Hacker News Reader

Not this crap again.

Fahrenheit maps to your human experience because it's what you grew up with.

To me, Celsius maps to my human experience, because it's what I grew up with.

That's it. That's all it is, and that's the only reason you think it maps better.

No. Fahrenheit really is better than Celsius regardless of your reluctance to admit it.

1. 0 Fahrenheit is close to the typical minimum temperature, and 100 close to the typical maximum temperature, which can be tolerated by humans. Celsius 0 and 100 aren't even remotely close to this: it's surprisingly awful.

2. Fahrenheit has nearly twice the integer resolution of Celsius. This is particularly helpful in describing body temperature.

3. Celsius seems like it'd be useful in cooking but it is not really. Sure, 0 and 100 describe the boiling point and freezing points of water -- in Paris only -- but these are phase changes in water. Nobody says "now hold the temperature at 100C", because water naturally fixes to 100C the entire time it is undergoing its phase change. You don't need to measure it. And Fahrenheit's higher resolution is again very helpful.

4. Scientists don't use Celsius or Fahrenheit. They use Kelvin. Both Celsius and Fahrenheit convert to Kelvin with a linear transformation.

I have no problem believing that Americans find the Fahrenheit system intuitive, but you are failing to see that for metric users, Celsius is 100% intuitive. I know exactly what sort of clothes to wear for each gradation (assuming it's not something insane like -60C in which case I'm staying home anyway), whereas I genuinely cannot do the same calculus with just Fahrenheit, even taking into account the 0-100 concept of Fahrenheit, which is in itself busted in various regions from Canada to Saudi Arabia. I have to constantly remind myself of what Fahrenheit means, because I never grew up with it.

This is a theory of mind issue.

> for metric users, Celsius is 100% intuitive.

For metric users, Celsius is 100% familiar. Intuitive means that there are a bunch of hints that can give you an idea of values that you don't know. The fact that 0° Fahrenheit is about as cold as you can easily stand and 100° Fahrenheit is about as hot as you can easily stand is something that gives you an intuition: if it's as hot as I can stand, it's about 100°, if it's as cold as I can possibly stand, it's about 0°. I can figure out intermediate values through converting my feelings into a percentage.

You can of course do that with Celsius too, but Celsius hasn't been built around a range of human feeling, it's been built around the phase changes of water. You just have to remember the human comfort range in Celsius, do all the same calculations as above, multiply the result by the difference between your low human range and your high human range, then add that to your low human range to get an answer. There's no help, and it might as well be arbitrary.

This is a post-facto rationalization for two reasons.

The first is that the Fahrenheit scale wasn't built around a range of human feeling, but originally around a substance whose temperature is easy to stabilize. Following this, the human temperature was set at 96 (later adapted to 98.6) and the melting point of ice at 32 to have 64 gradations to help with measurements linked to specific hardware. If it was designed around human feeling, then 98.6 and the freezing temperature of brine is hardly helpful.

The second is that if it were intuitive on its own, people who don't use the system regularly would still be able to remember the logic, but that's not what typically happens. Every time I need to think in Fahrenheit (which tends to only be online debates about this very topic), I can't picture anything clearly even with the purported 0 to 100 concept. It is a subjective appraisal of temperature that can't help me convert feelings into anything useful because the limits are still arbitrary. If -17.7C is the lowest a person can possibly stand, then Canadians are gonna be iced out. If 37.7C is the highest a person can possibly stand, Saudis are cooked. The number of regions that hold a range that's appreciably close to that 0 to 100 is quite small.

> The first is that the Fahrenheit scale wasn't built around a range of human feeling

How and why Fahrenheit was developed is completely immaterial: what is material is whether and how its range is of value to its users now.

> Every time I need to think in Fahrenheit (which tends to only be online debates about this very topic), I can't picture anything clearly even with the purported 0 to 100 concept.

Your argument is "because I'm not used to Fahrenheit, it must not offer an advantage to people used to it"? Really? C'mon.

> Your argument is "because I'm not used to Fahrenheit, it must not offer an advantage to people used to it"? Really?

But that's your argument for Fahrenheit...

