Language Selection

English French German Italian Portuguese Spanish

Beastie of an OS

Filed under

Once a distro goes into beta 3, most of the major choices are in place. In looking at the 3rd testing versions of distributions, one can get a fairly good idea of what a distro might be like once it's released. The only experience I've had with a BSD clone or derivative was with my PC-BSD review some months ago. That install was as simple as 1, 2, 3... or click, click, click. I'd heard the horror stories about other BSD installs, yet downloaded 6.0 beta 3 with hope. Was this going to be a brain-burning, hair-pulling, data-losing proposition? What happened with my attempted FreeBSD 6.0 Beta 3 install?

As this is my first foray into FreeBSD, this isn't so much a "what's new" as it is a "what's here".

First off, the install was much easier than running it... at first. But as with many new things, once you learn how, you wonder why you were nervous to begin with. The installer was easy enough. I had read the docs on FreeBSD before and as I recall, it sounded like a cross between lfs and gentoo. But if that was true then, it certainly isn't true now. The FreeBSD installer was a nice ascii graphical type installer that walks one through the install in much the same manner as Slackware. Can you install Slackware? Then you can install FreeBSD. In fact it even looks very much like Slackware's.

The most difficult step for the newcomer might be the fdisk step. I even experienced a sweaty-palm moment. The FreeBSD fdisk didn't seem to see all my partitions, or rather it saw the extended partition as one big empty slice. I toyed with the idea of inputting the start and end block numbers in and seeing if it installed on the correct partition, but chickened out of that. It was already complaining that it didn't agree with the geometry reported for the partitions it did see. I chose to put FreeBSD on the first partition of the drive - former spot reserved for, if not the current home of, Windows. It is now a Unix slice.

The rest of the install is fairly straight forward. One picks out the type of install they'd like, if I recall correctly, something like: developer, developer + kernel, developer + kernel + X11, etc., or as I chose ALL. It takes about 10 minutes to install the system, then it asks about packages and ports. I chose many many packages I'd need including KDE, gnome, and all the other window managers available during install. Turns out there are many many more available through their package manager. This step takes some time, probably a 1/2 hour or so, but then it gets to the configuration portion. It asks some questions about your net connection preferences, root password, setting up users & groups, and some other hardware. All this was quite easy to follow and complete.

I didn't choose to install its bootloader, instead I googled and discovered one only needs an entry in their linux lilo.conf very similar to ones we used for Windows. In fact, it's almost exactly like that. Mine looks like so:

Then run lilo and yippee! Upon reboot, lilo hands off to the FreeBSD
bootloader and your new system boots as desired.

One is booted to a terminal for logging in. First thing I always do is setup X. I fired up my console browsers in an attempt to download the NVIDIA drivers, but that failed as NVIDIA changed their site since I last downloaded their drivers with a text browser. I used to think how nice that one could use Links/Lynx to do that, but now their stupid javascript license agreement ruins it. So, I improvised. Since my FreeBSD is still not seeing anything in my extended partition, I had to make other arrangements. This was all in vain as the install bombed out very early on. It shot an error something about NOOBJ is deprecated to NO_OBJ or some such and I knew it was vesa for me. Xorg NV drivers lock my machine up fairly tight no matter the boot options I use.

However, there was no /etc/X11/xorg.conf skeleton in place and copying one from another install wasn't an option, so I was left to run Xorg -configure. This sets up a test file in /root called, and one can test their configuration with Xorg -config If it works well, then you can cp it to /etc/X11/xorg.conf, and I did.

Now to start KDE, or actually more accurately, KDM. I wanted to be able to check out all the window managers and figured KDM was my best bet. But where the heck was it? As with many Linux commands, fortunately "which" is in my BSD Unix clone and it worked quite well. I found xinit was located in /usr/X11R6/bin and kdm was located at /usr/local/bin/kdm. So su to root and issue the command /usr/X11R6/bin/xinit /usr/local/bin/kdm and we are in business. In the future to expedite things, I learned startkde was in /usr/local/bin/startkde. One finds the standard and complete KDE 3.4.2 upon startup or one of many other window managers.



Many ports get installed into /usr/local with FreeBSD and there is no /opt directory. In fact the directory structure may be similar in some ways to Linux, but to me, it was more different than alike. Many binaries are located in /usr/libexec and /usr/X11R6/libexec. But how does one find something not in their path? As you might recall in Linux systems, you can't use locate or slocate until you build the database, and regularly update it. But "which updatedb" didn't turn up anything. Thank goodness for google. To build and update that locate database, one needs to issue /usr/libexec/locate.updatedb

The kernel sources are located in /usr/src/sys/i386/ and the modules reside in /boot/kernel. I don't know which kernel I'm actually running, as uname -a reveals

tuxmachine# uname -a
FreeBSD 6.0-BETA3 FreeBSD 6.0-BETA3 #0: Mon Aug 22 22:59:46 UTC 2005 i386

I supposed I was still thinking Linux and expecting 2.6something. I try to remember we're dealing with a horse of a different color here. Anyway, at this point, if support for something wasn't in default, then I just won't use it. Maybe later.

