On Nov 19, 2008, at 4:08 PM, Peter Clifton wrote: > On Wed, 2008-11-19 at 15:44 -0700, John Doty wrote: > >> If you insist on no artifacts, it's zero. >> >> If you round imperial units to the nearest 0.01 mil, you have no >> additional roundoff error if your fundamental unit is 1 nm, because >> 0.01 mil is *exactly* (by definition) 254 nm. > > Good point... meaning it would be sensible to move towards some > kind of > metric internal units, or an intrinsic coordinate system with units in > metric being given the divisor of 254 * 10^n. > >> It's much harder to go the other way with a humanly comprehensible >> rounding because of the factor of 127 in the definition of the inch. > > I get that 127 * 2 = 254.. which definition of the inch mentions 127 > explicitly though?
I have multiple reference books that state 2.54 cm = 1 inch exactly. Presumably there's some SI or NIST definition. > >> 2^32 nm is about 4.3 meters, large enough for any PCB I've ever seen. >> 2^64 nm is about 0.12 astronomical unit ;-) > > Well.. I can't recall if we use signed or unsigned numbers throughout, > so we might end up limited at half those. If it's signed, you can go 2.1 m either way, so it's the same. > Aren't 64 bit integers going > to be slow on a 32bit CPU? How slow. What fraction of computation is arithmetic in world coordinates? > > > Practicalities of changing: > > 1. Change internal units: ought not to be _that_ hard since DJ nicely > abstracted things away when he refactored to introduce the HID (IIRC). > Make loading / saving convert to the old units, (rounded to integers). > > 2. Add unit specifiers to file-format, so users could, if they want, > write out coordinates such as: > > 5mm, 10mm or 5 mm, 10 mm. from their footprint-generation scripts. > > Save the files out with native units, e.g. "123456000 nm" > > Perhaps, for optimal flare, we could make it more human readable, by > writing out something decimalised, like: 123.456 mm (*) > > If the coordinates for a point happened to be a convenient decimal > fraction of an inch (requiring less decimal places / significant > figures > than the metric representation), then we could choose to emit in mils, > or inches etc.. > > * (Taking care not to loose precision if we were to read it back via a > floating point number). > > > > -- > Peter Clifton > > Electrical Engineering Division, > Engineering Department, > University of Cambridge, > 9, JJ Thomson Avenue, > Cambridge > CB3 0FA > > Tel: +44 (0)7729 980173 - (No signal in the lab!) > > > > _______________________________________________ > geda-user mailing list > [email protected] > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > John Doty Noqsi Aerospace, Ltd. http://www.noqsi.com/ [EMAIL PROTECTED] _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
