Used it remotely before I ever installed it. Lots of vulnerable 1.3.x boxes floating around on Efnet/Dalnet/Undernet back in the mid-90's, with public IP addresses, in unlocked channels.. For a teenager with a lot of free time, it was a wonderland. 1st distro I installed at home was Slack 3.x, just the A, D, and N floppy sets. Got it dual booting with Win 3.1 on a 486 DX2/33MHz with 8MB of RAM. Kernel was 2.0.25, I was told on IRC that recompiling it and taking out all the drivers you don't use would speed it up. And on a 33MHz machine, it was true! Compiled 2.0.29 and got LILO to boot it, with the distribution kernel as a fall-back just in case. I removed the windows partition the next year and for the rest of the 90's and all of last decade I ran Linux on the desktop. WindowMaker was my WM of choice back then (a NeXTStep clone interface). I had a separate machine for gaming, but if Blizzard had released the WoW Linux client to the public, I probably wouldn't have. Tried it in Wine many times, and I hear it's actually pretty stable now, but that's all in the past for me. If Linux is ever going to 'win', it will be because OEMs like Dell/HP/Lenovo pre-install it. People getting a computer for the 1st time have no idea what an OS is, they use what they have been given. If we had a "Universal App Store" that was OS-agnostic (and populated with the software we actually use to run modern businesses) then Linux's sticker price might convince a few OEMs to jump ship. The "Microsoft Tax" is tolerated because that is where the majority of useful business software exists, tied to legacy applications in that environment. These days I work in a heterogenous environment with many different OSs, Linux is there as a firewall and a web server, as the kernel behind phones/tablets we use, and on the laptops we run Kali on for pentesting. We use the OS that matches the function, and I'm typing this on an iMac from 2008 because it was a free donation and it has a larger screen, lol. There will be so many OSs after Linux (ESXi was Linux before VMware optimized it as a hypervisor). We put so much time and so much code into it, why not start there to fork a new tree and try a new direction? I see Linux in 20 years as the great-grand-father of the OSs we'll all be arguing about then, just like UNIX/POSIX is now.