Ha ha. You hit on my exact problem with -er vs -re endings in British English. I could never keep track of which to expect. American English is generally -er, with -re used only in cases like acre, where an -er ending would make a soft c. American English still has abundant opportunity for simplified, more phonetic spelling, but in my view, it has made some progress over British English. I prefer to go forward, not backward. However, I respect the right of the British to spell as they wish, and most times, I can understand the word. It is obviously a YMMV situation.
--- On Mon, 1/31/11, [email protected] <[email protected]> wrote: From: [email protected] <[email protected]> Subject: [USMA:49713] Re: [SI] A mole hill into Himalaya To: "U.S. Metric Association" <[email protected]> Date: Monday, January 31, 2011, 11:54 AM John, In addition to surveying errors of refraction (curved path) errors, there are "group speed" (packet, or group pulse speed) errors. These errors illustrate the difference between high "precision" (small random errors) and high "accuracy" (small absolute errors). A laser theodolite can be "precise" (small random errors of the "same" observation, but not very accurate (large systematic errors of the falsely assumed straight-line path ) much larger than 2 ppm. To continue your play on a mole hill, my fathre taught physics, chemistry, and biology (at the high school level), my mothre taught English (American standard), and my brothre taught mathematics (at the college level). I nevre had a sistre, but I do have a daughtre, and a granddaughtre. Gene. ---- Original message ---- >Date: Sun, 30 Jan 2011 14:00:50 -0800 (PST) >From: "John M. Steele" <[email protected]> >Subject: [USMA:49700] Re: [SI] A mole hill into Himalaya >To: "U.S. Metric Association" <[email protected]> > > I did further checking, including looking at Leica > specs. Those numbers like 1.5 mm + 2 mm/km are not > worst case specs, they are standard deviations. > Statistically, we know there is reasonable chance of > individual measurements having errors out to 2 or 3 > standard deviations, rarely even more. > ...
