Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The super-slow conversion of the U.S. to metric (2025)

thefabricator.com

Post reply on HN
Print view View original

361–370 of 422 posts

Re: The super-slow conversion of the U.S. to metric (2025)

#361

Post by pandaman »

nayuki wrote: post #354
pandaman wrote: post #326

Earlier quoted context omitted.

>Lots of useless names and numbers to memorize Do you have problems with time too? I mean, 1 minute = 60 seconds, 1 hour = 60 minutes, but one day = 24 hours? Wtf??! And one week = 7 days! And one month is 30 unless you mean an actual month which is anywhere from 28 to 31. And the year is 365 days, unless it's a leap year with 366. How do you cope with that?

> Do you have problems with time too? How do you cope with that? I have memorized how time and dates work, but I do not enjoy the system. Time is my biggest sore point. For starters, doing any kind of arithmetic is an exercise in pain. For example when I rent a shared bike, the system tells me the start and end time to the second - for example, from 13:26:08 to 15:54:39. To calculate the duration, I have to combine t…

I don't know what do you do, but most people need to know how many yards (or inches, or feet, or chains, or whatever) in a mile as often as they need to know how many seconds, minutes, or hours in a quarter i.e. never. Yet it's the strongest point proponents of the metric system have, so makes me wonder why are they so agitated?

For the record, I am from a metric country and immigrated to the US as an adult. I still find American system to be more adjusted to human needs. E.g. temperature in F does not need decimals unlike temperature in C, tool sizing in inches is simpler (look at the sets of drill bits in different systems for example), tire pressure in psi (e.g. one of my bikes is 53 psi rear and 51 front, or 3.65 and 3.51 bars, you could say I could remember just the decimals, but another bike is 33 and 31, or 2.27 and 2.13 so no, it's 3 digits with bars vs 2 in psi), house dimensions are in even number of feet so much easier to find furniture, which is designed with this in mind, obviously. Miles are great to estimate time of travel by car, take 1 minute per mile of distance on a highway and 2 minutes in the city and you will be pretty close.

But, of course, the reason the American system is never going away is because it would be insanely expensive: you either will have to rewrite all building codes/standars/recipes with stupid conversions e.g 50.8x101.6 instead of 2x4 even though the lumber dimensions are not really 2 and 4 inches or scrape them and write the new ones using the more sensible metric dimensions but then you will need to scrape all the tooling you had and buy new, metric tools. All so you could say how many micrometers in a kilometer and feel smart?

Re: The super-slow conversion of the U.S. to metric (2025)

#362

Post by mark-r »

pandaman wrote: post #326
nayuki wrote: post #173

Earlier quoted context omitted.

> I find the imperial system fits in better with that warmth of humanity Right, so you enjoy warmth such as: 1 foot = 12 inches, 1 stone = 14 pounds, 1 pound = 16 ounces. Lots of useless names and numbers to memorize compared to kilo = 1000, milli = 0.001. > in favor of cultural uniqueness and fun over conformity You're writing in impeccable English. As we know, English is an international language and most definitel…

>Lots of useless names and numbers to memorize Do you have problems with time too? I mean, 1 minute = 60 seconds, 1 hour = 60 minutes, but one day = 24 hours? Wtf??! And one week = 7 days! And one month is 30 unless you mean an actual month which is anywhere from 28 to 31. And the year is 365 days, unless it's a leap year with 366. How do you cope with that?

Back in the 1970's I tried to come up with a metric time system by breaking a day into powers of 10. A centiDay was 14.4 minutes.

I realized it would never catch on, because a 30 minute TV show would have to fit into 28.8 minutes, and the only way to do that was to lose a couple of commercials. Never gonna happen.

Re: The super-slow conversion of the U.S. to metric (2025)

#363

Post by pandaman »

mark-r wrote: post #362
pandaman wrote: post #326

Earlier quoted context omitted.

>Lots of useless names and numbers to memorize Do you have problems with time too? I mean, 1 minute = 60 seconds, 1 hour = 60 minutes, but one day = 24 hours? Wtf??! And one week = 7 days! And one month is 30 unless you mean an actual month which is anywhere from 28 to 31. And the year is 365 days, unless it's a leap year with 366. How do you cope with that?

Back in the 1970's I tried to come up with a metric time system by breaking a day into powers of 10. A centiDay was 14.4 minutes. I realized it would never catch on, because a 30 minute TV show would have to fit into 28.8 minutes, and the only way to do that was to lose a couple of commercials. Never gonna happen.

Back in the 1790s some Frenchmen did the similar thing but, despite not having TV and commercials it still did not catch up even in France.

Re: The super-slow conversion of the U.S. to metric (2025)

#366

Post by jrm4 »

tzs wrote: post #345
jrm4 wrote: post #322

Earlier quoted context omitted.

