[tzs]

> I also prefer Fahrenheit now that I'm used to it. Fahrenheit has a nicer range for the temperatures we live our lives at, whereas Celsius compresses that range so much.

What I don't understand is why Celsius even exists.

For length, area, volume, and mass replacing the old units makes a lot of sense to move to a power of 10 relation between the base units for those things and the smaller and larger units derived from those, and to name those derived units by prefixing the base unit, and to use the same prefixes across different kinds of units.

So with metric we just need to learn one basic unit (meter, meter^2, meter^3, gram), one set of prefixes (milli, micro, centi, kilo, etc) and then we are all set. Compare to English where the ratios between consecutive names units for a given kind of unit vary, and knowing them for one kind doesn't help with other kinds (e.g., knowing inch x 12 = foot, foot x 3 = yard, yard x 1760 = mile does not tell you anything about how many ounces are in a gallon).

Although one might argue that the multipliers in English units tending to be of the form 2^n 3^m makes a lot of problems easier than the 10^n multipliers of metric, there are also cases where the 10^n multipliers makes things easier, and metric using the same prefixes across different types of units is definitely a win for usability.

For temperature though we generally don't use smaller or larger units, so we don't have a question of what kinds of multiplier to use for bigger or smaller temperature units.

All it appears that Celsius does is make the degree bigger, and change the origin. (When first proposed it also changed the direction...0 C was the boiling point of water and 100 C was the freezing, but not long after Celsius revised his system so increasing numbers corresponded to hotter like most other systems).

Some say the argument for Celsius over the older and already widely used Fahrenheit is that basing the scale on the freezing and boiling points of water was easier to reliably reasonably reproduce than the reference points of Fahrenheit.

But that argument really doesn't fly. If the problem was the reproducibility and reliability of the reference points the obvious solution would be to change the definition of Fahrenheit so that 32 F is the freezing point of water and 212 F is the boiling point of water. That would fix the definitional problem just as well as Celsius did without requiring new thermometers.

[grujicd]

While temperature of boiling water has mostly academic meaning for everyday life, freezing point is much more important. I would say the most important temperature point for us humans. That point affects driving, walking, taking care of pipes, outside plimbing, plants, etc. Everything else doesn’t have to be that precise, 20C or 25C doesn’t make much difference. But whether it’s +2C or -2C could have huge impact.

Because of that it makes sense that it’s zero instead of some magical number that you have to compare in your head. I’ll take zero as a win for Celsius.

[fingerlocks]

I mean, not really? We usually measure air temperature but ground and water are much slower to change and are usually not the equal to the air temperature. I think Fahrenheit still wins here: There might be ice on the road when the air is as warm as 40f during a passing winter warm front. There is definitely ice on the road at 0f. That’s a useful 40 degree range of increasing levels of caution.