I started with various British computers (Sinclair QL and Acorn Archimedes), pretty soon acquired a RISCiX (Acorn ARM based Unix) machine. Moved to AMD/Intel hardware with Linux as OS. DOS/Windows is for work... and usable when it comes with a system administrator.

Nothing in the "Encrypted Media Extension" specs prevents or forbids proxying of both the key and the encrypted media stream to an external "decryption and caching" service. And all of the usual "how do we prevent the plaintext from leaking from the user's machine" questions are still in full force. It is unlikely that the W3C will get "effective protection".

KPN tried "carrier grade" IP4-NAT in the Netherlands a decade ago... Unfortunately the router software was too buggy and made the routers trash and crash. And how can the customers of the ISP run servers on their computers? NAT has implications for the peer-to-peer nature of the Internet.

There already is a picture of Mickey crossing light-sabers: http://news.bbcimg.co.uk/media/images/63819000/jpg/_63819199_starwars.jpg (Minnie is in the picture too. *sigh*)

Canadian $, Australian $ or Zimbabwe $?

I agree that the biggest danger is close to home, family and friends of the family. And while there are "predators" on the net they are far less dangerous than the predators the child may meet in real life. Children are pretty safe with the online equivalent of "don't go with the stranger offering you candy."

What are some good rules of the thumb:

• Don't talk to people you are not comfortable with.
• Don't tell where you live. "Near Big City" is good enough for someone until you trust him/her.
• Be careful with what you show on your webcam.

If you following the advice the Internet is a good place to experiment with political and sexual discussion, pregnancies and STDs come from meeting IRL.

It depends on the exact bug that is triggered. When a security researcher mentions "potentially exploitable bug" it could be serious. Very often a memory corruption is a first step into more serious exploits.

For the people that were lucky enough to use it on the Acorn Archimedes...

While I don't mind putting some restrictions on someone while (s)he's on probation; the laundry list of conditions sounded like something typed in a decade (or more) ago with some conditions added over time. It is so convenient to have a standard list, without regard for the probatee or his crime!

It is good to have relevant restrictions as conditions for porbation (no alcohol for people convicted of intoxication related crimes), but I don't see any good in a total restriction of computer use for a petty thief (unless he brokers on ebay).

You insensitive clod!

But there is no reason why I should make it easy for my ISP, upstream providers and the NSA to eavesdrop on communication with a webserver. Encrypting my communication plugs a big privacy hole: so, why not use it?

I have a program that generates CSV output, for import into a spreadsheet. Open Office Calc creates the sheet I expect (slowly but correct). However, Excel sees it fit to mess op the data; it arbitrarily breaks long lines mid-field, creating a mess that requires more time to clean up than waiting for OOo to import correctly.

N.B. OOo suffices for most of my business correspondence; I prefer (La)TeX/LyX for the more scientific documents.

They say that some oil companies are heavy into alternative energy, buying start-ups and shelving their products. For them, "Never" is the answer to your question!

I assume the consumer does not have to pay to be part of a beta test program. Consumer protection law in my country (.nl) takes the price paid for the product into account when determining how much quality a "reasonable consumer" might expect. There is no need for a computer game to be perfect, as long as it is playable. Our judges are likely to handle "paid beta" software as any other paid for software: It should work, for reasonable definitions of work.

Most of the storage media in common use (disks, tapes, CD/DVD-R) already do use ECC at sector of block level and will fix "single bit" errors at firmware level transparently. What is more of an issue at application level are "missing block" errors; when the low-level ECC fails and the storage device signals "unreadable sector" and one or more blocks of data are lost.

Off course this can be fixed by "block redundancy" (like RAID does), "block recovery checksums" or old-fashioned backups.