|< Free Open Study >|| |
Linux systems are extremely flexible. Linux (the kernel) can be scaled up to large multi-processor systems, or scaled down into tiny memory-constrained PDAs. Obviously, the software requirements for a distribution running on the multiprocessor system are going to be rather different that the software requirements for a PDA-unless, of course, someone's lucky enough to have a 32-processor PDA with 8GB of memory (that won't spontaneously burst into flames!).
So, the first decision that a distribution vendor has to make is what, exactly, are the goals of the distribution. Is the distribution to be for servers, or workstations? Is the goal a highly user-friendly environment suitable for desktop office use, or is the goal a high-performance traditional Unix environment for engineers? Is the distribution supposed to run on Intel, Alpha, PowerPC, SPARC, MIPS, ARM, or IBM S/390 mainframe?
This is, of course, a marketing decision (whether for a commercial entity or for a volunteer free effort), and different vendors target different markets. Red Hat Software, for example, produces a well-rounded distribution suitable for use as either a server or a desktop system. The Debian project produces a distribution similar in scope to Red Hat, but consisting entirely of free software. Mandrake Software produces a desktop-focused distribution, and TurboLinux produces a high-end distribution for clustering and parallel computing. The Slackware distribution aims to create a traditional Unix-like system. In other words, the market each vendor targets defines the overall requirements of the distribution, and the particular selection of software that goes into the distribution.
Each user, meanwhile, also has specific needs. Some users need to set up a workstation for their own personal use, others need to set up a local file server, and still others need to set up extensive server farms to run a web site. Some Linux distributions, of course, are good at some tasks, and others are better at other tasks. Sometimes, the user has the luxury of picking the best distribution for the task. Frequently, however, the user prefers to use a specific distribution (perhaps by personal preference or corporate mandate) that might not have been optimized out of the box for the task. In these cases, it helps the user immensely to understand how the distribution was built and what its goals are, in order to customize the system for the desired task. This book will help the reader understand and customize the system, but it's up to the reader to know what her own goals and needs are.
This book covers three sample distributions:
Red Hat Linux
These distributions were chosen because of their huge influence on the evolution of Linux systems; almost every other distribution out there is derived from, inspired by, or related to one of these three. Despite their different focuses, however, each of these distributions consist of the same general set of software; they're snapshots of the same continuum, just from different angles. The rest of Part Two (in Chapters 4 through 6) is dedicated to discussing each of these distributions in detail, but first you need to understand exactly what a distribution is; that's the focus of the remainder of this chapter.
|< Free Open Study >|| |