David Both

I have written about how Linux can keep old computers running and out of landfills or recycling streams. My most recent article is “How Linux rescues slow computers (and the planet),” at Opensource.com and it discusses one of my computers that has a BIOS date of 2010 so is about 12 years old.

I now have an old Dell I am working on that has a BIOS date of 2005 which makes it at least 17 years old. According to the Intel web site the Pentium 4 processor was released in the 1st quarter of 2004 so that BIOS date is probably as good as anything to estimate the age of the computer.

You can see the specs for this system which are pretty minimal.

#######################################################################
# MOTD for Thu Sep  1 03:23:02 AM EDT 2022
# HOST NAME:            test1.both.org 
# Machine Type:         physical machine.
# Host architecture:    X86_64
#----------------------------------------------------------------------
# System Serial No.:    CXXXXXY
# System UUID:          44454c4c-5900-1051-8033-c3c04f423831
# Motherboard Mfr:      Dell Inc.          
# Motherboard Model:    0F8098
# Motherboard Serial:   ..Cxxxxxxxx00TQ.
# BIOS Release Date:    05/24/2005
#----------------------------------------------------------------------
# CPU Model:            Intel(R) Pentium(R) 4 CPU 3.00GHz
# CPU Data:             1 Single Core package with 2 CPUs
# CPU Architechture:    x86_64
# HyperThreading:       Yes
# Max CPU MHz:  
# Current CPU MHz:      2992.644
# Min CPU MHz:  
#----------------------------------------------------------------------
# RAM:                  3.142 GB
# SWAP:                 3.142 GB
#----------------------------------------------------------------------
# Install Date:         Wed 31 Aug 2022 09:26:13 PM EDT
# Linux Distribution:   Fedora 36 (Thirty Six) X86_64
# Kernel Version:       5.19.4-200.fc36.x86_64
#----------------------------------------------------------------------
# Disk Partition Info
# Filesystem      Size  Used Avail Use% Mounted on
# /dev/sda2       298G  9.8G  287G   4% /
# /dev/sda2       298G  9.8G  287G   4% /home
# /dev/sda1       974M  226M  681M  25% /boot
#######################################################################
# Note: This MOTD file gets updated automatically every day.
#       Changes to this file will be automatically overwritten!
#######################################################################
[root@test1 ~]#

It has a single 3GHz Pentium 4 processor with Hyperthreading so the equivalent of 2 CPUs. It originally had 2GB of DDR 533 RAM so I added 2 GB for a total of 4GB which is the maximum amount of RAM supported by this system.

Naturally, I installed Linux on it just to see how it would work. This old computer works just fine with the most recent version of Fedora. My next test is to see how well it runs a Linux VM using VirtualBox.

So yeah – when I say that Linux keeps old computers running, it is absolutely the truth.

I borrowed this computer from Digital Bridge in Wake Forest NC to do some testing for classes I will be teaching. They have a number of these old computers that were regifted to them but Windows is pretty useless on them. The only way to get them to work reasonably well is with Linux. Linux can also keep your old computers working far beyond the point where it is normally assumed they should be discarded.

In addition, Linux keeps these old computers safe from malware and bloatware which is what really causes them to slow down.

Here is an interesting article by Steven J. Vaughan-Nichols  at “The Register.”

https://www.theregister.com/2022/08/10/opinion_column_drop_windows_for_linux

Some of the highlights

• Businesses should dump Windows for the Linux desktop.
• …Windows’ security mess has never been just because Windows is more popular. I’d argue Windows is insecure by design.
• It [Windows] was never, ever meant to work in a networked world. So, security holes that existed back in the day of Windows for Workgroups, 1991, are still with us today in 2022 and Windows 11.

Digital Bridge will be participating in the next Wake Forest Area Chamber of Commerce Job Fair. Make sure you stop by and check out what opportunities we have available to help advance your career or business.

Check out the DigitalBridge Twitter feed.

I have done some additional hacking on my primary workstation after initial installation in the CoolerMaster MasterFrame 700. The picture shows the current state. There will never be a final result because this frame just cries out to try new things.