> 4. Scientists don't use Celsius or Fahrenheit. They use Kelvin. Both Celsius and Fahrenheit convert to Kelvin with a linear transformation.

But the slope of the Celsius<->Kelvin is 1.

Celcius is more useful in weather forecasting, because for a lot of people it's important to know whether the water outdoors will be frozen or liquid. Example - it's useful to know if there will be ice on road and it's useful to know if the plants outside will die due to ground freeze at night. Giving the temperature in C answers both questions automatically.

"Typical" maximum and minimum temperature as limits are meaningless, different places have significantly different minimum and maximum temperatures.

The resolution might be better for body temperature (I have my doubts, we measure to decimal point in C), but certainly not for cooking. Nobody needs to know the difference between 200 and 201 F in cooking.

> Both Celsius and Fahrenheit convert to Kelvin with a linear transformation.

C doesn't need multiplication tho.

What on earth is "the typical minimum temperature that can be tolerated by humans" though? There are people who seem fine living in parts of the world where it's regularly -30C (somewhat less than 0F), and many of us that would find anything much less than 0C (32F!) regularly to be pretty much intolerable. Resolution is an issue for measuring fever, so we use a single decimal point (36.8 is normal, 37.1 is elevated etc.). Other than that it just never comes up. Agreed that 100C being the boiling point of water isn't that useful for setting oven temperature etc., but at least you know something that's 100C is going to be as hot to touch as putting your hand under scolding water (and likewise 0C is going to be as cold as ice. You're willing to sacrifice that?). Kelvin is far easier to mentally convert to Celsius than Fahrenheit, and at any rate Celsius is still used extensively by biologists, behavioral scientists etc.

> many of us that would find anything much less than 0C (32F!) regularly to be pretty much intolerable.

Really? Barely long sleeve weather.

For you.

Here we see the difference between subjective and objective plainly expressed.

You just proved my point by missing it, I'm impressed.

> Fahrenheit has nearly twice the integer resolution of Celsius. This is particularly helpful in describing body temperature.

Yet medical practitioners insist on using Celsius in the most of the world, what fools!

> 0 Fahrenheit is close to the typical minimum temperature, and 100 close to the typical maximum temperature, which can be tolerated by humans. Celsius 0 and 100 aren't even remotely close to this: it's surprisingly awful.

I am very rarely exposed to 0 F, in fact, I actively go out of my way to avoid it by seeking shelter during blizzards while on mountain tops.

You're more likely to encounter 0 F in a continental country, but in my temperate maritime climate country, a populated locale hitting -10 C / 14 F is big news, and only rarely would it get to -4 C / 25 F where I live. In other words, I have very little use case for 0 - 32 F.

So its intuitiveness to you, is opaqueness to me. Because you grew up with Fahrenheit, and I grew up with Celsius.

Scientists do use Celsius and Fahrenheit sometimes, unless your definition of scientist excludes everyone who isn't a chemist or physicist.

TL;DR - you prefer Fahrenheit because you think in it. I prefer Celsius because I think in it.

I think this is a theory of mind issue. You are used to Fahrenheit intuition and living in the US, so people arguing against your perception seem to be bikeshedding or being willfully obtuse/superior.

I can perhaps relate my subjective experience to better explain what's going on. I grew up with Celsius temperatures, so temperatures like 27C or -5C tell me exactly what I should wear when going outside. For Fahrenheit, even though we are told that it maps to human experience, I have to consciously try to remember what these 10F, 20F temperatures mean and often forget. The intuition just isn't there at all.

For height, I'm used to the metric system, but I can also understand the imperial system in that specific area because of seeing so many discussions about it over the years.

I actually learned to cook while in an Imperial zone, and now I am more used to certain temperatures in F.

The reason people find it shocking, is that we are essentially forced to deal against our will with Imperial measurements, because the US is a powerful country. The annoyance has more to do with the sensation of being trapped by another culture and having to adapt to their way of life rather than the reverse.

Celsius is better for cooking, the Fahrenheit scale is better for hot weather, for colder weather, I prefer Celsius, around 0°C means water is freezing, so, snow, ice, etc...

