Why not just put a QR code near the entrance of the museum, or a web address?

Coral Cache isn't working for me, but google has cached it.

Someone should host the PDF on S3.

You could create a lot of variants if you just randomly drop a word from each one.

I would estimate that roughly half of people have less-than-average intelligence.

It's easy to spoof, but it's hard to establish a TCP connection by only sending outgoing packets. (Spoofing is the most useful for attacks on UDP protocols.)

I don't think that spammers have any amazing tech, they just have different requirements. They can still send spam with a 1% success rate whereas with OCR you'd want a 99% success rate.

This seems like an amateur mistake. Who are these companies hiring lately?

At the very least, hash and salt. If the hashes might be stolen then hash it thousands of times (see PBKDF2).

When I moved out of my mother's basement I used the Albertson's delivery service until they shut it down. It was $14 per delivery, regardless of size, so I'd get all of my groceries for the month in one order.

It was a lot easier to avoid impulse buying and to plan out what was actually needed when I could place the order online. Albertson's would remember your previous order so it was easy to just adjust it slightly each month.

I imagine that google is just going to restrict the use of the android trademark, not source code. This seems like a sensible solution, and won't restrict the rights of the manufacturers.

Also check out Musopen, a large collection of public-domain classical music recordings and sheet music. They take donations and use those donations to hire professional artists to make new recordings of the pieces and then put them into the public domain.

I think you're confusing speech recognition with natural language parsing. They are two different components. The reason natural language processing is powerful is the same reason the command line is powerful. With a GUI, you normally have to find your way through menus to get to a particular functionality. When there aren't many options to present or you aren't familiar with the system, this is a good interface. On the other hand, once you're familiar, a multitude of functionality is only a command away. (I have 4503 commands in $PATH.)

Once the speech recognition reaches a certain point you'll be able to call your bank and say something like "transfer $500 from checking to savings".

Anyway, here's the use case I picture: Anywhere in the house, I can say something in a normal voice, like "computer, what's the best way to stop a bloody nose?" or "computer, how long do i need to boil an egg?" or "computer, turn on the front-porch lights". Being able to interact with a phone in the same manner would be appropriate in some situations as well. Speech has the potential to be a great input device, especially when there isn't a keyboard handy. (Insert joke here about how none of us get far enough away from a keyboard for it to matter.)

What's the difference between this and the SMS service already offered by Facebook? Facebook's service works even if you don't have a special SIM card, and the last time I checked (several years ago) it had lots of useful features.

I believe that it's just WHATWG that has dropped the version number, not the W3C.

I'm doing this right now. It's as easy as apt-get install linux-image-generic-pae.