One of those things not in the default kernel build was support for my bttv card. But sound was there and instead of modprobe snd_emu10k1, one issues kldload snd_emu10k1. For convenience I googled again and found that /boot/defaults/loader.conf is where one sets up their modules to autoload upon boot. Some commandline equivalents might be:

  • kldload = modprobe
  • kldunload = rmmod
  • kldstat = lsmod

But what about installing other software? I always like to have mplayer installed and GIMP is a must-have. But what do I do? Well, google of course. I found that the installer for FreeBSD is pkg_add. A lot of software is located in /usr/ports/. One could just navigate to the package directory of choice and issue a make install or one can use pkg_add <name of package>. Using the -r flag tells it to search remotely and get the latest available. It tries to sort out dependencies as well, but if there are issues, one might try portupgrade <package name> Mplayer isn't available, but gimp is as well as bash_completion.

There are many similarities between FreeBSD and Linux, but there are subtle differences as well. One major difference is the naming convention of devices. For example, ethX are vrX and hdX are acdX. As stated the directory structure is quite a bit different and I found commandline flags must be typed before the filename.

So, all in all, I found FreeBSD to be a capable desktop system. I've experienced a few konqueror crashes, but no other stability problems. I think their strongpoint is still in the server market and I'd probably appreciate it more there. If one checks in with Netcraft, they will find that almost 1/2 of the longest running systems by average uptime are FreeBSD.

I now recall how it feels to be the newbie stumbling around in a strange operating system. One wonderful resource where I found some answers to some of my issues is the BSDWiki. There is also some documentation as well as latest news on the FreeBSD website. I could very easily adapt to FreeBSD if something catastrophic happened where all the Linuxes (Lini?) suddenly vanished off the face of the earth. I can't say what's new in this release since the last stable or even the other betas, but I can state that many of the applications are of the lastest (stable) versions available. Try it, you might like it!

I have some additional Screenshots in the gallery.

More in Tux Machines

Android Leftovers

Intel Core i9 9900K vs. AMD Ryzen 7 2700X Linux Gaming Benchmarks

Complementing the just-published Intel Core i9 9900K Linux benchmarks with the launch-day embargo lift are the Linux gaming benchmarks... This article is looking at the Linux performance between the Core i9 9900K and AMD's Ryzen 7 2700X in a variety of native Linux games as well as comparing the performance-per-Watt. So if you are a Linux gamer and deciding between these sub-$500 processors, this article is for you. If you didn't yet read the main article that features a 15-way CPU comparison on Ubuntu 18.10 with the Linux 4.19 kernel, here is a recap of this new Coffeelake refresh CPU. The Core i9 9900K is an eight-core / sixteen-thread processor with 3.6GHz base frequency and 5.0GHz turbo frequency. This 14nm CPU has a 16MB L3 cache, dual channel DDR4-2666 support, and a 95 Watt TDP. There is also the onboard UHD Graphics 630, but if you're a gamer, that isn't going to cut it. The Core i9 9900K is launching at $499 USD. Read more

Intel Core i9 9900K Linux Benchmarks - 15-Way Intel/AMD Comparison On Ubuntu 18.10

Intel sent over the Core i9 9900K as their first 9th Gen Coffeelake-S CPU hitting store shelves today. With the embargo on that now expired, let's have a look at how well this eight-core / sixteen-thread processor performs under Linux. The Core i9 9900K is Intel's new answer for competing with the likes of the AMD Ryzen 7 2700X, but does come at a higher price point of $499 USD. While the Core i9 9900K is a Coffeelake refresh, rather than being six cores / twelve threads, they are matching AMD's precedent set by the Ryzen 7 processors in having eight cores / sixteen threads. This 14nm 8C / 16T processor has a base clock frequency of 3.6GHz with a turbo frequency at 5.0GHz, a 16MB L3 cache and supports dual-channel DDR4-2666 memory. Read more

Google: Desktop, Server and Kernel

  • Chrome OS Linux support to gain folder sharing, Google Drive, more
    Chrome OS has been shaping up to be the all-in-one system, combining the best of Google’s ecosystem, including Android apps, with the power of Linux apps. The latter is still in beta phase with improvements and new features in every update. Today we take a look at some of the features coming soon to Chrome OS Linux apps. Chrome OS first gained its Linux app support, also known as Crostini, with version 69. While it’s certainly not flawless, the support has been groundbreaking, enabling everything from full photo editors to Android Studio on Chrome OS. With upcoming versions of Chrome OS, Google is working to smoothen the rough edges of Crostini to make it easier to use.
  • Google Cloud CTO Brian Stevens on using open source for competitive advantage in the public cloud
    As all three continue to vie for the affections of CIOs, how they market their respective public cloud propositions to enterprise IT buyers has subtly shifted over time. For evidence of this, one only has to look at how little fuss the big three now make about rolling out price cuts for their services compared to several years ago, when one provider announcing a price drop would not only make headlines, but prompt its competitors to publicly follow suit too. This in itself is indicative of the fact enterprises expect more from providers than just access to cheap commodity IT services these days, and that ongoing cost reductions are simply an accepted part of using cloud, Google Cloud CTO Brian Stevens, tells Computer Weekly.
  • KUnit: A new unit testing framework for Linux Kernel
    On Tuesday, Google engineer Brendan Higgins announced an experimental set of 31 patches by introducing KUnit as a new Linux kernel unit testing framework to help preserve and improve the quality of the kernel’s code. KUnit is a lightweight unit testing and mocking framework designed for the Linux kernel. Unit tests necessarily have finer granularity, they are able to test all code paths easily solving the classic problem of difficulty in exercising error handling code.