and generally warmer temperatures is just too convenient.

[kphoger]

Why all of the 11's in the Imperial distance measurements?  5280=480 x 11.  43560= 11 x 11 x 360 or 66 x 66 x 10.  Again, why?

It was to avoid over-taxing people in late-16th-Century England.  It was either decrease the length of the rod by ¹/₁₁ or increase the number of rods per furlong by ¹/₁₀.  Keep in mind that the rod in those days wasn't just a theoretical distance, but an actual surveyor's rod of a government-mandated length, and the number of rods your property measured determined how much tax you owed.  The former solution would have increased the numerical size of everyone's property and therewith the amount of property tax owed, so the government went with the latter solution instead to avoid–you know, a nationwide riot.

In other words, before the year 1593, those 11s were 10s.

[TheHighwayMan3561]

The bigger issue is that, as usual, Poiponen never responds to what someone says. I mean, look at his reply to JN Winkler.

There does seem to be a habit of many outside observers looking at some things the US does and going "no one else does that so that automatically makes it dumb and wrong". Poiponen can't really articulate his arguments beyond that idea.

I'm not saying that to promote or to degrade imperial measurements (or other things like our date format that foreigners like to pick on), but I mean, I don't know why some people in Finland or France or wherever get so aggravated over me buying a half-gallon of milk.

[kkt]

Why all of the 11's in the Imperial distance measurements?  5280=480 x 11.  43560= 11 x 11 x 360 or 66 x 66 x 10.  Again, why?

It was to avoid over-taxing people in late-16th-Century England.  It was either decrease the length of the rod by ¹/₁₁ or increase the number of rods per furlong by ¹/₁₀.  Keep in mind that the rod in those days wasn't just a theoretical distance, but an actual surveyor's rod of a government-mandated length, and the number of rods your property measured determined how much tax you owed.  The former solution would have increased the numerical size of everyone's property and therewith the amount of property tax owed, so the government went with the latter solution instead to avoid–you know, a nationwide riot.

In other words, before the year 1593, those 11s were 10s.

That's really interesting.  I thought England was thoroughly surveyed by order of William the Conqueror and producing the Domesday Book.  Why 1593?  And if you don't like taxes being too high, wouldn't it be easier to, I dunno, lower the tax rate instead of changing the definition of a rod?

[kphoger]

Weighing in on the Celsius topic...

It's quite useful to know what temperature water freezes at.  For example, it's handy to be able to glance at my car's outside temp readout to know how concerned I should be about hitting icy patches.  Or it's handy to look at the cheap thermometer in the fridge and know how far above freezing it is in there.  So, in that sense, Celsius being pegged to 0°C makes sense.

But I have basically zero need to ever know what temperature water boils at.  What am I, a lab scientist?  Literally the only thing I need to worry about boiling over on a daily basis–the coolant in my car's engine–is mixed with antifreeze and kept under pressure and therefore has a much higher boiling point than 100°C anyway.

So let's just shift the Fahrenheit scale up by 32 degrees and call it good.  Freezing is now 0°F, room temperature is now in the 35°F to 40°F range, hot Midwestern summer days are now 70°F, and the hottest summer temperatures recorded on earth now just barely crack the 100°F mark.



In general...

Converting units more easily with the Metric system is only true if you're comfortable ending up with decimals as your final answer, or if you only really need to divide by ten or five or two.  This is great for a lot of things.  But not for everything.

Converting units more easily with the US Customary system is true if you're also frequently dividing by threes and fours and sixes and eights.  The threes and sixes are the real issue.

In the end, people should use whichever system suits their needs the best.  Which–*gasp*–they generally already do!  Living in a part of the country where the entire street and road system for hundreds of miles around is based on US Customary units, I would hate having to switch to kilometers (even though I'm perfectly find using kilometers in Mexico, where that mile-grid pattern isn't a thing).  But I could switch from buying milk by the gallon to buying it by the liter, no problem:  I never need even a quart at any one time anyway.  Like Scott, I often measure paper by the centimeter;  but I sometimes measure it by the inch instead, because it just depends on how the fractions or decimals work out for that particular project.  And measuring body height by the meter just seems silly to me, because almost every adult I know is between 1½ and 2 meters tall.

