In the last post we talked about computer memory and there was that weird line in the free command results called swap.  I also promised to talk about it this week.

So there you are in Math class one day and your teacher writes the most complex equation on the board you’ve ever seen and says ‘SOLVE IT!’.  What do you do?  Well, you pull out a piece of paper and get to work.

swap is kind of like that piece of paper.  The computer wants to do all its thinking in the memory it has installed for that, but sometimes that doesn’t work.  So instead of pulling out a piece of paper and doing work on that, it uses swap.  It would be more fun if your computer could stop what its doing and pull out paper, but it wouldn’t work as well.

So what exactly is swap anyway?  Well, when you set up your system, you designate a part of your hard drive to be the swap space.  This means that the computer is using its remembering memory for thinking memory.  Sounds weird, but it works.  You remember when you think sometimes, right?  You remember how to do the math problem while you’re thinking about the math problem.  Or you’re thinking about how Iron Man would solve it rather than actually solving it, but hey, whatever keeps the math teacher happy!

So let’s look at the result of that free command again:

Alpha:~ computerlamp$ free -g
             total       used       free     shared    buffers     cached
Mem:            31         29          1          0          0         12
-/+ buffers/cache:         17         13
Swap:            2          2          0

According to this, I have 2 gigabytes of swap to use and I’ve used both of them.  In other words, I’ve used up all the paper I had free to do my work on and there’s no more available.  Not exactly the best spot to be in, but I’m sure I can borrow some from the guy next to me… except that doesn’t really work with computers.    But your administrator can add swap if she decides she needs it.

The Linux® kernel is not one huge thing that can run every hardware ever made and talk to every application ever created.  Can you imagine that? It would be huge and every time something new was added, you’d have to add it on.  Even if it wasn’t anything you needed.  Instead, the kernel is a modular kernel, which means you can add pieces on as needed.  These pieces are called modules.

It’s like Lego® blocks.  You add the blocks on that you need but you don’t have to use every single one.

Lego Blocks

Just like that picture.  If I’m making a wall, I don’t need to add this head:

Lego Head

It doesn’t make any sense to add that to a wall, does it?  But if the rule is ‘use every single Lego® block you own no matter what’, then even if you’re making a wall, you have to add that on .  Maybe to the top, it would be good at scaring intruders, or at least making them laugh.

With the Linux® kernel, you don’t have to do that.  You can use modules to build your kernel so it does exactly what you need and nothing more.

It’s hard to find computing hardware that Linux® won’t support.  It runs motherboards, networking connections, webcams, security systems, CDs, DVDs, mice, keyboards, card readers, I could go on and on.  In fact, it might be easier to list things it doesn’t support, but… I really can’t think of anything.  If someone really wants to use Linux® to run a piece of hardware, they’ll either write the kernel module or find someone to write it for them.

You may not use all of the modules in Linux® but when you download the source code for it from https://kernel.org (that’s where the most recent source code is) you do get almost all of the modules along with it.  You can pick and choose which parts you want when you build it, which is something we’ll talk about in the future.  That’s because it’s about how to compile a program from source code, a more advanced topic than this.

I said ‘almost all’ because sometimes hardware manufacturers offer the modules separately.  It depends on the manufacturer and the people that maintain the kernel source code.

We’ve talked about how your computer gets places on the Internet.  Now let’s talk about how your computer, well, works!  How does Linux® do its thing, aside from  typing commands at the command line?  It’s called the kernel and it’s the core (or kernel) of the operation of your system.

So it’s not like the kernel of your popcorn, it’s more like the thing in the middle.  Like Nick Fury.  He’s always in the middle of the Avengers keeping things going.  Or making things go.  He’s always up to something.

Nick Fury

He showed up with that helicarrier at the end of The Age of Ultron, didn’t he? He made that happen and got the Avengers (and the towns people) from the flying in the sky town (something I never want to see) and on to safety.

That’s what the kernel does.  It’s in between the hardware and the software.  Like this picture:

What the Kernel Does

Linus Torvalds made the first Linux® kernel in 1991 for his own system. These days it runs on all sorts of things, but back then, he just wanted something for his own computer.

So how does it control your hardware?  It uses things called device drivers.  It lets you type on your keyboard and have it show up on a screen, or print to a printer, or read a CD, or read a USB, or… you get the idea.  It’s really flexible.

The kernel is also  modular.  That means that it isn’t one huge thing that keeps things going, but it’s made up of lots of little systems that put it all together and make it go.

It’s kind of like a Death Star built from Legos.  It needs all the parts together to make it a Death Star, but you can leave pieces off and still mostly have a Death Star.  You can add pieces to it and it’ll be a Death Star with more abilities.  Like if someone added an ice cream maker to the Death Star.  Or a spa for the emperor.  It’s still the Death Star and still acts like it, it’s just more!