LinuxFox, by Alan Mackey Tux the penguin, the Linux mascot

The actual installation


So, now you've got the distribution and all of the auxiliary files. You've created and formatted your DOS partitions. You've got your DOS formatted floppy disks ready. Now you're ready to do the actual Linux installation.

The first thing you'll need to do is create the boot and root floppies. Use GZIP.EXE to uncompress the boodisk image you downloaded from the MCA Linux homepage:

    GZIP -d [bootdiskname.gz]
where [bootdiskname] is the name of your bootdisk image. You'll wind up with an uncompressed file that has bootdiskname. Now you put it on one of the floppies:
    RAWRITE [bootdiskname] a:

Next comes the rootdisk. You'll do basically the same thing with it:

    RAWRITE [rootdiskname.gz] a:
Note that the rootdisk image file is still compressed. Step two of the MCA installation instructions says that it should be uncompressed, but I tried that and found that the size of the image was bigger than the capacity of a floppy disk, something like 1.7MB or 1.8MB. I felt that Linux could do a lot of things, but I doubted that making a floppy disk hold more bytes than it was designed to was one of them, so I recompressed the rootdisk image and wrote it to the floppy that way. Everything worked just fine.

Finally, if you have an ESDI hard drive, you'll repeat the process one more time for the esdi_slack file you downloaded from the MCA Linux homepage:

    GZIP -d [esdislackname.gz]
    RAWRITE [esdislackname] a:

As noted in step 3 of the MCA installation instructions, make sure that none of your floppy disks is write protected.

Put the bootdisk floppy (the first one you made) in the drive and reboot. The Linux Loader will load and you'll get a

    boot:
prompt. Enter
    ramdisk
followed by any of the options in step 4 of the MCA installation instructions that you may need. Read through all of the information in that step to see if you need any of those options. In fact, I hope you've read through those instructions first anyway.

Now you'll see a bunch of messages about your system as the kernel loads. Soon you'll see something like:

VFS: Insert root floppy disk to be loaded into ramdisk and press ENTER
Remove the boot floppy, insert the root floppy and . . . . press enter. Some more stuff will load and then you'll get a login prompt:
Slackware login:
Type
    root
and press enter. You'll see a pound symbol (and nothing else). # That's your command prompt. Congratulations, you've booted Linux. :-)s

Now it's time to create your Linux partitions. If you have ESDI, follow the directions in step 7 of the MCA installation instructions to prepare for this. Otherwise, you can just start the fdisk program. Enter

    fdisk [drive]
where [drive] is the hard drive device name as listed in section 4.2.3 of the installation HOWTO.

As you're considering how to partition, you'll need to think about your filesystems. If you decide to divide your drive into multiple partitions, you'll have to make sure that you give the different filesystems enough space, or Linux won't install. The filesystems themselves will actually be created during the setup routine, when you specify your target(s) for the install. We'll be getting to that soon. :-)

You can play it safe and just make two partitions, one for the swap space and the other for everything else, which you then give to the / (root) filesystem; Linux will divide it up as it needs to. Or, you can choose to partition the disk into several filesystems. If you do, first re-read section 4.2.1 of the Installation HOWTO to get a feel for what you're about to do, because it's completely different from DOS/Windows. Also note that if you are using more than one disk, you will be mounting at least two filesystems, plus the swap space.

As an example, I divided my disks up into filesystems for / (root), /usr, /home, and XWindows, plus the swap partition. I guessed that the system and networking stuff, and the like, would be put in directories in the root filesystem; applications would go in /usr; /home would be for user files; and XWindows would go in the filesystem for XWindows. I was told later that I had made a pretty good guess. Note that if you allot a partition (filesystem) to XWindows, it must be named /usr/X11R6. If you don't, it will be created; if /usr or / haven't been given enough space, the install will fail (I've seen this happen -- just not to me :-) ).

So, using the information given in section 4.2.3 of the installation HOWTO, create your Linux partitions. One of these will be a swap partition. You'll have to enable the swap space before you run the setup procedure if you have less than 4MB of RAM. If so, use the instructions in section 4.3, immediately following section 4.2.3, of the installation HOWTO to do that now.

Now, finally, you're ready to do the actual installation. At the prompt, type

    setup
and press enter. You'll be presented with a menu of options, which are outlined and explained in section 4.4 of the installation HOWTO. I won't go into them in detail here, just make a few comments about some of the little "quirks" that I came across. The installation HOWTO number the menu options in a recommended order, which made a whole lot of sense and which I follow here.

