car crash

[cellio]

No, not hardware.

I had to reboot my car this morning. I still don't know what happened.

My car, like almost every car built in the last N years, has a remote control. There are buttons for "lock" and "unlock"; for the latter, press once to affect the driver's door only and twice to affect all doors. (A single press of "lock" affects all doors.) The car does not appear to preserve state for any significant length of time; I have not conducted experiments yet to determine the timeout between presses of "unlock".

This morning I needed to get something out from the passenger's side, so I pressed twice, retrieved the item, and a few minutes later got into the car and drove off.

Now when you lock the car using the remote control, the car's alarm system is supposed to automatically activate. You know that this has happened because of a little light that comes on. There is no documented way to deactivate the alarm.

When I got to work and locked the car, the light didn't come on. I pressed the lock button again, thinking my car was somehow weirdly stateful after all; no change. I unlocked and relocked; no change. Eventually I unlocked the car, got in, put the fancy electronic key into the ignition, removed it, got out, and locked. That time it worked.

I still have no idea what happened. This is not covered in the manual, nor did the UI provide sufficient hints.

Comments

can we blame Bill Gates for this?

[sethg]

I have this sinking feeling that our grandchildren will consider it perfectly normal that objects they use in their day-to-day lives will just mysteriously fail to work every once in a while ("dang, the front-door lock has hanged again, I'll have to go in through a window and toggle the circuit breaker"), and nobody will seriously consider this a sign that the object is poorly made, because you know, it has some kind of computer with software inside, and everybody knows that software just fails to work every once in a while.

I mean, it's not like any reasonable person would expect a computer program to behave in a way that's predictable.

Grrr.

Re: can we blame Bill Gates for this?

[cellio]

I share your sinking feeling.

It's possible that in this case it wasn't unpredictable but that I don't have all the necessary information. If so, though, that doesn't make it any less annoying; even if a problem is completely predictable, if the set of rules for predicting it is complex or counter-intuitive, that's still a problem. For example, I can predict that one of our machines at home (which I'd like to salvage but not spend lots of money on) will fail within 30 minutes of being turned on, but that predictability doesn't make it ok. :-)

Re: can we blame Bill Gates for this?

[tangerinpenguin.livejournal.com]

As opposed to believing it's perfectly acceptable that when something stops working (or making a weird noise) it can be "fixed" by giving it a good whack on the side?

People work around low-level design flaws rather than insisting they be fixed in the first place all the time, doing more so the more complex the system in question is. People forgive random failures in software that they can work around quickly precisely because software is complex enough to do a lot of unusually cool things when it does work.

Not that better quality isn't a worthy goal, and not that a lot of design lazyness has crept into software development, just that this isn't really some new phenomenon that's unique to computer use.

[Now, the meme that if you aren't trying to make it impossible for anyone to break in you aren't serious about security, when in almost every other context known to history security's been a matter of making it generally cost-ineffective for crooks to break in, that we can blame on the world of computers... :-)]

Re: can we blame Bill Gates for this?

[tangerinpenguin.livejournal.com]

"... not that a lot of design lazyness hasn't crept into ..."

He says, having let design lazyness creep into his writing :-)

Re: can we blame Bill Gates for this?

[cellio]

As opposed to believing it's perfectly acceptable that when something stops working (or making a weird noise) it can be "fixed" by giving it a good whack on the side?

Point -- though I believe side-whacking has its origins with devices that had actual moving parts that could actually become wedged, which is a little easier to understand than rebooting because your web browser froze. (I also don't have a sense of how common it is. Everyone who owns a computer has had to reboot it for mysterious reasons somewhere long the line, possibly many times, but how common is it to whack your TV or your toaster?)

People work around low-level design flaws rather than insisting they be fixed in the first place all the time

Yes. The perfect product doesn't exist and the individual doesn't have the power to force design improvements, so selecting a product revolves in part around deciding which design flaws are acceptable. This has been true long before computers, of course. Also, there seems to be a fairly short "burn-in" period during which you'll actually work to fix something; e.g. if you buy a house thinking you'll change such-and-such about the kitchen, you better actually do it within months or it'll be too late because you've learned to live with it. I think this applies broadly -- once you've learned to live with a design flaw, you no longer notice it as a flaw. It certainly applies to software. :-)

[zare-k.livejournal.com]

[I have not conducted experiments yet to determine the timeout between presses of "unlock".]

On my car this timeout is relatively short (< 10 secs). Also you may have noticed that if you press unlock and then don't open a door, the car will relock itself after a minute or so.

