Hacker News Reader

Comment by Irregardless | Hacker News Reader

If this were the olympics of mental gymnastics, you would take the gold medal in every category.

> So, here's the story: He is expecting a number at around 100 miles because he thinks he's being more efficient now. He sees 50 miles. On a digital display at a hotel at night, he misreads the 5 as a 9, and this fits with his expectation, so he goes to bed thinking he has 90 miles of range. He wakes up and the Tesla has lost 5 miles of range inexplicably -- but it therefore has become 45 miles, which looks totally different. He calls them up complaining that the Tesla lost half of its charge overnight and some sympathetic tech support describes it as a "software glitch."

Pure gold. Thanks for the laugh.

View on HN

He does have a valid point in that both the reporter and Mr. Musk may earnestly believe they are telling the truth. For instance if you look at the cabin temperature graph, it's clear that it was uncomfortably cold for nearly 50 miles; but that it wasn't see your breath cold at any point.

But hey, let's all pick sides based on our allegiance to ( east or west coast, Tesla or The Times ) and have a nice little monkey dance about who's right or wrong.

> it's clear that it was uncomfortably cold for nearly 50 miles

No, that's not a log of the temperature reading. They logged the temperature SETPOINT. An important distinction.

I do this all the time, being from cold climates. I jack up the heat at the start of my trip, and then once I get going, I eventually turn the heater off completely.

>No, that's not a log of the temperature reading. They logged the temperature SETPOINT. An important distinction.

It's a thermostat controlled temperature, though.

There's no reason to turn it down if it's set to a comfortable temperature, because it will automatically shut down when the proper cabin temp is reached.

Eh. My car is thermostat controlled, and I often adjust the temperature because the vents are close enough to me that it can become uncomfortable. The size of a car makes it a little different than setting a thermostat in your house.

The difference between theory and practice is ... Nah, forget it, theory is always wrong.

Anyone who drives in the cold knows this is wrong. You turn the set point to 90 at the start of the trip, because the built-in logic is going to drive the fans harder when the delta between ambient and set point is higher, and the car will warm up faster. Yes, in theory the software should recognize that. But, in practice.... Nah, forget it. Theory is always wrong.

90 is not a "comfortable" setting, though.

I agree completely that's what you do in cold weather. I do that every morning with my car. But after 3+ hours in the car, it's pretty unlikely that you'd lower the temperature at that point.

EDIT: Downvotes? Really? If the car's climate control is controlled entirely by a thermostat, then setting the temperature up to 90 will warm it faster when it's cold, but once it's warmed up the cabin, you set the temperature to something comfortable(say, 70). You don't turn it off, because the climate control system is supposed to be doing that automatically.

> If the car's climate control is controlled entirely by a thermostat, then setting the temperature up to 90 will warm it faster when it's cold, but once it's warmed up the cabin, you set the temperature to something comfortable(say, 70). You don't turn it off, because the climate control system is supposed to be doing that automatically.

Hi, I live in Michigan. You do turn it off. Your theory sounds like it could be good on paper, but in reality, humans behave a little differently. Our winters are brutal. At some points in the winter time, it's a horrifying proposition climbing into the icebox that is a car's cabin. Reflexively, many people just crank the heat up as high as it will go and blast the cabin with warm air as soon as it's available. As much as you can, as quickly as possible. What tends to happen then is that at some point, you realize that you're hot. You've been blasting the heat, you weren't paying attention (listening to news/music, concentrating on the road, on your phone, etc)and when it suddenly occurs you you that you don't need the heat to be blasting anymore, it's because you are uncomfortable. You're still in full winter gear in your car, you jacket is smothering you, and the cabin is now hot with you in your coat. People aren't always thinking "Ah, let me just lower the heat to a more comfortable level," they're thinking "OMG turn this thing OFF. Could the climate control handle this in theory? Sure. If human psychology and behavior weren't continually getting in the way.

No telling what any given person would do when you don't have the data, but I can tell you what people do do in reality. They crank it then turn it off.

Also, I'll never purchase another car without heated seats in Michigan again as long as I live here. What was I (not) thinking?

Being a controls engineer, this behavior always drives me crazy. I love to observe the temperature controller do its job. But it does depend on the car.

A good controller will blast the heat at max until the temperature gets close to the setpoint, then back it down. Turning it up to 90 does exactly nothing extra for you, but does end up overheating you.

However, not all cars have good temperature controllers. My wife owned a Honda Accord a while back, that was simply awesome at this job. We never, ever changed the setpoint in that car because it just worked great.

