The real value of a self-driving car is just that, fully self-driving.

It's having something that can drive while you're asleep, reading, or maybe even working on your laptop.

It's something that can drive your 10 year old to school, drop them off, and then drive back to the house so that other people in the household can use the car.

And just as importantly, it's something that someone who is not fit to drive-maybe for medical reasons, maybe because they have not slept in 24 hours, maybe because they are drunk-can use to safely get where they need to go.

So no, the danger of salf-driving cars isn't that people will decide not to be in the driver's seat, the danger is that both automakers and regulators will try and give us supposedly self-driving cars that can't handle those cases, and then be surprised when things go horribly wrong, or when people just don't see the value in buying one.

Personally, I plan on ownning a true self-driving car very soon after I can buy one that can do the driving when I can't, and I bet that the vast majority of legally blind adults with enouh money will be right along there with me. But that won't happen anytime soon when people are acting like you need a driver for it to be safe.

The #1 priority, it has to be at least as good as my $30 Timex at what that $30 Timex actually does.

I could live with the battery needing charging every week, but not more often, half the point of my watch is being able to tell at a quick glance how much longer I have to sleep. (Without putting my glasses on, thanks, a clock on the night stand really doesn't help here.)

Better programmable alarms, alarm noises, and vibration alerts than I can get with a simple watch would be good.

Beyond that, give me a good heart rate monitor, and other basic sensors, and a good API to play with it all.

I'll probably be along, at least for a little while.

I'll go for either a new spine, or new lungs.

Just not sure which.

I'm a system administrator.

Well, alright, I also do my share of coding, but I'm a sysadmin first.

Assuming that it goes high enough, power disturbation. It's enough of a savings that every decade or so people talk about using current generation superconductors for it, need for cryogenic cooling and all.

Then making a lot of stuff that uses current superconductors cheaper, like MRI machines and particle accelerators.

Sure, I bet that there will be _plenty_ of new stuff, but I'm less convinced that anyone is going to be able to predict what that will be all that well.

Will the general public be able to buy the units? A lot of interesting low cost hardware has come about (like the OLPC), but it's been rare that people off the street have been able to buy them.

Even if there is a very explicit lack of support, it would be nice to just be able to buy them without having to be a school or having an order for 5000 of them.

When the weather changes, the leaves turn different pretty colors, and everything actually behaves like it's fall.

That's when fall starts.
(Date based methods are silly, especially these days.)

At least in the past, most ISPs will only block port 25 for non-business users. And most ISP level blocks are against the consumer DHCP ranges.

So spend the extra money on business accounts with static addresses, and take up the port blocking with your account manager if it's still an issue.

Barring that, spend $20 a month or so on a hosted VM somewhere, and run your mail through that. (Securely, please.)

A system administrator.