Linux Mint 19.2 Cinnamon vs Kubuntu 19.04 Review

It’s been almost 20 months since I last tested one of my highly respected GNU/Linux distributions; Linux Mint, the Cinnamon edition. Linux Mint 19.2 (based on Ubuntu 18.04), the latest release which I’m going to review, features the Cinnamon 4.2.3 which includes few new features, performance enhancements (such as a reduced memory footprint), improved theme and Kernel 4.15, among others. This operating system will be supported up to 2023.

As always, in this Linux Mint 19.2 Cinnamon review, I’ll be focusing mostly on its performance, such as the Boot-up speed, memory usage upon loading the desktop, power usage (idle), system responsiveness, shutdown speed, and system stability, although I will mention new features that I feel like worth writing about.

I have decided to compare it with the data I have gathered on a freshly installed Kubuntu 19.04. For this review, I downloaded the 64-bit ISO image (2.0 GB), and completely securely deleted the 3 partitions (‘/EFI‘, ‘/boot‘ and ‘/‘) on my rotational disk, that I used to install it.

I tested it on my Asus laptop and below are some of its hardware details:

Intel Core i7-5500U, Hybrid GPU Setup (Intel Broadwell HD Graphics 5500, Nvidia 920M), 4GB RAM DDR3, Hybrid Permanent Storage Setup (Seagate 5400 RPM, 500 GB rotational disk and a Kingston 24 GB SSD), Qualcomm Atheros AR9565 Wireless Adapter, Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller, Realtek ALC3236 Sound Card, LED Display (1366 x 768 resolution, 60 FPS/HZ). It's an Asus laptop (F302LJ-FN024H).

After installing the OS I opened the ‘Startup Manager’ and disabled a few apps (Update Manager, Print Queue Applet and SSH key agent if I remember correctly). Then I boot into it a few times (and shut it down a few times as well) to let things settle down. Booting into a freshly installed OS a few times is not always necessary, but, on certain ‘Linux’ distributions, there are boot-up services that do require a few boot-ups for them to properly set things up. The ‘ureadahead’ service (originally developed by Ubuntu, now abandoned) that Linux Mint still ships with, is one such example. On rotational disks, this service can speed up the boot-up times by around 30-40% (it’s useless on SSDs), in my experience.

The Installer

Linux Mint uses Ubuntu’s installer by default. It’s simple, stable and has worked great for me. I was able to install Linux Mint 19.2 Cinnamon with my custom settings (keyboard set up as Dvorak and custom partitioning scheme) without any issues. However, I’d say that compared to the installation times of Kubuntu 19.04, it was a bit slower, although I’m sure the installer had nothing to do with it, it was probably because of the amount packages it had to install or due to the level of compression they had been compressed with. The installer is also capable of automatically creating a ‘swap partition’ (a ‘pagefile’ type file that you find in Microsoft Windows), starting with Ubuntu 17.04, which it did in this instance as well.

The Desktop

Mint’s Cinnamon edition maintains a very Microsoft Windows looking user interface while rendering the desktop layout. Even though I have missed a few Mint releases, the overall look has not changed at all. That being said, there are subtle visible changes, if you look carefully. Take the desktop right-click context menu for instance (again, please remember that I don’t know whether these changes were there in the previous releases or not, simply because I haven’t used them. Therefore, I’m simply comparing them with the last Mint release I reviewed, which was Linux Mint 18.3).

Here’s a screenshot from Linux Mint 18.3:

Here’s how the same context menu looks like in 19.2:

‘Desktop’ entry which used to let users customize the icon arrangement on the desktop, is now replaced with an entry called ‘Customize’. Once you click on it, it takes you to a separate settings window that lets you set those same settings, though, using verticle & horizontal sliders (these sliders, for instance, let you further manipulate the spacing of the desktop icons compared to the previously existed fixed settings), buttons, etc.

Also, as you can see, the previously existed colored icons are now all converted into monochrome on the desktop context menus as well.

Under ‘Themes’ (‘System Settings’ --> ‘Themes’), you can also manually set the width of the scroll bars to your liking.

The file manager also comes with another new feature called ‘pin’. If you have files or folders that you access frequently, then by ‘pinning’ them (right-click on the file or the folder and select ‘pin’) will make them show at the top, whenever you visit that location. The names of ‘pinned’ files or folders will be shown in bold letters (below is an example).

Whenever you enter a password in the terminal app on Linux Mint 19.2 Cinnamon it now displays asterisks for each letter of your password. It looks pretty neat and even useful (for those that ‘hunt and peak’ while typing), but I don’t like it. It’s one thing to let others guess the length of your user password at login, but to show it every time you enter the ‘root’ (administrative) password on your terminal, I just don’t like it.