What precisely is "better" about that? It's more predictable over -- some things that you don't know how they're going to scale? Again, the general thrust of "imperial" is better -- base your units on "utility of the most people using them the most for real life things" Do whatever you want for distances between stars, but no, walking off a room in "feet" can't be beat.

To pace off something in feet most people actually try to pace in yards and multiply by 3 if they want to express the result in feet. They could just as easily pace in meters. What is better with metric is the consistent way to name multiples and divisions of the base units. Metric uses power of 10 prefixes but another power could work fine. Power of 2 for example actually fits well with Imperial volume measurements,…

[flagged]

Re: The super-slow conversion of the U.S. to metric (2025)

#367

Post by jrm4 »

timbit42 wrote: post #331
jrm4 wrote: post #320

Obligatory "what it's based on doesn't matter." Like programming languages and UIs, it's what and how people use them. Imperial tends to be better because it's more "evolutionary."

Base 10 is more evolutionary because we, and many animals, evolved with 10 digits on our hands and feet.

Yes, and that is why it "won" even though there are pretty good arguments for 8, 12, 16 and so on.

Measuring things is "for humans."

Re: The super-slow conversion of the U.S. to metric (2025)

#368

Post by rstuart4133 »

I actively avoid stuff made in the USA, precisely because they don't use metric. For example, you can get a USA bolt with all sorts of threads and dimensions, or you can buy a 100mm M8 that is compatible with every other M8 bolt / nut made in the world. That boils down to bolts made in any country, bar the USA.

Every time the USA manufacturers something that isn't metric, you've made it incompatible with the rest of the world. The USA got away with that when I was young because they were the world manufacturing powerhouse. Now, those powerhouses are based on Asia. They define the units most of the world sees, and they use metric. So if I buy a Chinese mower, all the bolts are metric and I'm guaranteed the local hardware store stocks them.

Time has moved on, the USA is now a follower, not a leader in most things bar digital services. If they want to return to selling those things to the world the speeds have to be in km/hr, weights in kg, sizes in mm or meters, the temperature in Celcius, pressures in Pascal's.

Re: The super-slow conversion of the U.S. to metric (2025)

#369

Post by AngryData »

fragmede wrote: post #316
AngryData wrote: post #308

Earlier quoted context omitted.

0F is useful in areas where they salt roads because thats the temperature many salts stop being effective and previously good road might turn to ice sheets. But it isn't a good reason all by itself.

How hard is it to remember -17° C when that's something you grew up with since childhood though? As a trade off, you get 0° C for water freezing and you don't have to remember 32°F for water freezing instead. Or you remember all 4 of those if you leave near a borders.

I would argue that -17C for brine water is far more difficult to remember than 32 for pure water freezing, 32 is an incredibly common number in fractional measures, but I will concede the point because that seems like more of a happy accident than intent.

The biggest problem I have is I don't see either as inherently better, both are relative scales defined by easy to setup but completely arbitrary measures that lacks any real relevance to modern life. We could of defined or scaled our thermometers based on the boiling and freezing points of mercury and nothing would really change, just shifting some numbers around and then still making another absolute scale that can actually be used for any sort of decent thermal calculations.

Re: The super-slow conversion of the U.S. to metric (2025)

#370

Post by adrian_b »

toolslive wrote: post #214
technothrasher wrote: post #149

Earlier quoted context omitted.

> the useful scale of Fahrenheit is far superior 0-100 versus 0-32. Well, first of all, I'm not sure why you're defining those scales as the "useful" ones. They don't even equate to each other. But why are you arbitrarily using 100 as the end of your Fahrenheit scale? Just so you can declare it 'Metric-y'? If you read his paper, Fahrenheit's scale is actually 0-96.

As an engineer, you should not even use temperature at all. All thermodynamic formulas simplify (a lot) if you use the inverse temperature.

You are indeed right that in most numeric computations using the inverse temperature, a.k.a. the reciprocal temperature, is more convenient.

Nevertheless, there are many important quantities which are proportional to temperature, e.g. pressure, internal energy, voltage generated by a bandgap reference and so on. Because of this, there are many cases, especially in qualitative reasoning, when using temperature is more convenient than using its inverse.

This is similar to waves, where in most numeric computations wave-number and frequency are more convenient, but there are also many cases, e.g. when reasoning about resonance frequencies or stationary waves, when using wave-length and periodic time is more convenient.

Another example is in electrical circuits, where for some problems using impedance and resistance is more convenient, while for others using admittance and conductance is more convenient.

Perhaps one would need a simpler name for reciprocal temperature, to facilitate its use wherever this makes sense. However, when implementing a physical model in a program, where one should always define distinct types for each kind of physical quantity, using a short type name like "RecTemp" would not stand out among the many abbreviations typically used in programs.

Post reply on HN