Feet or inches are, I think, a bit more human friendly than meters, but only when used by themselves, that factor of 12 is just weird in a base 10 system. I know that the argument is that you can divide by 3 and 4 easily, but when it matters, metric countries simply make standard sizes a multiple of 12 (ex:120mm).

For volumes and weights, I find absolutely no redeeming feature to the imperial / US customary system. The kilogram is a perfectly fine unit, so is the liter, and that both are the same when we are weighing water is a very useful property. Because many things are around the density of water, you know that most things that fits in a 1L bottle weight around a kg, and you can often measure volumes with a scale without conversion.

Metric is not the most practical for every single situation, but it is more practical when used as a whole, that is, when you need to deal with with several quantities at the same time (ex: masses and lengths). And there are everyday situation where it is simply better in my opinion, like cooking.

About as reasonable a take as you can get.

Can you elaborate on the cooking though? I've seen that mentioned here a couple of times, and I'm not sure I follow. I'm from a Fahrenheit country and my wife from a Celsius country, and it's never caused any issues or confusion in the kitchen. But maybe you mean also for the volumetric measurements as well (or primarily)?

It is not about confusion, it is about convenience. I like the idea that water boils at 100°C since it is maybe the most important temperature in cooking. I like the fact that I can measure 500mL of water on a scale as 500g (and other liquids with a small error). Small details that make it more convenient in metric.

But there is something else, and I am asking you the question. Since you and your wife live in countries with different unit systems, isn't problem when doing each others recipes?

I mean I live in Europe, and if the recipe calls for 4 ounces of butter, I don't have a 4 ounces stick of butter, I have a 125g stick, which is not exactly that. A quart of milk is not exactly the liter I have. And my oven doesn't do 350°F, it does 180°C which is not exactly the same. In the end, I will have to adjust the recipe to my standards, and things may be lost (or gained) in translation, but it won't be the same.

For what it's worth, if looking for an English recipe I just add "UK" to my search to avoid American measurements. I find British recipes are also less likely to require some processed American ingredient unavailable in the EU. ("Add the miracle whip," what?)

> Celsius is better for cooking

Not really. Celsius maps to the boiling and freezing points of water in Paris. Live in Denver and you're screwed. And -- this is important -- nobody measures the freezing points and boiling points of water, because during its phase changes water holds to that temperature automatically.

Fahrenheit is also twice the resolution of Celsius. This is highly advantageoous for body temperature, medicine, cooking, and other functions. Celsius's integer resolution is horrible.

You give temperatures for cooking with 5 C resolution anyway, why would you need more than 0.5 C resolution?

Okay, highly advantageous for body temperature and medicine, only marginally advantageous for cooking.

Unnecessary precision is a disadvantage. That's why speed limits are 60 mph not 316800 feet per hour.

This sounds like trolling. Fahrenheit nearly perfectly nailed the intervals of importance in body temperature when it comes to child fevers, say. As no doubt you know well, 100 is "mild", 101 is "significant", 102 is "requires medicine to bring fever down" and 103 is "call the doctor". That's a huge difference in a range of 3 units. Now ask yourself what those values are in celsius.

I meant cooking. You wrote that Fahrenheit has "marginal advantage" in cooking, I responded that unnecessary precision is a disadvantage if anything.

As for body temperature the rule of thumb as I know it is sth like this:

    36 - fine
    36.6 - the theoretical perfect temperature (but it's a lie [1])
    37 - observe
    38 - go see doctor (but no rush)
    39 and higher - go see doctor immediately

These also depend on where you measure the temperature, mostly people measure in armpits here, if you do oral or rectal then they will differ. Remote measuring on your forehead is almost useless.

As for "requires medicine" - I don't remember having to make that decision, if I had 38+ C fever I just went to a doctor anyway.

[1] the actual perfect temperature changes by almost 0.5 C between people, time of day, time of month (for women) and other factors (measurement error). If you used thermal method of anticonception you'll know how variable it really is.

> Celsius is better for cooking

As far as I'm aware, there's no winner for cooking. I've never seen a recipe that required temperature more than as a value that could be provided in any units as long as you have equipment in those units (i.e. I'm not doing math with them). Putting "400°F" or "200°C" or "475 K" or some arbitrary "oven level 5" into my oven shouldn't really change how it comes out.

