Slashdot Mirror
25 Years After DOS - Lessons for Linux?
E IS mC(Square) writes "Microsoft is planning to celebrate 25 years of DOS. An article at ReallyLinux discusses what lessons Linux can learn from the history of DOS. The article begins with 'What can the Linux world learn from Microsoft's past 25 years of unique experiences and domination?', and ends with 'Only question now is not if but when will Linux become the number one OS on earth?'" From the article: "First, we must admit openly once and for all that the 'best solution' is not always the 'most used solution.' There are few who would be foolish enough to argue that back in 1981 PC-DOS was the best solution. There were obviously a number of choices. PC-DOS was the least robust, the most temperamental, and arguably not very compatible with the IBM hardware and BIOS it was sold to work on. Yet, somewhat like the odd but obvious dominance of the VHS over BETA, this simple, cheap OS stole the show."
...cheap & simple will always win.
Only question now is not if but when will Linux become the number one OS on earth? How many times does this REALLY have to be asked? It's beginning to become cliche...
Real programmers can write assembly code in any language. -- Larry Wall
DOS must have been the best at SOMETHING, because it eventually prevailed in the end.
And, honestly, in terms of design, I don't think linux is that good. I don't think BSD is all that good either. Windows is awful too.
In my mind, Unix is overcomplicated and antiquated. Why can't we design a system with the needs of modern home computers in mind rather than mainframes in the 70s? Hell. We're still using the TTY acronym to describe a terminal (TTY=TeleTYpe machine). Unix is a definite case of too many cooks spoiling the broth.
Granted, steps are being taken in the right direction -- the abandoning of the traditional Unix Filesystem, apple's amazingly cool launchd, etc... but the fact remains is that it's all too complicated and obfuscated. The fact that we're complaining about the lack of a good packaging system points to a serious flaw in that Unix is the only type of system that NEEDS this sort of thing.
In the 80s, there was tons of diversity in the OS marketplace. There was DOS, MacOS, Unix, and all the other failed OSes. It looks like Unix won out simply because it already had an established market, and, hey... it's owned by ma' bell. The only successful newcomer to the OS scene in the past 15 years has been NT. How sad is THAT?
I really don't think we should be looking at linux, but, instead, looking at (creating) other operating systems that take advantage of all modern features (build the GUI into the underlying OS, a logical approach to hardware support, sensible APIs, etc.) while also maximizing simplicity. This CAN be done!
Just look at the number of failed OS pet projects. There are so many out there that never made it simply for lack of developer support in favor of the lurking hulk that Unix has become. Don't get me wrong, I think some wonderful things have been done with Linux/BSD and I commend those developers, but I think it is time to move on. It's a miracle that something as complicated as Linux/Unix even WORKS, let alone works WELL. The size and complexity of Unix is snowballing rapidly, and it's time for consolidation and simplification. The open source community has already proven that it has been able to keep the antiquated hulk of unix working well all these years, and it's time that it proves that it can create something even BETTER from scratch (and while we're at it, let's abandon X-windows and destroy the concept of a 'separate' GUI all together)
BeOS has been jumping out in my mind as I write this post as one such OS. It got lots of stuff right from the start. It was built upon sensible principles, supported as much hardware as the small development team could allow, and was FAST as hell. Perhaps more importantly, it was the only operating system that wasn't explicitly written to imitate another previous operating system (I'll acknowledge that MacOS was a big influence, but BE certainly wasn't a derivative work). It also wasn't written by a group of researchers, but by a business with a proper marketing department that knew what attributes could be changed to imporove the concept of the Operating System. Microsoft and apple have both done this to some degree of success, but, likewise, NT was built to be compatible with Win9x which was built to be compatible with DOS, and OSX was built to be compatible with Darwin/NextSTEP. The only real OS that was built from scratch by non-academics that achieved some degree of success was MacOS Classic which was revolutionary in its day.
Unless an open-source OS can be written from scratch and throw all the crap from the past out, we're not going to replace the Commercial OS market anytime soon. I for one do think that open-source will be important in the future, not as a political idea (it fits in nicely with communism and fascism), but simply because it has the potential to throw away the past and start from scratch and create something revolutionary. Linux was a me-too Operating system to imitate Minix/Unix. It will always be lurking in the past of the gigantic steel IBM mainframes. Open Source may be the future. Unix will not.
-- If you try to fail and succeed, which have you done? - Uli's moose
...Linux is quickly taking over the server market segment...
People who run servers are usually computer professionals or at least computer savy power users. Linux is an excellent system, made by geeks for geeks, but it is not likely to be taking over the desktop until the geekyness is sufficiently hidden from ordinary users. Mac OS 10.4 is just as powerful for server use as Linux, but Apple has also manged to hide the Unix geek stuff very well from the ordinary Joe or Jane user. The Mac also has some very easy to use software and very good hardware support for many neat gadgets and accessories. Windows is more frustrating because of all the malware, but still easy to use for the masses.
All theory is gray