The Subaru Forester we have now is much, much worse, and we do end up fiddling with the setpoint because the controller's not doing its job well.

Being a controls engineer, these assumptions of thermostatic perfection drive me crazy. The things only measure air temperature, not that of the seats, dashboard, wheel, etc. When these objects are cold a person will feel colder than if they were the same as the air temperature. See also thermal conduction and thermal radiation. Also, that thermostat is located where? At the driver's fingertips? Their rear end? Not likely.

So cranking the air temperature setting beyond the desired air temperature for a while will indeed get the cabin more comfortable quicker.

Hi, I live in Ohio. And what I just read was a PERFECT description of human behavior around temp controls during Midwest winters.

Really? No. The fans spin up as quickly as they can in a cold car. If you turn the fan up and the temperature up all the way when you start the car, you are just blowing cold air on yourself. The fans spin up when they can actually deliver warm air. Keep in mind of course this is in a climate controlled car, like the Tesla S.

Indeed! And here's you've given a perfectly good theoretical answer for why cars do the verifiably incorrect thing. The problem here is your handwavy use of "as they can" to invoke some kind of hidden expertise that doesn't actually exist in the real world. The car has some vague idea of what the current fan outlet temperature is, based mostly likely on one or two thermocouples in the chain, or maybe just on the engine coolant temperature. And some egghead firmware engineer somewhere built a very solid, justifiable model of "rider comfort" and optimized the control response to that, thus giving the theoretically optimal experience for...

Nah, forget it. Theory is always wrong. I spend an hour in my car every day. Don't you[1] dare tell me how best to warm it up. Just don't.

[1] To head off the wounded screaming: no, I'm not talking about you personally here. Just the idea of "firmware knows best".

Edit: grammar

Some cars are really good at this. Some are not. It depends on so many factors, including the egghead quality.

I'm one of those firmware eggheads, btw.

As another example, when I fuel my car from emtpy to full, as soon as I turn the key, the fuel gauge ramps rapidly up to Full. In my wife's car, it finds its way to Full after a minute or so. I don't know, but I'm pretty sure the difference is the filtering algorithm employed. One egghead knew about Kalman filters or similar, and the other didn't.

It's the same with all the other software features in the car. Some had good eggheads on the time, and some didn't. I think it's impossible to buy a car that does every single thing "right", by any single person's definition.

> the fuel gauge ramps rapidly up to Full

That's funny! Does it also fluctuate when you take sharp turns?

No, but that's just the point. It's smart enough to know that at key-on, it has very little information about the actual fuel level, so it allows the needle to move rapidly. As it continues to operate, it gradually (probably over a few seconds) filters the fuel level measurement more, so you don't get rapid fluctuations while driving.

Your anti-intellectualism is quite obnoxious. It doesn't help your case that you are actually wrong on top.

The poster above you has it correct. Climate control cars will not turn up the fans until the engine block heats up. Turning up the temperature does nothing to change this.

Please stop with the attitude and name calling.

I love that you're telling me how my car (cars, actually, this has been true of all of them) works, and I'm the obnoxious one. :) Clearly I'm not wrong. You are arguing on the basis of what can be done by software given perfect inputs. I'm telling you what is true on real systems. And the fact that you can't distinguish the two is precisely the problem I'm getting at.

That's not "anti-intellectualism", it's called "correct requirements analysis". And it's something I wish more software people knew how to do.

I am certainly not telling you about your car. Your car is very much irrelevant. The topic is very recent luxury cars with climate control. Even if your car would qualify for such a category, it's behaviour does not generalize to all such cars.

A great deal of people beyond yourself drive fancy cars with climate control, myself included. Your car is a single data point. My car is another. That my car, and other posters cars, behave as we say is enough to show that your generalization is wrong.

It is still all anecdote of course; to really determine truth you would need to do a study of as many cars as possible. Further, there is no theory in such a generalization.

This is where your anti-intellectualism comes in. You repeatedly make false statements about theoretical works: "theory is always wrong". It is almost as though you do not understand the definition of "theory." Additionally, your use of "egghead" when referring to engineers is classic anti-intellectualism. I mean, why not just throw out nerd, geek, or pointdexter too? Using anecdote to support your generalization is another extremely common undesirable behavior.

Please stop.

And you're letting your implicate trust and bias into everything that Tesla releases show. Tesla has just as much to lose in this whole ordeal and could just as easily be fudging the numbers.

It seems like Tesla would be demanding an apology and a retraction if they were fudging the numbers. Instead, they are asking for an investigation.

Two words:

Jayson Blair.

Reply on news.ycombinator.com