You can, however, easily disable this feature if you don’t like it. For that simply open up the terminal and enter the below command:

sudo vim.tiny /etc/sudoers.d/0pwfeedback

This will open up the configuration file in the command-line. Once the file is opened, press the letter i on the keyboard to enable editing the file. Then simply type # at the beginning of the line (as shown in the below screenshot).

Once you’re done, press the ESC key and enter wq! letters (without spaces) and press the Enter key, to save the changes and exit. That should do it.

I have always been a big fan of the default wallpaper selection of Linux Mint. They always include quite a nice collection of wallpapers with every release, and this release is no exception. Plus, you can also choose wallpapers from some of the previous versions too, from the ‘Change Desktop Background’ settings window.

These are of course only a few new features to mention. If you want a list full of what’s new in this release, please refer to this official page for details.

Web-browser Video Playback & Multimedia Codecs

I’m a big fan of the Google Chrome web-browser. Therefore, I installed it once I had gathered all the performance-related information. In my past ‘Linux’ reviews, I used to manually enable Adobe Flash on popular websites such as YouTube to test its performance. But nowadays, almost all popular websites have adopted HTML5 standards. Therefore, I no longer bother to test Adobe Flash performance. That being said, both on my Intel & Nvidia mobile GPU drivers, I tested the HTML5 video performance of Google Chrome on a few video sharing web-sites (I have been using Linux Mint 19.2 Cinnamon for about 3 days now) and so far, it has worked flawlessly.

As for the multimedia codecs, Mint does not come with the proprietary multimedia codecs by default but it lets you install them through the Installer as long as you have an active internet connection. Or you can install them later by using the ‘Install Multimedia Codecs’ utility as well.

Skype and Teamviewer

I did not run extensive tests on either of these utilities, other than making a very short video call on Skype & remotely connecting to a computer through Teamviewer. They worked without issues on both those occasions. However, I could not send Skype to the system tray whenever I closed its window, for some reason (Skype is known to have these types of issues on Cinnamon anyway). And also, every time for opening it, I had to open Skype twice to see its application window (I’m pretty sure this too is due to its system tray issue).

On a side note, so far I have installed three major proprietary software in Linux Mint 19.2 Cinnamon (Google Chrome, Skype & Teamviewer) without any issues whatsoever as well.

Performance comparison

As mentioned in the beginning, I take careful measures to the best of my ability to make sure the data I gathered is as ‘pristine’ as possible. That’s why I securely deleted all the partitions into which I installed Linux Mint 19.2 (‘/EFI’, ‘/Boot’ & ‘/’). Then, I disabled a few startup apps and added the system monitor app icon to the taskbar (I do this to keep the memory usage readings accurate. Otherwise, if I used the start menu to open it up, it would’ve taken more memory -opening the start menu means loading additional data into the RAM- therefore reducing the accuracy of the memory usage readings) for getting the memory usage data.

Then, I boot into the system a few times to let things settle down. Only then did I start to gather the performance data.

Boot-Up Delay

I used my Android smartphone to count the seconds it takes for the system to boot. I started the measurement as soon as I hit Enter at the GRUB screen and stopped the moment the desktop was fully loaded. Below is the graph I created based on the data.

As you can see, Linux Mint 19.2 Cinnamon is the clear winner here. Even though it used more memory (which you’ll see in a minute) compared to Kubuntu 19.04, it was still able to boot 19 seconds (35% roughly) faster. The reason is pretty simple. Linux Mint includes the ‘ureadahead’ daemon which as mentioned in the beginning, can dramatically speed things up on rotational disks, but Kubuntu 19.04 does not ship with it.

To prove my point, I disabled this daemon to see by how much it would negatively affect the boot-up speed. Below is the graph that shows, on both instances Linux Mint 19.2, with and without the ‘ureadahead’ daemon.

In summary, without its aid, on the rotational disk, the system booted 39.5% (14.3 seconds) slower!

Memory Usage

Once the desktop is fully loaded, I waited for about 15-20 seconds before opening the system monitor app. Once it’s opened, I gave it another 5 seconds before taking note of the memory usage readings. Below is the graph that I came up with.

Kubuntu 19.04 is the clear winner here as Linux Mint 19.3 Cinnamon consumed about 112% more memory. KDE these days is impressively lean on memory and in my experience, this 700 MiB something memory consumption is quite usual for GNOME 3 based desktop environments.

