I like to think of C++ like this: Just because a person plays a game that is difficult to master and loses all the time doesn't make it a bad game. It's that the player doesn't bother or is incapable to learn the finer points that will allow him/her to become better. Like a good game, C++ has layers to its strategy that takes time to discover and master but if you're just starting out with basic knowledge it is still "playable". The problem with C++ is that it is suffering from feature creep and bolted on fixes to the more broken aspects of the language all for the sake of backward compatibility. I've run across code from 2001 and it still compiles with the modern compilers and runs on the latest kernels of your favorite flavor of linux.

In this day and age of VMs and high bandwidth connections, I've downloaded quite a few ISOs (or grabbed a few from Linux Format) and at least poked around a little. I've tried Sabyon (?), Mandriva, Damn Small Linux, Puppy Linux, Mint, Knoppix. But the one's I actually do anything useful on are as follows in order Slackware (floppy) -> Redhat (floppy to CD to DVD) -> Fedora (CD to DVD) -> CentOS -> Ubuntu I develop seriously on CentOS and Fedora. Redhat was painful to develop on up until 5.0 then settled down and became more standardized.

I agree with other posts about using XAMPP (http://www.apachefriends.org). I've used it quite a bit for development work and I found it super easy to install. It supports Linux (debian and redhat among others) and Windows. It comes with MySQL and SQLlite with the phpMySQLAdmin admin tool, PHP and perl for development). The best thing about it is Apache, MySQL and PHP are all pre-configured. There's even an FTP server included. I'm not too sure about security though as I never went live using it, but if you read up on server (esp. web server) security, you can configure it pretty easily from what I understand.

I agree with writing actual applications to hone your new knowledge. My first real applications in java were a calculator and a CD collection application. I wrote these without an WYSIWYG IDE. It helps to really see how things work if you don't use a GUI "builder". If you want to get ideas for good projects you can read "Java Cookbook" published by O'Reilly. It provides some programming example projects that can be expanded upon. Finally, read a good algorithms book (can't think of any examples, sorry). Java is great for supplying many objects that hide the implementation, but it's still helpful to know the theory behind these implementations.

Right. In the U.S., fingerprints are kept indefinitely. At least in the local law enforcement offices. But they are viewable/searchable by other agencies as well. The reason they keep these is for identification purposes other than crimes such as after death or missing person, etc. I'm sure DNA will be kept for the same reasons.

Let's say the hardware DOES cost $150. I think over the lifetime of the hardware they can more than recoup the cost. I think it's in the realm of possibility to get $10 add revenue per month per user. That includes search revenue and adds splashed all over everything. They get their money back after 15 months. Let's say the average lifetime of the hardware is 2 years, they make money after a while. Of course, they make money sooner as the hardware gets cheaper.

Yes, deprecation of MySQL is my fear. It seems to me that not keeping some form of MySQL free is like biting the hand that feeds it. There is plenty of cash flow in the form of services that can be had. I'm hoping that if Oracle cannot support MySQL development/maintenance, it will be spun off into a separate business arm ala RedHat/Fedora. BTW, there already is a fork of MySQL, Drizzle (http://drizzle.org)