Whatever you grew up with is going to be more intuitive. There’s nothing inherently intuitive about either temperature scale. And 0=freezing is an important temperature in everyday life - it’s the difference between rain and snow, ice on roads vs. puddles, refrigerated food vs. frozen, etc. There’s nothing special about 0F or 100F.
True, but I would concede that for climate control, 1F precision is sufficient, but 1C can be marginal. AC thermostats in Japan typically have 0.5C steps, and I think that’s very much desirable. I’ve been in situations where setting it to 25.0 seemed too cool and 26.0 too warm.
Standard gravity was originally based on the acceleration of a body in free fall at sea level at a geodetic latitude of 45° but was set to a consent.
Without standard gravity you can’t actually accurately derive the newton (kg⋅m/s), pascal (kg⋅m^−1⋅s^−2), watt (kg⋅m2⋅s^−3) or other units.
This is part of the reason they want to find a replacement for the IPK, which is a physical artifact. Because you need to know air pressure to offset buoyancy when calculating weight to correct for standard gravity. It is a circular dependency with the last remaining physical prototype.
If representing 1/3 of a meter as 0.33333 m results in rounding errors that actually matter to you, then you should be representing it as 0.333333 m instead.
And using different units would have made no difference in those examples.
Most people don’t use measurements in any serious fashion, so the system doesn’t matter very much to them.
The people who do (cooks) are either too invested in the US system and are unaware of the advantages of Metric (scaling, for one thing), or like my wife, actually kind of clueless about measurements in general.
I wonder how many people actually understand that fluid ounces and weight ounces are unrelated?
Gas marks are rare to non-existent in the US. I personally have never seen one on this side of the pond. The reason that they exist in the UK and Commonwealth countries is that prior to the conversion, everything was done in Fahrenheit, so recipes were in Fahrenheit. After the conversion, the recipes didn’t change, but converting 350 to 177 was strange at best, so they just added gas marks. Gas marks are a Fahrenheit scale that starts at 1 = 275 and then you add 25 degrees for every mark. 2=300, 3=325,4=350, etc. In the US, we just use the Fahrenheit scale, so you probably bake using 4 and 6 most commonly. We would bake at 350 and 400 respectively. Canadians solved this problem by largely still using Fahrenheit ovens. Canada is a strange place, but a good place.
4 inches vs:
0.333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333 M
4 inches is exact and the metric equivalent is still an approximation.
But once again the issue is that 0.1 *does not exist *in Floating-Point.
You can stay with rational numbers like 1/3, but a main point of a decimal based system is to get rid of those fractions.
I disagree completely, and it has little to do with familiarity. Sure, I could get used to C, but it is less useful as a scale for living (as opposed to measuring water freezing and boiling). Most of us live our lives between 0 and 100 F. If it’s below 0F, it’s damn cold. If it’s above 100F, it’s damn hot. In between, each decile has an easily observable and relatable amount of coolness or warmth. That simply doesn’t happen with the C scale.
I’m probably not going to convince you about F, and you are definitely not going to convince me about C.
In the real world,those 4 inches are also an approximation.
All measurements have errors, unless you’re measuring spherical cows, you will never ever have exactly 4 inches.
I’ve never thought about the decile thing. How do Celsius countries do it? If asked the temperature, we’ll frequently give the decile. “It’s going to be up in the 90s today.” Or “It’s going to be in the 40s tonight.” What do you Celsius people do? Just refer to the actual projected high or low? If I say it’s going to be in the 70s, that means something to most people. Do you say the low 20s? What if you’re trying to differentiate between the low and the high decile? Does that just not happen? I’m actually interested.
Hmm. Room temperature is a nebulous standard; there is no way it is precisely defined to within 1 degree or 0.5 degree (and does the ventilation system even keep the air mixed that well?) Recommended temperatures are always given as ranges depending on the season, humidity, and possibly other considerations. Also, if 25.0 seems too cool, I would suggest wearing at least a T-shirt.
This is a completely ridiculous argument, and has nothing to do with units. All measurement systems will have numbers that can’t be represented exactly in binary. So what? It’s possible to represent them with any required accuracy.
To add to this, because it is not a reason to not adopt SI units but because the implications are important.
Floating point is a form of finite mathematics and the distributive law and Associative law for multiplication do not always hold in finite math due to these representation errors. This breaks a lot of assumptions by compilers and programmers.
a * ( b * c ) may not equal ( a * b ) * c
a * ( b + c ) may not equal ( a * b ) + ( a * c )
These assumptions not holding will lead to compounding inaccuracies well beyond the issues caused by the normal loss of precision with floating point.
Decimal based systems, which have intrinsic representation errors can be really problematic because of their non-existence in base 2 when converting to smaller prefixes.
This is why banks and other critical users resort to using computationally expensive decimal data types, and why your online loan amortization tool, phone app or calculator never seems to match your statement.
The pound is 16 ounces, the mile is 5280 feet, where the base prefixes in the Metric system are hard tied to a number that can’t be represented in binary in the metric system. Or in other words, any move to a smaller sized unit has a rounding event.
exa E 1000000000000000000 10^18
peta P 1000000000000000 10^15
tera T 1000000000000 10^12
giga G 1000000000 10^9
mega M 1000000 10^6
kilo k 1000 10^3
hecto h 100 10^2
deca da 10 10^1
(none) (none) 1 10^0
deci d 0.1 10^−1
centi c 0.01 10^−2
milli m 0.001 10^−3
micro μ 0.000001 10^−6
nano n 0.000000001 10^−9
pico p 0.000000000001 10^−12
femto f 0.000000000000001 10^−15
atto a 0.000000000000000001 10^−18
Are you really suggesting people don’t use the prefixes of the system? Once again, I am not anti SI, but I can go from a mile to an inch and even to a mil with no loss in precision, but I can’t do the same for a meter to a centimeter.