CPU Usage (idle)

When letting to idle, Linux Mint 19.2 Cinnamon did not bother the CPU at all. System monitor consumed about 1% here & there, other than that, nothing major happened. I do not have a screenshot for Kubuntu 19.04 (truth be told, I gathered these performance data as soon as Kubuntu 19.04 was initially released a few months back. I was originally planning on writing a review, but somehow that didn’t take place. Fortunately, I had the performance data written on a notebook but the screenshots I somewhat recently deleted from the disk) but it delivers pretty much the same amount of efficiency compared to LM 19.2 Cinnamon (I know this because Kubuntu 19.04 is my main operating system).

Power Usage (idle)

As far as power usage at idle is concerned (while measuring power, I disabled any screensavers, screen dimming and prevented it from turning off. I also had my mouse attached to the laptop on both operating systems), Kubuntu 19.04, out of the box, was about 20% energy efficient compared to Linux Mint 19.2 Cinnamon. That’s a big number as far as energy consumption is concerned.

But then again, if you’re not happy with your ‘Linux’ operating system’s power efficiency, these days, you have a few options. One such option is to install the awesome ‘TLP’ power usage optimizer. Almost all popular ‘Linux’ distributions host it on their servers, therefore, it’s very easy to install. I installed it on both Kubuntu 19.04 and LM 19.2 Cinnamon and it was able to reduce about 0.6 Watts on Kubuntu and 2.1 Watts on Linux Mint 19.2 Cinnamon. Thus in the end, once TLP was installed on both OSs, both performed almost equally efficiently. Below is the screenshot I took after measuring power (to measure power, I use the amazing ‘powerstat‘ utility) usage on LM 19.2, after installing TLP.

Hardware Recognition and Stability

Linux Mint 19.2 Cinnamon, just like Kubuntu 19.04, recognized all of my hardware correctly and offered me to install the proprietary driver for my Nvidia 920M mobile GPU (which I did later on).

As mentioned earlier, I have been using it for 3 days now (I’m writing this review on it), I have switched from Nvidia to Intel (back and forth), so far everything has worked without any issues.

As far as the OS’s stability is concerned, I have not yet seen a single crash so far either.

Bug? Not exactly…

This is not exactly a bug. But I’m not that happy with the battery indicator on the notification area. Take the below screenshot for instance. That’s my battery at its 63% charge, but those pixels indicate or at least give the user the impression, of a much higher value for its remaining charge. The issue is with the number of pixels used in each ‘bar’. If I had not enabled showing the remaining percentage in number, this would’ve been a very misleading picture.

KDE, on the other hand, uses a much smaller ‘bar’ scale, thus is more accurate in its presentation. This may not be a huge deal, but then again, the battery charge indicator is kind of a big thing nowadays.

System Responsiveness

To get a sense of the system’s responsiveness, I run a very simple test. I copy a file (about 2-2.5 GB in size) between two locations within the ‘Home’ folder. And as soon as I execute the file copy job, I start the web browser (Firefox on this instance because back then I had not installed Chrome), try to play a video file, open the software manager, LibreOffice Calc, GIMP (since it came pre-installed) and the terminal (to see if I can run a command or two). If I have the time, then I also try to open up a location that includes a reasonable amount of files & folders (such ‘/usr/bin’) in the file manager, to put additional pressure on the system. If the OS ables to open most of my apps, play the video file with the minimum amount of interruptions, all the while not losing the sensitivity of the mouse pointer that much, then I consider it to be a responsive system, under stress.

This is only an illustration…

How did Linux Mint 19.2 Cinnamon perform then?

Well, on the first try, the apps and the video was only opened with 4-5 seconds of delay. After, the OS was able to keep up with the load. But after those initial 4-5 seconds, it had finished copying most of the data of the file, therefore, I don’t think it was under heavy stress afterward. In any case, Kubuntu 19.04 in this regard was far superior. Not only was it able to open the apps and play the video from the very beginning, but it also carried it all out with very little interruptions, if none at all. If I had to assign it a value, I would’ve given it (Kubuntu 19.04) 9/10 stars (according to my notebook). Linux Mint was nowhere near that performance. But I didn’t give up.

So I changed the I/O scheduler (I/O scheduler is the utility that controls input/output operations on storage volumes) from ‘CFQ’ to ‘deadline’, rebooted the computer and ran the test again. And this time things were massively improved. Not only the video playback and the app opening started to take place at the very beginning (and most of the apps were opened before the file copy job finished), but video playback was also carried out without any noticeable interruptions. Plus, the mouse pointer did not lose its sensitivity either. I wouldn’t say it was still as good as Kubuntu, but I’d give it 8.3/10, which is still a very good number.

