--- Bill Hooper <[EMAIL PROTECTED]> wrote: >... > But it does NOT make sense to have either the length > unit or the time > unit linked to something so inconstant and imprecise > as the earth's > size or its rotation. > ? But Bill you don't seem to have quite caught the idea here, my friend. While it IS true all you said below, the fact of the matter is, there IS usefulness IN THE CONCEPTUAL definition, which, BTW, is STILL valid until today!!! Sure, there are the imprecisions and stuff and a much more rigorous way of defining it *today*, but there is NO denying that the *principle* behind it is what allowed for this... "size" to get a... "human face"!
I'll explain more on what I mean later below. > The early definition of the metre in terms of the > size of the earth is > not sufficiently precise for our modern technology. (True, but...) > That's why it was > abandoned long ago. (the *idea* or concept, though, wasn't!) > Why should a universal > measurement depend on the > size of one small planet? The size of this planet is > not perfectly > constant, anyway. (True, but just like 365 days a year is NOT accurate either, and a day of 24 hours, etc, there's still value in defining and quoting measurements with these things. Don't we still do it EVEN TODAY? Example, how old are you?... ;-) It's not the fact that this is not 'perfectly constant' that matters, but that relating a *rigorous* size that can be refered back to something THIS tangible that does! NO, I'm NOT using the same argument ifpists use, but rather proposing a side effect of an "old" definition that STILL has use today!) > It varies with erosion, infalling > cosmic dust, tidal > effects, the particular latitude along which it is > measured, whether it > is measured at sea level or some other elevation > (and sea level is not > constant and all other elevations are measured from > sea level). > Yet it's still useful that one uses something like the monster nautical mile, for instance! Just because of it's *convenience* relationship to the arc-angle - remember, aviation and maritime navigation STILL do it! Obviously, the size of these things are INVARIABLE *because* they're tied to the meter, a near immovable immutable size! But my point, Bill, is that there is merit in adopting "sizes" that *make sense* on a "larger picture". If these... "sizes" are flawed, fine, let's make them "immovable"! Case closed! Like it or not, here is the thing. IF there is *rigor* in such sizes, using them in a framework that has a *rigorous* unitary/decimal relationship would serve a formidable purpose! Let me illustrate. Aviation STILL clings to knot/n.m. 360 degree angle... Why? It doesn't matter that surfaces are unequal, erosion, blah, blah, blah, what matters is that there are *useful/valuable* relationships tiing these entities together!!! Alas, if this is so, imagine how much more useful it would be if instead of these trashy things one would deal with 40 Mm *exactly* to a "hypothetical/ideal" sphere, which would contain 400 gr causing 0.01 of a grade/gon to be *exactly* 1 km! If pilots/aviators/navigators can swear by kt/n.m./deg and the likes, they'd *certainly* do the same if they had a chance to use m/ m/i / gr (and that's it!). > The length of the day (and indirectly the hour and > minute) is also not > precise or constant enough to serve as a universal > standard for time > measurement. Of course they aren't, you're right, but... If a "hypothetical" day can be *rigorously* defined as: (the equivalent of 86400 s) why can't I do likewise, but say it's 100 ki??? Please remember, both the s and i are *rigorously* defined! In other words, please compare: Today: 1 day (imprecise) s (precise) Tomorrow: 1 day (imprecise) i (precise) Same *EXACT* thing! >... > The purpose of standard units in SI (or any other > quality system) is to > provide CONSTANT and REPRODUCIBLE values for the > fundamental size of > each unit. AGREED! And (ultimately) there would be *nothing wrong* with doing it vis-a-vis a new i (ip)! In the meantime though I'd be pretty content with just **using** it for my time reckoning till the day, who knows, when: Future: 1 day (*decimally* defined, but imprecise, still...) i(precise)! > Relationship of such sizes to physical > features of the earth > (or the human body) ARE IRRELEVANT! True, BUT if we could capitalize on the usefulness of such relationships when the fundamental unit IS 'constant' (as you put it), why not? > Constancy and > reproducibility are > the only things that are important. > Agreed. Therefore, *on technical grounds* you SHOULDN'T oppose the *principle* of what I shared here! (and NO, again, I'm not advocating a change to the second!) > One can define them in such a way that the sizes are > convenient in > general, but to try to make them VERY convenient for > certain > applications, automatically makes them less so for > other applications. ? However, given that we ALREADY deal with a day as a unit, what's the difference between doing it via s or i? Please note I'm not advocating a *new* unit, it's still the *exact same* day length, but written on different terms, as in: 1 day = 86 400 s (rigorous) 1 day = 100 000 i (rigorous) Both entities above have the *exact/same* size!!! > Therefore, the usually followed procedure is not to > try to make them > convenient for ANY particular application. (Although > this was done in > the past, the practice was flawed from the beginning > and was quickly > abandoned.) > ? Interesting take. I'm not sure you catch the points here though. You claim (and agree) that 'convenience' WAS done AND achieved, but also indicated that this was 'flawed from the beginning'. Hmm... The new "model" would NOT change relatively to the present in terms of defining a *fundamental* unit for time. The difference would be in the length of that entity. BUT, such a small change *ultimately* could lead to so much *convenience* for *so many* applications that it may in the end be worth the while! BUT even if BIPM finds it not, fine, STILL using a new framework would STILL be advantageous EVEN in spite of an... "eternal" conversion factor (0.864)! (well... don't we already have one, err... actually, 3?! 60, 60 and 24) > The SI solution to obtaining units of convenient > size for many > different applications was to create new units that > are simple > multiples and sub-multiples of the basic unit and > then identifying them > by the SI prefixes. Excellent, agreed! And a "permitted to use with SI" NEW ip would serve the same "purpose" that hideous things like knots, nautical miles and the likes have today, with the exception that, like the hectare, they would be FAR more useful (if only to rid us from such sillinesses!) > This allows ALL measurement to > be expressed in > whole numbers that are reasonably small (between > units and thousands) > or decimal fractions not smaller than tenths or > hundredths. > Again, good point! So... we all agree. > No particular measurement turns out to be some exact > simple whole > number and there is no reason it should. But with the ip they still wouldn't as amply demonstrated above, so I'm not advocating any change or something new here! > The metre > is not exactly the > length of anybody's arm (and need not be) and the > second is not an > exact fraction of the earth's rotational period (and > need not be). Etc. >... All true, of course. But how much more **useful** it would turn to be if IN ADDITION to its harmony with the rest of the SI units the time framework had *finally* become rational after all! (food for thought) Respectfully, Marcus ______________________________________________________________________ Post your free ad now! http://personals.yahoo.ca
