The case for LINUX

The case for LINUX

When you go back to the situation that you had in the mid to late 60s, computers were invariably expensive and large and operating systems for them were written in assembler language. Selling computers was a game which only large corporations could play. Aside from the time and effort which it took to develop a new piece of computer hardware, there was the huge effort to develop corresponding operating systems, programming languages, and other software tools required to make the new piece of hardware usable. The advent of the minicomputer did not fix the situation, that only brought the price down from millions to the hundred thousands range. The effort to develop operating systems (such as VA X/VMS) and software for the minicomputers remained beyond the reach of small companies.

With the advent of microprocessors such as the Motorola 68000 in the late 70s and early 80s, the situation was aggravated. It seems clear enough that the cost and effort to develop new generations of hardware would go down substantially and that in theory at least it should become possible for small to mid-sized companies to market computers, but you still have the same problem with operating systems and software. In other words, a company might very quickly develop a new piece of harbor but it would still take a year or more to develop operating systems and software for such computers. That said that a company might spend some long period of time to develop a new computer along with software for it, only to have a much shorter period of time in which to market that computer before being leapfrogged by others who simply started later. That still made marketing computers a very expensive proposition.

At that point the world was forced to sit up and take notice of the C programming language and the UNIX operating system (written in C) which had been developed at Bell laboratories. C was a revolution in programming languages, a highly structured low-level language which could do everything which you use assembler language for while simultaneously allowing the creation of logically complex code. An operating system written in C meant that other code could directly address devices in a high-level language and essentially have access to the entire functionality of the operating environment.

And the clincher was that for the new piece of hardware you were developing, all you had to do was develop a C compiler and, basically, with just a minimal bit of other coding, you had UNIX up on your new piece of hardware along with all of the software which needed to market it, problem solved.

Thus by the early 80s there were two or three hundred bright new small companies in Silicon Valley preparing to take the world by storm with microcomputers running various flavors of UNIX. Those companies were largely killed off by the PC revolution and the associated advertising blitz which began around that same time, at least in the case of the microcomputer world. UNIX did however begin to take off in the minicomputer and mainframe worlds at about that time.

Thus by the mid-90s there was a proliferation of proprietary UNIX systems being sold for servers and graphics workstations of various sorts. This presented a problem of its own. There was no sufficiently large user base for any one of those proprietary UNIX systems to guarantee that bugs got killed off quickly when a new version of UNIX was released for any of those computers. What was needed was a standard version of UNIX for all manufacturers. That need was met by a Linux, which had started out as a hobby shop project.

My own first look at Linux came around 1995. I was working at the Aberdeen Proving Ground’s using Sun workstations and I installed a copy of Linux on a 486 based PC at home to see how close I could come to replicating my environment at work. The shocking answer was that I could come extremely close. I kept expecting to see some kind of a flag or warning come up saying “hey, this is just a toy version of UNIX, you can’t do that sort of thing on a toy operating system”, but it never happened. Generally, things which were graphics intensive were faster on the Sun workstation while things which were disk intensive were actually faster on the 486 Linux system. The reason was that the 486 computer was new while workstations which cost $50,000 tend to be used for several years before being replaced, meaning that the disc technology on the 486 was several years ahead of that of the workstations at the office.

Meanwhile, as I’ve mentioned, Peppermint Linux is now in a position to compete successfully against all Windows/Microsoft operating systems. Microsoft, which set American computer science back years if not decades when they shut OS2 down in the early 90s, is now powerless to stop the Linux revolution.

The thing which prevented ordinary users from wanting to use Linux originally was the extreme difficulty of installing any particular piece of software on a Linux system. You generally had to find and install every other piece of software which the item you wanted required in a chain of dependencies which cascaded downwards. All of that is history now. Ubuntu and mint flavors of Linux come with a software manager and a synaptic package manager which provides point-and-click interfaces to easily find and install items from a gigantic software world, all of which are free. Aside from Open office software which is filesystem compatible with corresponding Microsoft office products, there are native versions of Google Earth, Skype, and other common packages which are equally easy to install.