On the stovetop, I've never had an instruction with a temperature attached besides "interior temperature must reach X" or keeping something at a near-boiling temperature, which I've never needed a thermometer for, and 90% of the time I'm improvising without caring what the temperature is outside of "hot enough to cook the way I want but not too hot".

Sous-vide? It needs the temperature within a tight range, but the scale attached to the numbers doesn't matter as long as it's the one my thermometer uses. Double boiler? Doesn't even need you to use a thermometer to keep the temperature right. Give me a kitchen with everything in unpronouncible alien temperature units and I'd do just fine.

Temperature near 0°C is very important when we travel (meteo, ...). Temperature near 100°C is important when we cook (eggs, pasta, rice).

We can leave hands in water at 50°C. Above, it slightly start to cause burns.

This is a range that make a lot of sense IMHO.

0 degrees C is certainly nice for freezing, but Celsius loses its usefulness, in my opinion, at the upper ranges. Fahrenheit is okay though, because 32 isn't that hard of a number and 0 degrees F is a big milestone because it's a big (i.e., dangerous) milestone in outside temperatures.

> Temperature near 100°C is important when we cook (eggs, pasta, rice).

When does that matter on the scale being used? Someone else mentioned this, but I've never known anyone to be taking direct measurements of these things when cooking aside from baking. It's pretty clear when water is boiling.

> We can leave hands in water at 50°C. Above, it slightly start to cause burns.

When has that been useful and dependent upon a temperature measurement (and thus a temperature scale)?

You are putting far too much weight in arbitrary numbers, probably due to personal familiarity.

0°F is a big milestone? I'll claim 1°F should be the milestone instead, or -5.4°F, or 2.33333°F.

15°C or 12°C or something is an important temperature if you live in the tropics without a heating system.

> You are putting far too much weight in arbitrary numbers, probably due to personal familiarity

bmitc reacted to the comment that said "Temperature near 100°C is important when we cook (eggs, pasta, rice)." By asking when that info is important. They are not the one here that put weight in arbitrary numbers, at least in that last comment.

For the tea preparation, we shall use hot water below 100°C. french site: https://www.fauchon.com/fr/mag/conseils/quelle-temperature-d...

Or just; don't use boiling water.

I don't feel like either Celsius or Fahrenheit are any more natural than the other in most contexts. It just comes down to what you're used to. They're both relatively arbitrary.

I'm not sure why 0f is a big dangerous milestone in outdoor temperature. Is -18c/-1f really that different to -17c/1f? If anything, there's a much stronger argument that 0c is the dangerous milestone because that's when you get ice.

Really the only temperatures I care about on a day to day basis are like, freezing temperature (0c/32f), room temperature (20c/70f-ish) and various benchmark outdoor temperatures, like 30c/90f being hot and 40c/100f being potentially dangerous. You can +- a few C either way on any of those (except freezing) and it doesn't really make a difference. Cooking, even baking, is usually fine within around 10c. Sous vide is apparently a lot more sensitive but I've never done that.

Very occasionally I deal with something something like a kettle thermostat or boiler thermostat, where I might choose something like 100c/212f or 50c/120f respectively. I also use an IR thermometer for cooking pancakes, pizza and monitoring my wood stove - but neither Farenheit or Celcius are any more useful in that context.

I agree with all that for the most part. Having used Celsius at various times, I still feel Fahrenheit is better for outdoor temperatures because of the range, but yea, there's undeniably a bias there no matter how much I think I reason about it.

I think the idea that it is indeed relatively arbitrary probably summarizes a more accurate reaction to the comment I originally replied to.

Nope. Have you ever used a thermometer to measure the temperature of water for boiling pasta and rice? Ask yourself why.

Freezing point being 0°C is handy because it is important to know when outdoor temperature crosses this point for multiple reasons - water is too common and important substance to ignore its properties. In Fahrenheit freezing point is a hard to remember value (for someone who used to Celsius).

I see nothing special about -17.8°C +37.78°C range except that 37.78°C is very hot. But why 37.78 and not some other value? It is arbitrary. And -17.8°C is even more arbitrary - it is not the coldest weather in many northern regions.