Shutdown Delay

Here too Kubuntu 19.04 won by being 22.6% faster while shutting down. But 6.5 seconds of shutting-down time is more than acceptable in my experience.

Final Words

Linux Mint 19.2 Cinnamon still ships with ‘ureadahead’ daemon which as I’ve shown, has the capacity to significantly speed things up (again, it only works on rotational disks) when booting up, even though the memory usage is not as impressive compared to modern KDE desktops (that then again is not Cinnamon’s fault, because, that’s how it is with most GNOME 3 based distributions these days).

If you’re not happy with the power usage, then install TLP, if the responsiveness is not impressive, try changing the I/O scheduler, and you’re good to go. It’s easy to use, quite stable and even though Cinnamon’s apps are forked from GNOME 3’s apps, due to its will to retain its own identity, nowadays, to a great extent, Cinnamon is its own thing. If interested in, download it from Mint’s official web page. Good luck and thank you for reading.

14 thoughts on “Linux Mint 19.2 Cinnamon vs Kubuntu 19.04 Review”

  1. Excellent review with actual subjective data on minutia, as we’ve come to expect from HecticGeek. Thanks!

    The case of Ubuntu dropping the `ureadahead` daemon is a classic example of how Linux distributors these days tend to not care much about their users. Have an older 32-bit processor? “Tough luck.” Have a spinning disk? “Not my problem, go buy yourself an SSD.” Excessive RAM consumption? “Get more RAM, all the cool people have 32GB+ now.” It’s obvious how the system is designed to perform perfectly under the hardware and workloads that developers run, which are usually orders of magnitude more powerful than what you find on the average consumer’s desktop.

    I’m not a Linux Mint user, but I greatly appreciate their focus on “real” users, and I am extremely thankful that they maintain Cinnamon and promote Mate, both of which maintain a level of sanity amidst the constant churn of “change for the sake of change” in the IT world.

    Reply
  2. You’re right Sam!
    Right now Plasma, in terms of DEs and Cinnamon are the bastions of sanity.
    The transition of Mate and Xfce recently to GTK3 are also great news.

    The Gnome DE got some minor usability improvements and fixes for memory leaks since it got back to being used by Ubuntu.

    I just wish the Mint team would again release a Plasma version of their distro. It would join the best of both worlds. I find Kubuntu lacking in certain aspects that Mint usually polishes.

    Reply
    • That’s very misleading. Gnome System Monitor uses the GTK UI toolkit (among other things), therefore, in order to open it, KDE which uses the QT toolkit, has to load GTK and other dependencies to RAM. That, of course, is going to increase the memory usage reading unnecessarily.

      Reply
      • It’s not misleading. You can compare with neofecth. KSysGuard can’t show cache or buffer. This was the case in older versions of the Gnome system monitor too. Just like on xfce4-taskmanager.

        Reply
          • I know that’s mean, but your comparison is incorrect. Basicly you can run “free -m” command in both distro and and see the result.

            According to this, You make an incorrect RAM usage comparison.

          • I don’t understand the question. First, you said that I should install the GNOME system monitor to monitor memory usage in KDE. But that obviously is going to quite unnecessarily (thus negatively affect the calculation) increase the memory usage in KDE since KDE has to load GTK libraries (etc) to memory, before loading the GNOME system monitor (I have not seen memory usages that go above 1GB in freshly installed KDE desktops -it could, however, increase the memory usage reading under certain circumstances, such as depending on the ‘weight’ of your GPU driver, for instance- in a long time, and in all those occasions, I have only used KDE’s in-built system monitor to monitor the data for the very reason I mentioned earlier).

            Secondly, ‘cache’ or ‘buffer’ has nothing to do with the type of memory usage readings I’m interested in.

            Thirdly, both ‘free’ and almost all desktop-based ‘system monitors’, gather memory-related data by parsing ‘/proc/meminfo’. Sure there could be a few mebibytes indifference (since the ‘system monitor’ GUI app itself consumes a few mebibytes, whether it is GNOME or KDE matters not) if its data is compared to the readings of the ‘free’ command. But that is an ‘error’ I’m willing to accept.

            That being said, there can be occasions that force one to switch to command-line tools for measuring the memory usage to get an accurate reading (for instance, when the memory usage goes beyond 1GB, the current KDE memory usage monitor has a tendency to ’round-up’ the readings a bit aggressively once it switches to gigibytes to display the readings. In such instances, it is useful to switch over to the direct command-line tools because they give greater control over the units that will be used to display the readings, and as long as the user force them to deliver the data in mebibytes, for instance, you can get very accurate readings).

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: