Well quite, I said that it illustrated the mode of argument, not that the arguments were valid. The arguments made on behalf of 'astronomers' are of course made by assertion without bothering to ask what astronomers might think. Every time someone proposes removing some archaic piece of junk from the Internet specs we have people saying 'but its being used in rural Africa'. I saw that argument being made seriously with respect to UUNET bang path mail routing when it was finally being euthanized.. And the adage 'if it ain't broke don't fix it', is invariably used after someone has asserted that something is broken. So a more honest version of the usage would be 'If I don't think your problem matters, don't fix it'. The point about the leap seconds is that the same phrase can be used to support both the status quo (constantly changing the measurement of time in unpredictable ways) and the proposed change (stop adding leap seconds). And people whose interests are in preserving the status quo for the sake of the status quo can always dismiss the claims of the people who were not at the table when the original decision was made. Unless the earth is slowing down faster than I thought, it would be several thousand years before the accumulated error is as much as an hour. But since we never use UTC time for daily use, why would this matter? We always use local time. Making UTC time slip constantly means that we need three separate time systems: TAI, UTC and Local when all we really need is one fixed series and a second one to make up the adjustment. On Thu, Mar 18, 2010 at 12:38 PM, Marshall Eubanks <tme@xxxxxxxxxxxxxx> wrote: > > On Mar 18, 2010, at 12:00 AM, Phillip Hallam-Baker wrote: > >> Well the US pint is 16 fluid oz which is 1 lb of water. The UK pint is >> 20 so a pint of water is a pound and a quarter. Go figure. >> >> But since we are on the subject of time, why accept UTC as the basis >> for Internet time? Leap seconds are unpredictable and lead to system >> errors. The only group with a colorable benefit from leap seconds are >> astronomers, the one group that might be expected to be able to fix >> leap seconds retrospectively. >> >> The ITU has been discussing plans to abandon leap seconds in >> perpetuity, but the astronomers always seem to win in the end. If we >> moved from UTC to Internet Time, we could abolish leap seconds. >> > > This is backwards. Most astronomers I know regard UTC as a nuisance. In > their calculations, astronomers use TAI (or, if they need to know the > rotation of the Earth, UT1). Solar system ephemeris work uses ephemeris > time, for historical reasons (ET - TAI = 32.184 seconds). GPS internally > uses GPS time, which has the leap second adjustment appropriate for the > start of the series in 1980, supposedly because the GPS program office > didn't understand leap seconds (TAI - GPS = 19 seconds). > > The push to create and maintain UTC came primarily from mariners & various > navies, who wanted to be able to do celestial navigation using civil time > (i.e., to treat UTC as an approximation of UT1, so that you could do km > level celestial navigation using time straight from NTP or WWV). Now, with > GPS/Glonass/Galileo, this seems largely moot. > > Now, it is true that Ken Seidelmann is an astronomer, and he is against the > change, but that is mostly in a "if is isn't broke, don't fix it" mode, and > also because he is thinking of the long term (in 500 to 600 years the > UT1-TAI offset should be order an hour, and people can be expected to start > complaining). > > The biggest thing stopping any change is apathy (and the aforementioned "if > is isn't broke, don't fix it"). > > This site has a lot of information on this subject > http://www.ucolick.org/~sla/leapsecs/nc1985wp7a.html > > I used to say that computer time should be TAI (closest to the actual > clocks, easy to calculate elapsed times), but that never seem to get any > traction. > > Regards > Marshall > >> >> OK, I am not seriously proposing the IETF try to do this (well not >> unless we get into a real fight with the ITU). But if you read some of >> the idiotic arguments advanced in favor of introducing random, >> unpredictable changes into the measurement of time, they are rather >> interesting. There are astronomers who seem to think the earth >> revolves around them. There are dire predictions that stopping >> fiddling with the time system would be a 'major change'. Every >> argument is thrown out, regardless of whether it makes any sense. >> People who point out that leap seconds really do cost real money are >> poo-pooed as having insignificant importance in such lofty debates. >> Quite a few of the protagonists attempt to claim it is only the >> ignorance and stupidity of the objectors to leap seconds that makes >> them unable to see the reason that they are essential. >> >> Over the course of a year, the length of a day varies by several hours >> at this latitude. And the time at which noon occurs varies by several >> minutes. And twice a year the state decides that we will all get up an >> hour earlier or later. So what benefit are those leap seconds to me? >> Absolutely none that I can see. >> >> >> >> On Wed, Mar 17, 2010 at 12:18 PM, Iljitsch van Beijnum >> <iljitsch@xxxxxxxxx> wrote: >>> >>> On 17 mrt 2010, at 17:02, Michael Edward McNeil wrote: >>> >>>>> (Although the exposure to non-standard ways of doing things may make >>>>> this harder for Americans.) >>> >>>> Since Americans habitually use month-day order anyway, why would >>>> YYYY-MM-DD be especially difficult for them? It's Europeans and others who >>>> typically use day-month order that would seem likely to incur difficulties >>>> -- except that putting the year first is a pretty glaring clue that the >>>> order shouldn't be regarded as it usually is for them. >>> >>> Absolutely. But Americans don't expect this kind of stuff to make sense, >>> because they're used to having a different way of measuring everything, >>> while in the rest of the world we're used to the metric system so we assume >>> things make sense. So an American wouldn't necessarily consider yyyy-dd-mm >>> inconceivable while people from elsewhere probably would and just assume >>> yyyy-mm-dd. >>> _______________________________________________ >>> Ietf mailing list >>> Ietf@xxxxxxxx >>> https://www.ietf.org/mailman/listinfo/ietf >>> >> >> >> >> -- >> -- >> New Website: http://hallambaker.com/ >> View Quantum of Stupid podcasts, Tuesday and Thursday each week, >> http://quantumofstupid.com/ >> _______________________________________________ >> Ietf mailing list >> Ietf@xxxxxxxx >> https://www.ietf.org/mailman/listinfo/ietf >> > > -- -- New Website: http://hallambaker.com/ View Quantum of Stupid podcasts, Tuesday and Thursday each week, http://quantumofstupid.com/ _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf