Assuming you were lucky and had only jumped from one good Microsoft Desktop OS to the next good Microsoft Desktop OS, I reckon your upgrade path would have gone:
DOS3.2->DOS6.22->Windows 95c->Windows 98 SE->Windows 2000->Windows 7
All the other steps like DOS 4.0, DOS 5, DOS 6, ME, Vista, 8 are all wild mis-steps which are best forgotten (in my humble opinion of course). But from this list you can chart a history of progress from the early command-line days, through a mixed-mode of command line and windowing system through to the GUI-first methodology of the post-2000 era. I’m not including Windows 10 cos that killed off all my laptops that ran Windows 7 just fine for zero added benefit, and Windows 11 is a flaming dumpster fire if you ask me
I feel that Linux has progressed in similar steps. If I was to write the same for Linux, I might put something like this (which is completely subjective), but I’ll explain below
Slackware 2 ->SuSE Linux->Red Hat->Debian->Ubuntu->Linux Mint->Kubuntu
The early Slackwares of the early 1990’s felt very much like DOS.. Primarily command-line, with a janky (X11) windowing system built on top of it. Most configuration was done by editing config files, things crashed all the time and software support was.. spotty.. SuSe was a great distro, and I loved it, but I spent more time recompiling things than actually getting work done. It was solid, and well documented and was great for coding on.
When Redhat appeared IBM took an interest… And suddenly we were propelled into “Linux: Phase 2″… We had properly backed distros with actual support. IBM, and later Canonical, put a lot of work into making the windowing system feel better, the configuration madness feel more controlled, and now hardware vendors started to take a little more notice and created actual drivers for the system. Mostly, the drivers were because Linux was being used a server, sharing things like modems and printers, so it was beneficial to do it for office environments. But the advent of the “SOHO” meant that anyone could have, say, an HP Laserjet on their desk.. and thanks to IBM and HP, they could now use their printer on Linux..
But even in the late 1990s, Linux was still a bit of a config hell. Recompiling kernels, failing dependencies, and cryptic error messages just turned people away in droves. But, so long as you stayed on the straight and narrow, email, web browsing, etc, Linux was pretty solid… I’d say by the late 1990s, Linux had entered its “Windows 95” phase. It was good, and was gaining traction, but was still very wobbly and prone to hard crashes and weird driver problems.
In my humble opinion, progress on Linux felt like it stalled from about 2000 to 2010. It was a period of consolidation. While it had made amazing progress, it seemed to disappear into being a “server OS”, andd the desktop side kind of went quiet. Other than “beardy-weirdies” like me who used it from time to time, it seemed that it had completely fallen out of favour with the general public.
While the number of distros exploded, the majority were seemingly just based on the same four or five base distros: Arch, Debian, Red Hat. Some offered more bleeding edge releases (with all the wobbliness that that entailed), while some were more stable. Plugging away in the background, Redhat, Debian, and the other main distros, just kept the candle burning, gradually and slowly improving while all these other little community-made distros based on their cores rose and fell.
If, like me, you were a debian user, the 2000s felt very solid. Debian just kept getting better and better. Support was growing for more devices, it was available for more things, and each release just felt more solid. I tried to ignore all the “upstart distros”.. They claimed too much in my opinion.
There didn’t feel like there were any major breakthroughs until Ubuntu became popular towards the end of the 2000s…. Ubuntu really did feel like the “Windows 98” era had begun. The UI was clunky and slow, but felt solid. It was clearly no contender for the “best UI” award, but it worked, and Canonical did some good work in standardising things. As a debian user I tried Ubuntu a few times, but bounced off the lack of software support and the godawful GNOME UI.
Since Canonical showed us that a Linux desktop wasn’t just possible, it was here already, it’s felt like we’ve been revving our engines, waiting for the take-off ever since. It just doesn’t feel like we’ve emerged from the history of needing to edit config files, of having to side-load random patches to fix little issues.. it’s getting less and less necessary, for sure, but we’re still not free of the command line.
It’s a shame because, unlike Windows 98 whose multitasking was janky and compartmentalisation woeful, we have a proper 64-bit pre-emptive multitasking core that’s secure and corporate-grade. We need to make that final break from the command line and having to hand-edit the config files in /etc to make changes to things. People are still scared to shift to Linux because they see people needing to work with the command line to fix issues, or search blogs. It’s just not quite idiot-proof.
Windows took the approach of moving all the config crap into the Registry and providing users with “regedit”. It was a halfway house, but I feel that’s what Linux needs to cut out a major percentage of the command-line wrangling that people may still need to do. Couple that with a proper Service management UI rather than “systemctl start…” on the command line, and I feel we’ll have emerged fully into the Windows 2000 era. Until then, for me, this feels like we’re stuck in the command-line Windows 98 era… Everything works… until it doesn’t!