The first thing I did was add the pair of fans on the right side of the main frame. The frame I mounted the fans on is not part of the MasterFrame. It was originally part of my Thermaltake Core X9 case and I used it there for fans to cool the hard drives. It works perfectly on this MasterFrame 700 and I mounted it using the small flat-head bolts with washers provided with the MasterFrame. I use these fans to move cool air over the motherboard with its two M.2 SSDs, four memory DIMMS, and two plug-in adapters. One of those adapters is the Radeon Pro WX2100 graphics adapter.

I lowered the 5.25 media dashboard to leave more room at the top of that wing. I also added the BASH sticker.

You can see the original I/O panel just underneath the media dashboard with the power LED in the shape of the CoolerMaster logo. I have a problem with this panel because it has no reset button or storage drive activity LED. So I purchased the add-on box at the bottom of the picture. This little box contains power and reset switches as well as LEDs for power and storage device activity. It does not have USB or Thunderbolt connectors but the original I/O panel does have those. In order to have both panels active with their common functions, I purchased a set of cable splitters designed for just this purpose.

I originally installed the AIO radiator cooling fans on the left wing to suck air through the radiator. Unfortunately that did not allow a good option for air filters on the back side of the wing because the radiator mounting screws interfered with that. So I reoriented the fans to blow air through the radiator but installed 120mm screens with holes already in the right places between the fans and the radiator. So now I have clean air and a clean look on those fans.

It looks great and works extremely well. The motherboard and CPU temperatures are as much as 5 degrees Celsius less than they were in the Thermaltake case.

I just rebuilt my primary workstation in a new frame, the CoolerMaster MasterFrame 700. This is not a review but I just want to show it off a little.

This is an amazing frame for a gaming computer or a build just to show off. “Frame” is a much better word for this unit than “case.” It is really just a frame on which to mount things. Because of the many options provided, it is not necessary to mount things like the I/O panel with the power switch in one of the two “official” locations. I hacked that a bit and mounted it low on the right wing. The other locations would not allow me to see the power-on light.

The media dashboard that I mounted in the center of the right wing and just above the I/O panel, usually fits into a 5.25 front panel drive bay on most cases. Because there are no front panel bays like that, I drilled a couple small starter holes in the bottom of the dashboard and used the supplied screws to fasten it. Be careful not to drill too deep!!

All the parts in this build were moved from its previous case so there is nothing new here except for the frame.

This computer frame is a hacker’s dream.

I hope you enjoy this.

Some of my books and articles require you to create some files for testing. This can be done easily from the command line.

• Open a terminal session on your desktop.
• Copy the line below from this post.
• Paste it into the terminal session.
• Press the Enter key.

This Bash program creates 200 files with “Hello world filexxx” in each, where xxx is the file number. You can edit the program and change “200” to any number of files that you need.

for I in `seq -w 200` ; do echo "Hello world file$I" > testfile$I.txt ; done

Some files for experiments do not download properly

It was brought to my attention by reader Benjamin Robertson that experiments that use the wget command to download files from GItHub result in unusable files. This is due to an error in the URL and can easily be corrected.

For example, the command:

wget https://github.com/Apress/using-and-administering-linux-volume-2/blob/master/Experiment_6-1.txt

Should be:

wget https://github.com/Apress/using-and-administering-linux-volume-2/raw/master/Experiment_6-1.txt

Note that /blob/ should be changed to /raw/ in the above URL as well as others that use the wget command. The known instances of this problem are in:

• Volume 1, Experiment 12-1
• Volume 2, Experiment 6-1
• Volume 2, Experiment 6-3
• Volume 2, Experiment 17-1
• Volume 3, Experiment 17-1

This problem does not occur when the wget command is used with other sites such as RPMFusion and WordPress. It only occurs with GitHub.

I apologize for any problems this has caused you. I am working with Apress to determine why this problem has just shown up.

It has been a while since I posted anything here. For that I apologize, but I have been busy with a number of projects and life in general. I do think the release of Fedora 35 today does deserve a bit of attention.

As soon as I received the email notification from the Fedora Project of its availability, I upgraded my primary workstation to Fedora 35 and am now using it to write this post. I have only spent about 45 minutes with Fedora 35 so far. I think the best thing that anyone can say about any upgrade is that it went well – and it did. It was an easy upgrade and there have been no surprises so far. My Xfce desktop looks and works the same. Everything underneath also seems to be working as expected. This is a seamless and – so far at least – painless upgrade.

