OSNews

Let me start by saying that my Mac SE running OS 6.0.3 on my desk holds a special place in my heart. I like to boot it up now and then and write. It's bare bones, simple, uncluttered interface is a breath of fresh air compared to the jumble of today's UI's. It's elegant, simple and classic. It is, however, largely useless outside of games and MacWrite. Even if Apple open-sourced Mac OS version 6, it would still be as useless for doing day to day work. My MacBook is much better as a platform.

I realize my attachment to classic Mac OSes is nostalgic and not practical. What makes an open source project popular is its value to people. I don't mean to be flame bait, but I just don't see the value in running OS/2, especially since as much as it was advanced in some respects it was totally retarded in others. I used OS/2 in graduate school and thought it was okay. It was better than Windows 3.1. The best part was the SOM Objects library (which might be worth porting out) and VisualAge for C++. The presentation manager API, IMHO, not so great. I really can't comment about using Amiga OS, which I have no experience with (besides AROS). However, I doubt there is much new or interesting software for Amiga OS 4 as compared with Linux, Mac OS or (gasp) Windows outside of 'classic' games.

QNX is an interesting candidate because it does have practical value. It's competing against embedded Linux (it's largely an embedded OS). In that sense open sourcing it would make sense in the same way open sourcing Solaris made sense. It keeps the OS relevant in the face of Linux and makes it easier for developers to adopt it in their projects. Open sourcing for something like QNX makes a good bit of business sense. It's easy to find embedded x86 SOCs that are essentially shrunken x86 desktop systems. Developing and debugging on Linux on your desktop PC and just copying to your embedded target of choice makes development much easier. As opposed to having to run a cross-compiler, move the application to the target platform, and then start your debugging.

I think HP/UX and AIX would be interesting candidates to open source, however, I don't think there's much developer interest in yet one more open sourced Unix. OpenVMS would be an interesting open source operating system. First, it's not just another Unix. Second the clustering technology might have value if it found its way into Linux or a BSD variant. I think VM or z/OS might be interesting if it went open source. Again, it's not a Unix, it has fairly efficient virtualization. The downside is most IBM z/OS customers probably don't care if it's open source meaning IBM doesn't care if it's open sourced. I also doubt too many open source developers are going to be that interested in participating on an O/S that's essentially the big brown panel truck of operating systems. But for z/OS and OpenVMS the underlying O/S is only part of the picture. The real value in these systems comes in their transaction monitors, associated compilers, and application monitors. Without those they are little more than 'interesting.' There may also be significant issues in open sourcing these additional systems.

The other quest is what do you run on these systems? Amiga has some interesting games, but for the most part it would be ports of open source projects that exist on other operating systems. Let's say HP decides to Open Source OpenVMS and it gets ported to x86. As a hobbyist I might find that interesting. As a business owner, I'm more than happy running Apache on Solaris. I doubt I would change the underlying O/S just to run Apache. None of my clients would be interested in OpenVMS (the same people who edit a CSV report produced from the database and wonder why the database wasn't changed). In fact, with too many open source systems, there may not be enough interested developers to go around. Open sourced OpenVMS might still wither on the vine if it's one of 27 newly open sourced commercial products.

Going back to the original article. I think BeOS should be open sourced because the company that owns it is just sitting on the goods and letting them rot. As good human beings, release the code and maybe its value will show up in other projects. Amiga OS 4 might be able to get some market and mind share if it went to x86. Frankly, I think everyone involved in that cluster-f**k is trying to protect their potential profits as opposed to making actual profits. SkyOS - might make more progress if more than one guy worked on it. QNX is open sourcing because they understand it will be a Linux world in embedded space unless they stay relevant. Outside of that I'm not sure there's sufficient interest in open sourcing yet another Unix like HP/UX, Irix, AIX, Xenix, A/UX .... Maybe some non-Unix operating system (other than ReactOS) would be interesting, but they're pretty boring. (Okay, VMS has empire).