I'm not so much disturbed by the idea of people learning to put up with flawed design-- as someone else observed, rebooting your PC or thwacking the toaster on the side to unstick it aren't all that different in concept. What I do find disturbing is the move towards everything being disposable. If your printer cost $40, why go to the trouble of repairing it? Just toss it and buy another. A friend went looking for a replacement mop sponge, and found that it was actually cheaper to buy a whole new mop than to get the replacement sponge. As annoying as it is to be forced into finding workarounds for faulty design, I'm not sure that a culture that encourages people to just give up and buy anew when their widget breaks down is really an improvement.

[cellio]

Also you may have noticed that if you press unlock and then don't open a door, the car will relock itself after a minute or so.

I've noticed that. They specifically document this behavior with respect to the hatch, but not generally. I have many complaints against the documentation, actually, but it's not something I'll need to refer to frequently so that's ok.

Disposables: I also find this frustrating. I get odd looks from jewelers when replacing watch batteries that cost almost as much as the watch, and as you've pointed out, sometimes reuse actually costs more. (I've had the mop-head experience too.) And the landfills get bigger and bigger... not a good thing to be teaching people.

Mind, I'm not perfect by any means; I use my share of disposables and decide to replace rather than repair sometimes. I try to be good but don't always succeed; such is life. I like to think that if I had kids I'd be more careful because of the message I'd be sending, but who knows if I actually would? (I don't plan to conduct the experiment.)

[alienor.livejournal.com]

There is no documented way to deactivate the alarm.

You should be able to get the dealer to do it if you want it off. The cavalier I had a few years ago was bought new in 1996 (when the remotes were just taking off). We had 3 (soon to be 4) drivers in the house, but the car only had the capacity for 2 remotes, so we had to get the remote-tied alarms turned off. I imagine they should still be able to do that.

[cellio]

There is no documented way to deactivate the alarm.

That wasn't a complaint -- just an observation. (If there were a "how to turn off the alarm" section in the manual, I would have checked to see if I'd done so accidentally.) In general I want the alarm, though I can imagine that there might be specific circumstances when I would want to turn it off temporarily. As far as I can tell, I "officially" don't have that ability and I "unofficially" do, though I'm not sure exactly how or if it's reproducable. :-)

[miz-hatbox.livejournal.com]

We have a car that has that sort of lock/alarm feature. The alarm won't unless all the doors are shut, so if the alarm doesn't come on when I lock the car, then I know that I have left a door ajar.

I am guessing that your alarm system might have a similar feature, and that when you thought you were rebooting the car with the fancy key, you were actually resolving the "ajar" state.

[cellio]

Hmm, could be. Near as I can tell all the doors are firmly shut now, though it's possible it passed through a funny state and that's why it needed to be reset.

[jducoeur]

There is no documented way to deactivate the alarm.

Hmm. In my car, there's actually a physical switch. You unlock the car and deactivate the alarm, then throw the switch into the down position; from that point, until you throw the switch to the up position and relock using the remote, the alarm is deactivated.

Of course, this isn't documented anywhere. I found it by accident: the switch is by the driver's left foot, and I simply knocked it by accident. I brought it down to the dealership, and realized that this was a FAQ when the girl at the service desk just said "hang on a second", crawled under the steering wheel, and said "try it now"...

[cellio]

Heh. You'd think that if it's that much of a FAQ they'd give you some hint ahead of time, whether in the documentation or from the dealer directly when you buy it. But maybe I'm reacting to the painfully-long lecture I sat through on how a dead battery could reset the anti-theft device on the stereo (yes, the stereo has its own anti-theft device), and what arcane steps I'll need to take if that happens. Mind, the arcane steps will have been long-forgotten by then, but I suppose the point of the lecture is to help me remember that there are arcane steps and I can find them in the manual. Shrug. Seemed a funny thing to focus on when, I realized after the fact, he had never told me how to get the car into reverse! (It's a funny push-then-move operation, which is different from my previous car. I hadn't realized until I tried to back out of my parking space that day at work that the test drive had not involved reverse. Oops.)

[jducoeur]

You'd think that if it's that much of a FAQ they'd give you some hint ahead of time, whether in the documentation or from the dealer directly when you buy it.

Well, this is the same dealership that was responsible for msmemory's horrible experiences this year. They didn't suck *that* badly when I bought mine, which suckered us into coming there again. But I should have gotten suspicious earlier...