SI isn’t a bad system, but it has very real limitations with very real implications.
Yes. So you can write 6 feet 4 inches, and it’s exactly 6 and 1/3 feet.
So?
Yes, we have a name for a 1/12th of a foot. It’s called an inch. How does giving 1/12th of a foot a name help us avoid rounding errors?
Any decimal expansion rounding error that applies to meters also applies to feet, inches, miles, furlongs, parsecs, light years, chains, rods, yards, leagues, or fathoms.
Again, you’re not understanding the problem. You can write 4 feet 4 inches, and get exactly 4 and 1/3 feet. That doesn’t give you a precise decimal expansion of 4.333333333… feet. It just gives you another way to write 4 1/3 feet. If you need to keep that fraction, then keep the fraction. You can use precise fraction with any unit. One third of a meter. One third of an inch. One third of a yard. One third of a furlong. One third of a light year. The fact that some of these fractional units are equivalent to integer values of other units is completely irrelevant! It doesn’t matter that there’s a vernacular name for 1/3 of a yard, but no name for 1/3 of a light year. If you absolutely positively must preserve that exact fraction, then preserve that goddam exact fraction.
Again, the only way you can preserve those exact fractions in a traditional system is to…preserve those exact fractions, and write them out as fractions, and then give them a special name so that instead of a fraction you’ve got an integer. Writing six feet four inches is EXACTLY THE SAME THING as writing six and one third feet.
Metric has nothing to do with it. Seriously. The only thing is that there are traditional names for some fractions of some traditional units, and there are not such traditional names for metric units. And? The fact that there is no name for one third of a meter is not a problem, just like the fact that there is no name for one third of a chain or one third of a rod or one third of a light year.
And of course, all these standardization of traditional units happened by happenstance. Different people used different traditional measures for different reasons, and it didn’t matter how many firkens were in a hogshead, because there was no precise definition of a hogshead anyway. Then along came the industrial revolution, and we had to measure precisely how big a hogshead was, and then how many drams that equaled, and maybe we’ll fudge the sizes so the relationships are integer multiples of each other.
Yes, there is no exact binary representation for 1/10th. There is also no exact binary representation for 1/7th. If those French Revolutionaries would have known that we were going to be using binary, they should have made metric hexadecimal rather than decimal, but they didn’t. However, the fact that we have decimal names for things in the metric system doesn’t make metric unsuited to binary representation, because traditional measurements have the exact same problem, you can’t represent 1/10th of an ounce in binary either. Oh, there’s no traditional named unit that’s 1/10th of an ounce? And there is one that’s 1/6th of an ounce? So what? How does that help? Oh, you can keep track of exact fractions using mixes of two named units and thus get integer representations of fractions? Guess what else gives you integer representations of fractions? Integer representation of fractions.
Speaking only for myself, and from a Mediterranean-type climate that never freezes: I think most often I say something like “it’s going to be about 25 today” or “it’s going to drop to about 12 degrees tonight”. For higher temperatures, often something like “over 30”, “over 35”, “over 40”, as the case may be. Similarly for cold temperatures, “it’s going to be under 10 tonight”.
Is the foot, the mile or the inch based on 1/7? No.
But yes, like I said up thread, had they ignored the thumbs and been octal, or even duodecimal or hexadecimal some modern problems would be easier to solve.
There are some problems like calculating the stable Lagrange points for the asteroid belt with the influences of Jupiter and Mars which are unsolvable in practical compute timelines due to these limitations. JPL has even reduced the precision of their ephemerides due to compute time because they have to use arb precision to get more accurate results which is not hardware accelerated. It may be easy to say just move to another number base, but we have a century of proofs often in non-linear domains that does not allow for this without a lot of work.
But lets be clear, I was responding to people making the claim that there is no reason for someone to use the customary units, but when you are a machinist building an airplane that is a few meters across, and your tolerances need to be within 25.40 μm (1 mil) if you use the customary units your job is a lot easier.
As I have stated multiple times that all systems have their limitations, but some systems have intrinsic limitations. So I don’t know what your post relates to my claim anyway.
All systems have flaws, and this is one of the Metric system, and as I stated before the dollar has similar challenges in modern computer based world.The post I was replying to was pointing out that there are lots of use cases where this loss in precision is problematic, and that it can’t just be ignored.
But it’s not constant; the whole weather report is based on its variations, yet you speak as if it were constant.
And the 100ºC also requires distilled water, which isn’t normally what’s used for cooking.
How do you give someone’s approximate age? The same expressions are used in Spanish. Couple of examples from Spain: Hoy hará… veintitantos (today it’s… twentysomething). ¿A cuánto estamos? Pues no sé, pero diría que veintimucho. (What’s the temp? Dunnow but I’m guessing twentylots.) And “under 36” or “over 36” are of course important, that being more or less body temp (actual body temp will vary by person, time of day and health status). ¿Qué hace ahi fuera? ¡Más de 36 y subiendo! (how hot is it out there? Over 36 and climbing!)
And ? 1m = 39.3701 inches = 3.28084 feet. Y’all don’t have an **exact **measure of this specific, arbitrary distance wot my unit system is based on, so yours is pure shite !1!!
(I jest, of course, but you get my point, right ?)
