Verified VCDS User
- Feb 13, 2014
- Reaction score
- NC USA
- VCDS Serial number
Which kinda validates my original question: Why the hell not?
The lamp on my bedside table comes on at exactly sunrise - calibrated to my stratum 1 redundant time servers.
I like it! I'll add those to my itinerary if I ever get over there.Glad to see that I'm not the only person geeky enough to get off the Concorde in London and make a beeline to the Royal Observatory in Greenwich to see the Shepherd's Gate Clock in person and synchronize my watch. I also took pictures of the former BBC atomic clock at the Science Museum.
A real nerd would discipline a pendulum clock with NTP. The NTPD guys already know how to do clock slew (tiny adjustments to the clock tick rate) instead of outright time corrections. You could use that along with a small servomotor somewhere to tweak spring tension or whatever it is you do to calibrate the rate of time passage on pendulum clocks. The real challenge would be tracking the pendulum clock's mechanical state, i.e., what time it's displaying. The gear teeth might be big enough to serve as encoder wheels for an optical pickup. Or, put a small camera in the works and do some image processing.Ironically, I was looking at what it would take to construct an NTP digital clock for my office just last week. My assistant restores and collects old pendulum clocks, and the one he has over his desk just doesn't belong in a high-tech office space. Hence, wanting to make my own clock to trump his.....
A real nerd would discipline a pendulum clock with NTP.
Nah... think of it like keeping the carburetor but adding servomotors at the idle and main jet screws (and I guess an O2 sensor somewhere in the exhaust path). You'd keep the old-school mechanism, but use hidden modern feedback controls to keep it in perfect adjustment.Sounds too much like replacing a carburetor with throttle body injection!
The Children's Museum in Indianapolis has a three-story water clock, the largest in the nation. It's mesmerizing... when we took the family there, I spent probably 30 minutes watching it work and trying to fully understand it.I wanted to build an atomically-synchronized Roman water clock in a three story atrium at work, but nobody else liked the idea.
Yeah, that's sorta how VW added Lambda Control to CIS-injected cars in the early 1980s. An oxygen sensor, a little control box that generated a PWM output, which controlled a Frequency Valve that bled pressure from the mechanica fuel metering unit.Nah... think of it like keeping the carburetor but adding servomotors at the idle and main jet screws (and I guess an O2 sensor somewhere in the exhaust path). You'd keep the old-school mechanism, but use hidden modern feedback controls to keep it in perfect adjustment.
I loved CIS injection. A lot of people found it incomprehensible,
but all you needed to trouble shoot it was some pressure gauges and a multimeter.
The only problem is that I fear the sound of running water all night would subconsciously lead to multiple trips to the bathroom.....
Dammit - I keep coming back to this picture and wanting that. [19" rack Jason... see where I'm going here?]
That does indeed look like a 19" rack, although I think our datacenter staff would probably frown on having several gallons of glass-contained liquid above the raised floor.Dammit - I keep coming back to this picture and wanting that. [19" rack Jason... see where I'm going here?]
I don't think your typical water/marble clocks try to give you an accurately readable seconds display (at least, not using water/marbles). That would be a pretty busy, noisy clock. The way the rolling marble clock design handles 60 minutes is with a five-marble one-minute accumulator and a twelve-marble five-minute accumulator, both of which you need to add to get minutes.The one gotcha that I foresee is flushing the seconds marbles after 59. I do not think that a solenoid gate at the end of that tube could dump 59 marbles in the ~ two second interval until the next minute's 1 marble is entering the top of the tube.
I don't think your typical water/marble clocks try to give you an accurately readable seconds display (at least, not using water/marbles). That would be a pretty busy, noisy clock.
Happened to run across some tzdata updates and tweaks in some release notes for CUCM today. These changes are all since 2013. It might illustrate the complexity involved. Yeah, auto manufacturers should be able to handle this sort of thing in a high-line nav unit, but it would be difficult in anything less, particularly something that can't receive updates. It's not really that much computation, but the rule-set is complex and malleable.Which kinda validates my original question: Why the hell not?
- Africa/Cairo (Egypt) observes DST starting 2014-05-15 at 24:00. -- The 2014 Ramadan-based transitions are 2014-06-26 and 2014-07-31 at 24:00 -- Africa/Casablanca and Africa / El_Aaiun's Ramadan transitions are 2014-06-28 at 03:00 and 2014-08-02 at 02:00. - Antarctica/Troll is a new entry - Asia/Istanbul (Turkey) began DST on 2014-03-31, not 03-30 - Europe/Istanbul (Turkey) began DST on 2014-03-31, not 03-30 - Europe/Simferopol (Crimea) switched to Moscow time on 2014-03-30 at 02:00 local time and will not observe DST - Pacific/Fiji ended DST on 2014-01-19 at 02:00, not the previously-scheduled 03:00