> I was also dubious of the claim that unit conversion isn't important -- that manipulating multiple values of like units is far more common. Well, maybe, but it doesn't mean unit conversion isn't often necessary. I convert between units all the time. Plus, if you add up enough measurements expressed in "inches," sooner or later you're going to need to convert to feet or yards or miles just to visualize the value.

For a person who claimed she/he was a "35 year-old engineer" I was pretty annoyed with her statement "The fact that metric units are base ten in fact has virtually no relevance either to day to day life or to scientific and engineering manipulation". Say again? That's complete rot. In my current job I adhere to the ASTM (American Standard Testing Materials) standard to evaluate the performance of metals. The first thing I do is measure in inches and milles (thousandths), since my specialty tools are in Imperial -- but then I immediately convert everything to metric. My calculations involve density, volume and mass and trying to resolve in them in Imperial would just add 4 steps to the calculation. In fact, the Reference Standards given by ASTM are in metric because it's sooooo much easier to compare values under that system.

Then there's my last job, where I worked using liquid concentration levels of 20 Picomolar. It's a trivial problem converting Liters to Picoliters -- just move the decimal to the left. Ounces, however...they can't even be divided up that finely.

I don't know what kind of units you work in Sam, but Imperial is fine for the Macroworld scale -- and completely inadequate for extrapolation into microscale analysis.


> What I *do* think is true, however, is that the base imperial units seem to be more suited to measuring the things humans tend to observe. It is true that degrees Fahrenheit have a finer resolution than degrees Celsius -- and as I often find the difference of a single degree Fahrenheit to be noteworthy, I wouldn't think using Celsuis, nearly half the resolution, would be desirable.

I never thought about any drawbacks to the resolution difference between Celsius and Fahrenheit. I mean, 37°C is 98.6°F, so already both scales are subdivided into tenths. In real-world terms the Celsius scale makes life easier for living in frozen subarctic climates (like Canada, eh). For ex. in the Winter: I might look at the thermometer outside the window and it reads 10°C. Then I look at the puddles in the road and see they're frozen. Normal water only freezes at 0°C, thus there's a wind-chill factor making it much colder than 10°C and you have to dress your kids up accordingly. Yes it's true you could do all that knowing 32°F is the freezing point. Honestly though, I wouldn't have intuitively observed the useful connection between wind-chill and freezing temperatures using the Fahrenheit scale...


> With regard to measuring length, I think the metric system is missing a base unit between the centimeter and the meter (even those that use the metric system aren't likely to use the decimeter or find it an intuitive base unit). Lots of things in the human world are best expressed in terms of feet that can't be as intuitively expressed in the metric system. "It's about a foot long" says plenty to me, but were I to give that measurement in centimeters or a fractional part of a meter, I don't honestly think that would be as effective.

Um...Well... Here, I have to agree with you. As f2m says, centimeters aren't comfortably "intuitive" units because they're overly small, and millimeters are way too small for everyday analysis. If only we could have something like a "daucen" = 2 cm, or something. Or, once all our elders have died off just redesignate 1 foot = 30 cm because they're close enough already :-). Frankly the only real-world advantage to metric length -- if you could call it an advantage, barely -- is that it takes only 5 seconds in your head to compute how many millimeters are in a klick, whereas you'd have much more trouble figuring out how many 1/32nds of an inch are in a mile. About the only people who'd need to work with that kind of data are folks designing laser and GPS satellite trackers anyway.

Just my $0.01's worth to the debate.