My daily experience absolutely fits into the 0-50 range.

0C outside means the ground is hard to work with. 32F is such an arbitrary number when dealing is concrete reality outside.

Well mine and most everyone's is almost exactly 0-100 degrees F, and 50 degrees C is on the extreme.

Yes, 32 degrees F is arbitrary, but so are several of the milestones in Celsius.

I think my argument is that Fahrenheit is more natural most of the time, but both scales have their sweet spots (which makes sense given their relative nature). But I can't see any clear argument that says Celsius is superior all the time.

> Well mine and most everyone's is almost exactly 0-100 degrees F, and 50 degrees C is on the extreme.

You don't cook?

Their point was related to outside temperatures, which is what most people use temperatures for. And although I've never used Fahrenheit for anything, I don't think the 100c = boiling water is useful in day-to-day life.

I've used temperatures in the 60 something to 80 something range to measure the inside of food, my oven goes anywhere from 160 to 220 degrees, but water boils when it goes all bubbly. I know that that's at 100 degrees, but I don't do anything with that info.

I don't agree with bmitc's point that Fahrenheit feels more natural. But it seems like that anytime someone has something positive to say about Fahrenheit or imperial measurement, 10 different people need to tell them how wrong they are, and how there can't be anything positive about those systems of measurement. That doesn't seem necessary to me. Especially with temperatures where everyone just remembers a couple of numbers and associates them to a certain feeling or application, and whatever you've used your entire life will feel better or more natural.

Likewise anytime anybody says that Celsius feels natural as well, you also have 10 people tell them that they are wrong and Fahrenheit is a much better fit for human beings.

Now I think these discussions often go about cherry picking examples and just limit the discussion to outside temperatures is such a cherry picking.

The real power of the metric system comes from going all-in. You can't use everything but the temperature scale. If you would do that it would be even more obvious how disconnected the temperature scale is to the whole system.

Yeah, having metric everything makes it easy to do conversions and correlating with different properties of matter, you're using the language of science (and most of the world), with proper definitions in terms of reality: atoms (for kg), distance in terms of c*some_time, volume related to mass (1ml of water => 1g of water, you can get very precise with temperature as an input). Having a standard for orders of magnitude that's consistent (micro, milli, kilo, mega, giga) is just the cherry on top.

I use the knowledge of the temperature of boiling water to calculate "how much boiling water I need to add to a water in room temperature to have my prefect tea brewing temperature".

Considering that 100F is almost 40C just put your scale from 0-40. It's completely subjective.

0C wear a woolly hat

10C wear a jacket or sweater

20C Teeshirt and jeans weather

30C Teeshirt and shorts weather

40C Get in the shade and keep hydrated

50C You want to avoid this end of the scale

It's almost all defending against the pro-Fahrenheit folks making strange arguments promoting it, like somehow you can't measure your life unless it's in F.

> It's almost all defending against the pro-Fahrenheit folks making strange arguments promoting it, like somehow you can't measure your life unless it's in F.

I didn't make that argument. Fahrenheit is more natural in a lot of cases because of the range and delineation between whole number temperatures. Celsius is slightly less so but not all the time.

My entire point, which was lost in the bikeshedding is that it is not "shocking" that the metric system isn't used everywhere in daily experience because non-metric systems were literally invented to be useful and practical. Are they perfect? No. Neither is the metric system for daily human use.

The U.S. is a big place and thus has wide-ranging temperatures, and that probably contributes to resistance to getting rid of Fahrenheit, amongst other things and other measurement systems. The U.S. built a huge amount of modern infrastructure before anyone else and has never experienced a "rebuild" period. For non-engineering purposes, there's no downside to using non-metric units, and a lot of downsides in converting. As the back and forth here shows, there's probably a strong argument to be made that the measuring systems don't matter at all in an ideal world where it's free to convert entire countries back and forth between them. But for some reason, a lot of the world feels very superior for using the metric system.

Again. Not "shocking" as the original comment I replied to mentioned.

Fahrenheit was defined the way it was not because it was "invented to be useful and practical", but because a mixture of brine to set 0℉ was easy to concoct reliably. That's as close as it got to being "practical".

