Deflating the tires 🙂? That works on sand, but the usual recommendation for snow is to have a narrower contact patch if possible, I thought. Most high performance cars have narrower winter sets. I guess it depends how deep the snow is. As long as you made it safely.
Seems like it's snowing everywhere this week.
Seems like it's snowing everywhere this week.
I don't wear winter tires. Here it's snowing once every five or ten years.
I do have snow chains for thick snow over long distances but it's a pain to put them on for snowed local streets and putting them off at cleaned main roads and back. For this in-town, low driving speed scenario, deflating tires works (increases traction). Drivability worsens (heavier steering)
George
I do have snow chains for thick snow over long distances but it's a pain to put them on for snowed local streets and putting them off at cleaned main roads and back. For this in-town, low driving speed scenario, deflating tires works (increases traction). Drivability worsens (heavier steering)
George
Attachments
Last edited:
I always swap to winter tyres, and back... Much more grip on cold or icy roads.
I keep snow socks (cheap quick get you of trouble, work well) and chains for emergency in the boot in winter. Plus all the cars are 4WD.
I keep snow socks (cheap quick get you of trouble, work well) and chains for emergency in the boot in winter. Plus all the cars are 4WD.
Hmm, 16 bits gives a full scale error of 1/65536. Not sure if that's "too much" or not.
To eliminate all possible future confusion between binary and decimal, I propose that when decimal digits is meant we use dits.
'16 bits' I don't think is the right nomneclature
16 decimal places more like it - I've just worked it out - gives 2" at 1.5B miles with at least a spare decimal place for uncertainty, and most of another for rounding
16 decimal places more like it - I've just worked it out - gives 2" at 1.5B miles with at least a spare decimal place for uncertainty, and most of another for rounding
There is confusion here arc second is not a measure of distance but angle, it is 1/3600 of a degree. 2" at 1.5e9 miles is 14,544mi. (tan(1/1800)*1.5e9). You need to specify how far off you need to be at a given distance (atan()) to specify the angle. As for angular resolution one part in 1e-16 of 2pi radians (360 degrees) gives 0.06" in 1.5e9 mi.
I'm still confused what maths we are discussing here. on the ground, in the craft or telemetry.
Right, if you are flying by joy stick using servos more than 16 bits would probably be silly.
Last edited:
Fun seeing the numbers.
Encoders at single digit nanometers and nanoradians is common nowadays.
Otoh, trying to do a course correction with a burn impulse accurate to 16 bits is not quite there.
Measurement of star position is gtg at that level.
The real problems tend to be other things, like consistency in units of measure.
Jn
Ps..Scott, I thought he meant two inches.
Pps. Imagine trying to use a mirror to reflect an x-ray beam, and having the spot 100 meters away remain stable at the micron level..boggling. Who knew this could be done.
Encoders at single digit nanometers and nanoradians is common nowadays.
Otoh, trying to do a course correction with a burn impulse accurate to 16 bits is not quite there.
Measurement of star position is gtg at that level.
The real problems tend to be other things, like consistency in units of measure.
Jn
Ps..Scott, I thought he meant two inches.
Pps. Imagine trying to use a mirror to reflect an x-ray beam, and having the spot 100 meters away remain stable at the micron level..boggling. Who knew this could be done.
Last edited:
I have been using an air gap for 20 years. No USB, actually upgraded to a cd.
When a machine and software is built with sufficiently advanced capability that it's still beyond SOTA after 20 years, it ain't broken. So money will not be spent to take it offline for 6 months to upgrade software.
Jn
When I was in space* the biggest problem was the ground controllers thinking they knew better and ignoring the folder full of procedures. With spacecraft if in doubt go into panic mode and point at the nearest thing you can lock onto and await instructions. Ground crew then have to decode what went wrong from the few hundred bytes of telemetry and coax it back to doing what it should.
For probes of course you only have set hours on the deep space network to talk to it.
I have huge admiration for what JPL manage to do given the limitations.
*In case Ed is lurking I wasn't actually 'in space' but working on space projects.
For probes of course you only have set hours on the deep space network to talk to it.
I have huge admiration for what JPL manage to do given the limitations.
*In case Ed is lurking I wasn't actually 'in space' but working on space projects.
Bits were changed to digits, my error on the double quote ambiguity. The thought that something would be launched from earth and be expected to arrive within 2 inches at 1.5e9 miles without any intervention or course correction ever never occurred to me. There must be some missing context here.
The context was that the use of pi to 16 digits was not a significant problem to getting the satellite to final destination, according to the web page I visited that I have not bookmarked, the Nasa engineer explained it was "less than 2 inches of error at the final destination," so they never need to store pi to better accuracy than that.
I assume taking calculations out to 24 or 44 or 64 decimal places is useless given the inputs are nowhere near that level of accuracy. 16 decimal places seems good enough given the inputs.
Multiple measurements over time can increase accuracy, but there are limits. Averaging helps..
jn
wrinkle....2 inches being good enough....as I recall, one mars orbiter....hit the planet..
good shot...too bad they wanted to miss...
I must admit. I do not understand the term "satellite" used with 1.5 billion miles. Satellites don't go that far, even geosync is 22.2K miles..
Multiple measurements over time can increase accuracy, but there are limits. Averaging helps..
jn
wrinkle....2 inches being good enough....as I recall, one mars orbiter....hit the planet..
good shot...too bad they wanted to miss...
I must admit. I do not understand the term "satellite" used with 1.5 billion miles. Satellites don't go that far, even geosync is 22.2K miles..
Last edited:
I'm sure they weren't doing floating point arithmetic, but you might (or not) be surprised about the number of "full stack" software engineers these days that don't understand how floating point numbers are represented in a computer. Makes me feel like somewhere there is a system where someone could pull off an Office Space type scam.
I’m not surprised by the number of software engineers these days that don’t have a clue what is a pointer, polymorphism, and the difference between a class and an object. Yes, I am old as dirt and clearly over qualified (to obsolescence) as a programmer, that’s why I am into management.
Sadly I am also on my way to obsolescence, not being a big fan of throwing web frameworks everywhere or making desktop apps that are basically sandboxed web pages.
I see the pointer issues often, yeah. Surprised about the object oriented stuff. Even though it's not trendy right now it still seems like a staple in most CS programs.
I see the pointer issues often, yeah. Surprised about the object oriented stuff. Even though it's not trendy right now it still seems like a staple in most CS programs.
Not to beat a dead horse here but why would you even need the value of pi? Pi has nothing to do with measuring the straight line distance between to objects.