Slackware 7, in the second half of 1999, shipped with KDE1. For KDE’s part, KDE 1 was released in July of 1998.
The UNIX landscape at this time was still largely proprietary, and CDE was dominant. Linux is, and has been from day one, a mix of different software projects. Every Linux distribution is like a Katamari Damacy of open source software. This shows in the way the Linux desktop looked in the late 90s. This didn’t sit too well with Matthias Ettrich, and thus he started KDE. In true Linux fashion, he announced it on usenet.
Given the problem that Matthias wanted to solve, KDE is very much unlike other desktops. It aims to provide a suite of visually coherent software. To this end, when KDE2 released in 2000 it included the Konqueror web browser which used its own rendering engine KHTML. This engine became WebKit which powers Safari, and a fork of WebKit is Blink which powers pretty much every other browser (except for Firefox and Seamonkey). KDE2 also included an office suite, KOffice, which had word processing, spreadsheet, vector drawing, and presentation software. It included a PIM suite. It included an email program. It included an IDE. The impressive part was that all of this software had a unified look and feel, and worked well together as a cohesive software system. In the early 2000s, this was huge. It wasn’t until the mid 90s that a “suite” of software was even a thing, as captured by Steven Sinofsky's Hardcore Software blog, and here was KDE a few years later where the entire desktop was a single suite of software.
For my own part, my first attempt at running Linux went quite well. There were always computers around in my youth, and I was a very curious kid. Slackware would find itself installed on everything I could get it to run on, and KDE was amazing to me. The first release of KDE turned Linux into something I actually used rather than just explored. KDE2 made Linux something that could be used by almost anyone (provided that someone technically savvy could set it up for him/her). Then came KDE3 on April 3 of 2002. This release had a printing system, LDAP integration, Microsoft Exchange compatibility, and some new applications. Throughout the KDE3 lifetime, the suite of software would grow and grow and become more and more refined. KDE 3.5 was forked to become TDE in 2010.
Why does this matter?
For those younger readers out there, every single time any environment gets significantly changed, there is a lot of acrimony and opprobrium fired off at the developers. In this case, the introduction of KDE4 was reviled by a large part of the community. It was, in fact, so hated that GNOME started to take away its market share. In 2011, GNOME 3 was released… and people hated it too. This made XFCE become more popular. XFCE has seen an increase in weight, and now a significant portion of the Linux community uses i3 or sway. Effectively, KDE4 caused the Linux desktop to fracture significantly.
The “traditional” desktop means vaguely Windows 95 and/or CDE. KDE was, IMHO, the pinnacle of the traditional desktop design. You have traditional menus, a system tray, notifications that aren’t so infuriating you want to murder your machine, a pager for virtual desktops, a task bar, and desktop icons.
There’s no need to completely leave your current application to an entirely separate screen, hit 5 buttons, and then have yet another useless screen transition just to launch another application. You just press the menu, slide your mouse through the menu, and then click an application. It opens.
People didn’t just prefer the “traditional desktop” out of habit. Habit was, for most computer users, MS-DOS or Windows 3. These are both great systems. I don’t really have anything negative to say about them. Windows 95 became a smash hit because it was simply a better system. It was intuitive enough that anyone could figure it out in a few minutes just sitting down with it. The addition of a task bar to show open windows that isn’t merely an icon and doesn’t smash everything into a single button was a good thing. The existence of the system tray for specific types of controls was a good thing. The existence of menus that didn’t attempt to “be smart” or categorize things in non-obvious-to-a-novice ways, was a good thing. The fact that it didn’t constantly bombard you with notifications was a good thing.
Experimentation is great, but all of the things that made Windows 95, KDE 1, KDE 2, KDE 3, GNOME 1, and GNOME 2 great … those things are still there. For some people, that older paradigm is the correct choice. KDE 3 lives on as TDE and that’s a good thing.
If you’re wondering why the screenshots of my desktop are 4:3… that cuz my 1080p monitor died, so I am using an HP L1710. Monitor prices are stupidly high, like the cost of everything is lately, and this works well enough for me at present. Plus, it adds to the wonderful “feel” of running KDE 3. The rest of the system is fairly modern. This Slackware install is not in a VM. It is running on some bare metal. Additionally, I purposefully left out KDE5, XFCE, and Emacs when doing the installation. The TDE install was accomplished via TDE Slackbuilds.
I have made the compiled packages available here.
First, 4:3 looks good. There’s a reason I prefer the 9:10 ratio in modern machines. 16:9 is great for watching movies and playing games but the older format was better for computing IMHO.
Also, when I got into Linux I used Gnome 2. By the time I got to try KDE they were on version 4 and I absolutely hated it. I should try the earlier versions as I’ve heard good things about KDE 3, and it looks quite nice. I also like the look of 2 in the screenshots but if I’m honest I really dig the whole vibe of 1.0.