Every other argument seems to boil down to "I'm not familiar with it, so it's hard", "dealing with 10s is easier than 5s", and "Fahrenheit has more resolution because I've forgotten decimal numbers exist".

The actual superpower of metric is that all the units relate to each other in simple ways. For instance most liquids you and I deal with on a daily basis have a similar density to water, so if you've got 100g of it, it's going to be ~100ml. You don't need to deal with arbitrary artifacts of the measuring system beyond the unit prefixes.

"The US is a big place" , but the world outside the US using Celsius is even bigger, more populated, and with more temperature differences.

Fahrenheit being more natural is purely subjective, there's no good reason at all for it.

Keeping the imperial system is just a matter of convinience because Americans are used to it and changing everything is difficult. Also some people think producers of food and any objects to sell might lose some margin by converting and rounding to the next round number of the new unit

> Fahrenheit is more natural in a lot of cases because of the range and delineation between whole number temperatures. Celsius is slightly less so but not all the time.

You mean "I'm already used to something and I dislike other things."

It comes down to familiarity, but -20 to 40 degrees Celsius isn't that weird a scale, either. You're going to remember more or less arbitrary values for certain things anyway (too hot to be comfortably outside, cold enough to need a jacket, cold enough to need gloves, cold enough to need multiple layers, etc.). In that sense, yeah, Kelvin would be impractical, as would be a theoretical temperature scale that compressed most human-relevant temperatures between 0 and 2 or so. But that's not the case and degrees Celsius is perfectly viable for everyday use, as is (I imagine and you say) degrees Fahrenheit.

Maybe your experience maps to the 0-100 Fahrenheit scale, but mine certainly doesn't. Temperatures where I'm from are typically 5-25C, 0-30C at the extremes. So 41-77, or 32-86. So how's Fahrenheit a more useful scale there? And what if you're from, I don't know, Bangkok? The temperature range there is, generously, 80-95F.

There are endless examples of where this argument doesn't stand up because the "averafe==ge human experience" is nothing like yours. But of course, it's more or less applicable in most of the US, so you imagine it's universal. Whereas, for the rest of us, 0 is cold, 10 is cool, 20 is warm, 30 is hot, and extrapolate from there. That's simple and relatable enough for me.

Your subjective experience of temperature has three digits of precision? Because mine sure doesn't.

> But they usually directly map to average human experience. > they don't understand anything about non-metric units or their practicality.

They don't see it the same way that you don't see that the units are totally irrelevant, the brain just have to form habits around the relation between the numbers and the world, but as soon as there are consistant and predictable, the extra effort or lack of doesn't change much about their practicality. The habits are formed early, and doesn't require any extra work once set.

Having physical relationships between units without having to remembering ratios can be handy time to times, and for your entire life.

Pretty much the only reason I like the Fahrenheit scale, miles, inches and ounces seems to be that it’s something no one else uses. On the other hand, computers make calculations with these non-metric units just as easy, and precision-wise computers have no advantage to be taken from the base 10 anyway. (The fact that US is stuck with base 10 for numbers doesn’t mean much, as there always are ‘dozens’ and ‘scores.’)

U.S. Letter size could behave like A4 (root 2) if it were 8.5" x 12" ... not a huge stretch.

(Or 7.75" x 11", approximately, if you prefer.)

But then it wouldn't be US Letter.

There are other paper sizes based around the A series that fit relate back to it as part of a geometric progression, which means stuff fits neatly together. For instance, the C and DL series, which are used for envelopes, with a C4 envelope being exactly the right size for a sheet of A4 paper.

To be sure, but the point I was questioning was that A4 is too skinny for most use cases, that U.S. Letter was preferred in that regard. Adding 9% to the length of U.S. Letter doesn't seem to make it suddenly overly skinny.

Bad for origami though. Ditto with regard to A4. :-)

Preferred by whom? I've never encountered a use case where A4 was too skinny. And it certainly never stopped anyone from making paper airplanes from whole sheets at school. A4 folds neatly to sheets of the same ratio, which is its party trick and why the DL and C series of envelopes are a thing.

Can you give an example of a non-metric unit that is more practical than it's metric counterpart outside of familiarity?