The second recommended option, source, wasn't especially clear as to how to specify where the source of the software actually was if you're installing from a hard drive partition. Although the installation HOWTO says to enter the name of the partition (as in /dev/sda1), you don't get any clues as to entering the name of the directory afterwards. What happens is that you do just enter the name of the partition, and setup then goes to another prompt asking for the top of directory structure for the source files. For example, I had all of my packages under a directory named "C:\slakware" on the first SCSI drive. When setup asked for the source I entered

    /dev/sda
then setup asked for the directory to look under and I entered
    slakware

The Target and Disksets options are pretty self-explanitory, except that the Disksets option was called "Select" in my setup. However, it was pretty obvious what the option was meant to do. And this brings us to the "Install" option, which does exactly what it's name says. It's pretty straightforward, except for one little thing. The Slackware MCA instructions say in several places to install the kernel from the floppy disk, but it's not clear when this happens. It happens at the end of the installation of the base packages (the A series). During the installation of the base packages, there'll be two times when you'll be given the chance to install a kernel. First, when installing the stuff from diska7, you'll be given the option to install an IDE kernel; then, when installing the stuff from diska11, you'll be given the option to install a SCSI/IDE kernel. Note that both of these are optional; you probably want to tell the install to skip them. That's what I did and it worked out great. At the end of the base installation, setup came up and told me I hadn't installed a kernel, that I had to do so, and gave me the option of specifying where I wanted to install the kernel from. It was then I told it that I wanted to use the kernel on the floppy.

Section 4.5 of the installation HOWTO discusses what happens after the installation, such as making a boot disk and installing LILO. If you are installing from a DOS/Windows partition, make sure that you tell the LILO installer to write the configuration to the MBR. The LILO installer gives you several options, and notes which ones are safest, but (in my experience) if you don't write the LILO configuration to the MBR it won't work. This doesn't seem to be as dangerous as it's made out to be by setup, however, and I believe you'll have to install LILO if you want to have a dual boot system. Like the HOWTO says, make a boot floppy; do it before you install LILO and then you'll have a way to recover if something does go wrong. If LILO doesn't install correctly, you can reboot using the boot floppy you've made, and then configure LILO using the discussion in section 5 of the installation HOWTO as a guide.

At some point, you should be returned to the setup menu, where there is an option to configure your system. In fact, I believe that this option simply gets called to complete the steps that we just discussed. If, for some reason, you didn't install the kernel from the bootdisk, you can select this and do it from here. There are other configuration options as well. One is setting the time zone, which is set as a number of hours ahead of or behind GMT. For example, I live in the Central time zone, so I set my time zone as -600, six hours behind GMT. There is also an option to set up the mouse. You can go ahead and do this, but it won't do much for you. Mouse support is not built into the kernel on the bootdisk (which you installed as your kernel), nor is (in my experience) serial support. You'll have to patch and recompile the kernel for these things to work.

So, now you've completed the installation and configuration process, and you can select the EXIT option from the setup menu. At this point you're ready to reboot using the kernel image you installed on the hard drive. If you have SCSI, you should be ready to just enjoy your system (or go on to patching and recompiling the kernel). If you have ESDI, you have to complete steps 10 through 17 of the Slackware MCA installation instructions, and you'll have to do step 10 before you can reboot. Note that you can not just turn the computer off! You must tell the operating system that it is shutting down. You can do this in one of two ways:
1). Pressing

    CTRL-ALT-DEL
2). Typing
    shutdown -r now
Linux interprets both of these as the second command, and lets the rest of the system know that it is shutting down. It kills any running processes and takes care of various other "housekeeping" chores, then reboots the system. If for some reason you find it necessary to turn the system off, enter
    shutdown -h now
Linux will perform the same shutdown procedure, but will then display two messages that the system is halted. When the second one is displayed, it's safe to turn the system off.

Reboot the system and log in as root at the boot prompt (you've done this once, so now it's old hat, right?). Use the bootdisk if you've got ESDI, as you still have the remainder of the ESDI configuration to do. Take care of that, then reboot again, this time with the kernel image on the hard drive. Whether you have SCSI or ESDI, there are a couple of other things you want to take care of as quickly as possible. The first is to set a password for root. Enter the command

    passwd
You'll be prompted to enter a password and to verify it.

The second thing you want to do as soon as possible is to create an account for yourself. You'll want to log in and work on the system as this user most of the time, saving the root login for when you need to take care of various configuration issues. Use the information in section 4.6 of the installation HOWTO on the adduser command as a guide to create and configure the account.

And that's for the most part it. You can change the hostname of the system as discussed in section 4.6 of the installation HOWTO if you like; my system was named "darkstar" not "slackware" so I left it alone for a while as I worked on other issues. Eventually I did change my system's hostname -- to panhead. :-) If you only installed the base packages you'll probably want to look into installing the remainder of what you want to use. Remember, you can't patch and recompile the kernel without the development and the kernel packages (d1 and k1). You may want to just familiarize yourself with Linux commands before you proceed to this step, but you'll eventually want to do it, as system support for many useful things are not included in the basic MCA kernel that you installed.

On to the kernel page

Back to my main PS/2 page

MCA Mafia Banner