What do servers need?

[Camel]

I always used to install svn. I'd host one server on my LAN, and then install the client on all other machines I frequented. I used to have 4 or 5 machines that I'd switch between, so it eliminated redundancies. I would log in to desktop A as tmp, hack around for an hour or two on a variety of different things. Then I'd commit to the server repository and log out. svn would automatically tabulate all of the files modified during the session and store them for me. I could leave for lunch, log in via laptop from the restaurant, check a copy of my home directory out, and be exactly where I left off; down to the wall paper and xmms volume. Rinse, repeat. I was always surprised how few people it seemed exploited this functionality.

Another is vtund, but depending on which version of the kernel you selected you may not have the necessary module.

That's a cool application for svn, I never thought of using it that way.

I use a glorified version of networked mounts for that. There are tools that will do the syncing in the background for you. If I can remember what the name of the one I use is, I'll post it.

[nslay]

From a programming point of view, MySQL is uber sucky.  Try PostgreSQL if you can.
(An example of why MySQL API sucks...no ability to do non-blocking transactions!)

[Camel]

From a programming point of view, MySQL is uber sucky.  Try PostgreSQL if you can.
(An example of why MySQL API sucks...no ability to do non-blocking transactions!)

MySQL is better for smaller databases, by a long shot. By "smaller databases," I mean everything smaller than the data mined by Google's search engine.

[iago]

From a programming point of view, MySQL is uber sucky.  Try PostgreSQL if you can.
(An example of why MySQL API sucks...no ability to do non-blocking transactions!)

I personally use MySQL because I'm used to it and comfortable with it. That's sufficient for me. :P