Slashdot Mirror
OpenSource.com Test-Drives Linux Distros From 1993 To 2003 (opensource.com)
An anonymous reader quotes OpenSource.com:
A unique trait of open source is that it's never truly EOL (End of Life). The disc images mostly remain online, and their licenses don't expire, so going back and installing an old version of Linux in a virtual machine and getting a precise picture of what progress Linux has made over the years is relatively simple... Whether you're new to Linux, or whether you're such an old hand that most of these screenshots have been more biographical than historical, it's good to be able to look back at how one of the largest open source projects in the world has developed. More importantly, it's exciting to think of where Linux is headed and how we can all be a part of that, starting now, and for years to come.
The article looks at seven distros -- Slackware 1.01 (1993), Debian 0.91 (1994), Jurix/S.u.S.E. (1996), SUSE 5.1 (1998), Red Hat 6.0 (1999), Mandrake 8.0 (2001), and Fedora 1 (2003). Click through for some of the highlights.
The article looks at seven distros -- Slackware 1.01 (1993), Debian 0.91 (1994), Jurix/S.u.S.E. (1996), SUSE 5.1 (1998), Red Hat 6.0 (1999), Mandrake 8.0 (2001), and Fedora 1 (2003). Click through for some of the highlights.
- Slackware 1.01 (1993). "The best part about trying Slackware 1.01 is that there's a pre-made image in Qemu's 2014 series of free images, so you don't have to perform the install manually... In more ways than I'd expected, the system feels surprisingly modern. What's missing is any notion of package management. All installs and uninstalls are entirely manual, with no tracking."
- Debian 0.91 (1994). "To try Debian 0.91, I used the floppy disk images available on the Ibiblio digital archive... The install process is surprisingly smooth... The dpkg command exists, but it's an interactive menu-based system... Even so, you can sense the convenience factor in the design concept... I sincerely see why Debian made a splash."
- Jurix/S.u.S.E. (1996). "Because I wasn't specifically looking for the earliest instance, Jurix was the first Linux distribution I found that really 'felt' like it intended the user to use a GUI environment. XFree86 is installed by default, so if you didn't intend to use it, you had to opt out."
- SUSE 5.1 (1998). "I installed SUSE 5.1 from a InfoMagic CD-ROM purchased from a software store in Maryland in 1998. The install process was convoluted compared to those that came before... Included desktops were fvwm, fvwm2, and ctwm. I used fvwm, and it worked as expected. I even discovered tkDesk, a dock and file manager combo pack that is surprisingly similar to Ubuntu's Unity launcher bar."
- Red Hat 6.0 (1999). "The disc I used was purchased in June 1999. The installation was fully guided and remarkably fast... The desktop bundled with Red Hat 6 was, as it still is, GNOME, but the window manager was an early Enlightenment, which also provided the main sound daemon... Unlike later implementations of GNOME, this early version featured a panel at the bottom of the screen, with an application menu and launcher icons and virtual desktop control in a central location."
- Mandrake 8.0 (2001). "Mandrake 8.0 was released in 2001, so it would have been compared to, for instance, Apple OS 9.2 and Windows ME... I'd thought the Red Hat installation process had been nice, but Mandrake's was amazing. It was friendly, it gave the user a chance to test configurations before continuing, it was easy and fast, and it worked almost like magic..."
- Fedora 1 (2003). "The Fedora Core experience is largely indistinguishable from Red Hat 6 or 7. The GNOME desktop is polished, there are all the signature configuration helper applications, and the presentation is clean and professional... A red hat icon marks the applications menu, and the lower GNOME panel holds all the latest Linux application launchers, including the OpenOffice office suite and the Mozilla browser."
This submission makes us ask an interesting question: what impact will systemd have on the historical preservation of today's software?
A lot of people have reported problems with systemd. A look through the Fedora, Debian or Ubuntu bug trackers and mailing lists will give some idea about the sorts of problems that are encountered. Many of them are quite strange, and some of them are downright stupid.
If we're running into these problems today, under ideal circumstances, I can't help but think the situation would be much worse in the future.
Will future historians trying to run Linux actually be able to do so with any ease? There's no guarantee that they'll have access to the same mailing lists, bug trackers, and people in IRC channels that we can use today to try to solve bugs with systemd, just so our computers boot.
The article makes it sound like it was challenging enough after only a decade or two, and we're talking about Linux distros from the era when Linux was much simpler, and typically just worked.
I suspect that there may need to be historical research specialists in the future who focus solely on debugging problems with systemd. That's not a job I would want to have! It's bad enough dealing with these problems now, when there are at least some support avenues available. Trying to do it in the future, perhaps without access to the resources we have today, would be a miserable experience.
Perhaps in the future we'll just see what we're seeing today: a decrease in interest when it comes to Linux, and an increase in interest in the BSDs, which avoid systemd and which just tend to work fine without needing to be hand-held and debugged frequently. Historians will study FreeBSD, rather than Linux, because FreeBSD will just work in the simulators they're using to do their research. They'll be able to focus on the historical research they want to do, rather than trying to debug pointless problems with systemd.
Bought the 4 disc Mandrake 8.0 in 2001. It came with an excellent book explaining the basics. Had been running Red Hat 6.1 for about a year and the Mandrake felt much more polished. As I recall at the time I had a Pentium II with 32 mb of RAM and it would run KDE, but worked much better with Fluxbox. A big deal at that time was to play a DVD, which it did with some help of packages from the Penguin Liberation Front. Got me on the internet as well, setting up PPPOE was stressful. Doesn't seem like 16 years ago. Have been running Linux as my main desktop ever since, no complaints.
Worked on ASIC development for F-22 back in the 90's. Management had stupid idea of putting us all in a big room where we had to share Sun IPX's to work on ASIC design. I got permission to build linux PCs so we could work from our desks. I used slackware for X windows terminals. In true defense contractor fashion the PC people bought us the SVGA graphics cards I asked for, but VGA monitors. Luckily I was able to return them and get state of the art 17" 1024x768 ones :)
This wasn't until a few years after 2003, but after trying to install Mandrake Whatever and Red Hat 9.0, I remember installing Debian Sarge. It wasn't magic, but I remember thinking, "So this is what an easy install feel like." No crashes or anything, it just installed easily, 1 step at a time.
I had a somewhat different experience. I started with RedHat 8.0 around 2001, since the school Linux cluster on which I was doing a class assignment ran RedHat (6.x or 7.x). A friend from class (who also worked tech support in the CS department) told me that I would be better off with Debian but that I would probably need some help to get through the installation. That was Debian Woody. I remember thinking that I would not have made it through the install without the hand holding.
That said, after some months of using RedHat, the Debian way just made so much more sense to me. It seemed far more natural than the way RedHat approached so many things. I liked so much that I start getting involved, contributing bug reports, writing documentation, and after a couple of years becoming a Debian developer. I think Debian has come such a long way that when I look back I am surprised at how far we have come, both in the sense on how Debian has been able to build on the work of others and how Debian itself has formed an incredible platform from which so many derivatives have sprung forth.