I did notice some interesting and different process names in htop, several sets of processes with numeric names from 0 through 31. I have an Intel i9 X series processor with 16 cores (32 CPUs) so I suspect that these processes are related to the CPUs. More research is needed.

I will post more when I have upgraded my other systems including my server and my firewall and have more data points.

In the meantime, here is a link to an article in Fedora Magazine, What’s New in Fedora 35.

One of the things I state in the Introductions about my books is that I use them myself as references and this week was no exception.

I perform many tasks on Linux computers for which I am responsible and a good deal of the time, that means remotely. And I don’t mean just from one room to the next. In many cases I need to do things over a distance of miles. I do a lot of things quite frequently and so I need no reminders of how to perform those tasks. But this week was different.

I had two problems on one computer for which I am responsible at a remote location and I did not have the time to drive there. Well – I didn’t want to drive there in any event. So, to fix both problems, I logged in to the problem computer from home using SSH.

And I used my own books as a reference resource for the tasks I performed.

Kernel problems

The first problem was with VirtualBox, the free open source software I use for running virtual machines. I knew I could fix this easily because I have done it several times; it is a problem in which VirtualBox cannot compile the kernel modules for the new kernel and I need to revert to the previous kernel. So I configured grub to boot to the previous kernel by changing the line GRUB_DEFAULT=saved to GRUB_DEFAULT=1 in /etc/default/grub. Normally the “saved” option specified in that line says that every time that Linux boots, grub saves the identity of the kernel that was booted and that becomes the default kernel for the next boot.

This resolves the problem until a kernel fix can be provided. I did not need much help with this but I did check Chapter 16 of Volume 1 of my self-study series “Using and Administering Linux: Zero to SysAdmin” to verify that I remembered the right option and the right file in which to make that change. I also needed a little help with the syntax of the command required to rebuild the grub configuration file using the revised grub default file.

[root@studentvm1 grub2]# grub2-mkconfig > /boot/grub2/grub.cfg

I don’t cover this specific situation in that series so I have added this information to my technical website at http://www.linux-databook.info/?page_id=6044.

Printer configuration

The second problem is that one of the printers available to the host was not available in the print menu when the office manager tried to print. The network printer was visible but the USB attached local printer was not. So I needed to create the print queue for that USB printer. Unfortunately, this is one of those tasks I only perform once every 3 or 4 years so I don’t really remember how to do it.

Using an SSH login session, I followed the step-by-step directions in Volume 2, Chapter 7, of my self-study series “Using and Administering Linux: Zero to SysAdmin,” Experiment 7-2, to add this print queue and test the result – successfully.

About my books

My books were written to be used as references and all of them started as memory aids for me. I would perform many tasks infrequently and it seemed a waste of time to keep doing the same research many times over. So after discovering new tools, fixes, and solutions, I started recording them so I could find them easily. Eventually I used that material as the basis to create a series of three, 4-day instructor-led courses that I taught for several years.

And then I decided to convert those courses into the books that I have written over the past several years. Because they are written in a way that can be used both for training and reference, I find them to be useful all the time. I think you will, too.

So, yes, I would really like you to purchase my books. You can find all of them listed here and specifically my self-study series “Using and Administering Linux: Zero to SysAdmin.”

I just learned something very cool.

Sometimes it is necessary to make remote modifications to a desktop using the command line. In this case I needed to reduce the number of workspaces on the Xfce panel from 4 to 3 at the request of the user. This is actually easy and only required about 20 minutes of searching on the Internet.

The xfconf-query command can be used by non-root users to query and set various attributes for the xfwm4 window manager. In the sample below I have first verified the current setting of 4 workspaces, then set the number to 2, and finally verified the new setting.

[user@test1 ~]# xfconf-query -c xfwm4 -p /general/workspace_count
4
[user@test1 ~]# xfconf-query -c xfwm4 -p /general/workspace_count -s 2
[user@test1 ~]# xfconf-query -c xfwm4 -p /general/workspace_count
2
[user@test1 ~]#

Also, the default workspace count and many other defaults for xfwm4 can be found and changed in the /usr/share/xfwm4/defaults file. So setting “workspace_count=4” to “workspace_count=2” changes the default for all users on the host.

Look for more details about remote Xfce desktop configuration and the xconf-query command here.