[kkt]

Body height is usually quoted in centimeters in the world outside the U.S....

[kphoger]

Why all of the 11's in the Imperial distance measurements?  5280=480 x 11.  43560= 11 x 11 x 360 or 66 x 66 x 10.  Again, why?

It was to avoid over-taxing people in late-16th-Century England.  It was either decrease the length of the rod by ¹/₁₁ or increase the number of rods per furlong by ¹/₁₀.  Keep in mind that the rod in those days wasn't just a theoretical distance, but an actual surveyor's rod of a government-mandated length, and the number of rods your property measured determined how much tax you owed.  The former solution would have increased the numerical size of everyone's property and therewith the amount of property tax owed, so the government went with the latter solution instead to avoid–you know, a nationwide riot.

In other words, before the year 1593, those 11s were 10s.

That's really interesting.  I thought England was thoroughly surveyed by order of William the Conqueror and producing the Domesday Book.

Before the 16th Century, England used both Roman and Anglo-Saxon/Germanic units of measurement (the former, of course, predating William the Conqueror).  Because surveyor's rods were the basis of determining taxation, its length was standardized in the 16th Century by Henry VIII, who had recently come into a whole lot of land to be granted or sold to others, land that suddenly needed to be assessed for taxation and sale price–you know, land that had formerly been the property of the Catholic Church and now... ummm... wasn't.

Even though the North Germanic (Belgic) foot had been a defunct unit of measurement since around 1300, the length of the surveyor's rod at the time of Henry VIII was still based on it.  Originally, using the North Germanic system, there were 40 rods to the furlong, and 15 feet to the rod.  England had since ditched the North Germanic foot in favor of one based on the Roman foot, and it was slightly smaller:  to be precise, the North Germanic foot had been 10% longer than the newer English (née Roman) foot.  The length of the foot had changed but, understandably, the length of the surveyor's rod had not.  After all, landowners' deeds the country over had been drawn up by the acre, which was an area equal to one furlong by four rods, and by the square-rod/rood:  change the length of the rod, and would you want to be the one responsible for doing all that math and rewriting all the deeds?

There were then, depending on how you calculated a furlong, two possible answers.  16.5 English feet to the rod, 40 rods to the furlong = 660 feet?  Or 625 feet to the furlong, as already established by law?  In 1593, the government officially maintained that there were 8 furlongs to the mile, and 40 rods to the furlong, but what to do about how many feet to the rod?  (1) They could shorten the length of the rod, so that there could be 15 English feet to the rod (as originally), 40 rods to the furlong, 8 furlongs to the mile, and therefore 4800 feet to the mile.  Or (2) they could keep the higher number of feet per rod, and therefore establish a greater number of feet per furlong and feet per mile, giving 660 feet per furlong and 5280 feet per mile.  Option 2 meant nobody had to rewrite the whole country's deeds, nobody had to make new surveyor's rods, nobody had to pay higher taxes, ...  so they went with that.  16½ feet to the rod, 40 rods to the furlong, 8 furlongs to the mile.

Why 1593?

Because that's when Elizabeth I passed that particular Weights and Measures Act:  35 Elizabeth c. 6 (1593)

And if you don't like taxes being too high, wouldn't it be easier to, I dunno, lower the tax rate instead of changing the definition of a rod?

I'm no expert on how late-16th-Century British tax rates, but my assumption is that it wasn't simply a matter of "tax them at 15.29% instead of 16.82%" or whatever.  But as I elaborated above, they were revamping the measurement system anyway.

[kphoger]

Body height is usually quoted in centimeters in the world outside the U.S....

But notably not in Canada, which supposedly "went metric".

[Brandon]

Body height is usually quoted in centimeters in the world outside the U.S....

But notably not in Canada, which supposedly "went metric".

Supposedly.  I just bought a carton of milk for the morning here in British Columbia.  It's marked in millilitres, but the size happens to be exactly one US customary quart.  BTW, it's a local milk in a Canadian-made carton.