From LQWiki

Timesharing came about in the first part of the elder days. (Actually, by the chronology in the History of Hacking article, the middle of the Iron Age - 1965.) The term means that multiple users use the same machine at the same time. It came about because computers became too fast for enough programs to be fed in batch mode to keep one busy 24/7. But at the same time, computers were still hugely expensive beasts, and needed hundreds of users to justify the expense. So large universities and businesses bought one or a few computers for their entire staff.

The advent of timesharing computers allowed the development of the first hacker cultures. Firstly, timesharing allowed programming students the ability to goof off and play around on a computer, without worrying about monopolizing a valuable scientific instrument. Hacking, after all, is based on the playful use of computers. Secondly, it allowed users to communicate through the medium of the computer. (Email was originally used for communication between users on the same computer, and only later adapted to networking.)

Timesharing was an early factor in the nocturnal nature of the stereotypical hacker. Since less users were on at night than during prime time, (and hence programs ran faster, and CPU cycles were cheaper) the night belonged to hackers.

Unix was inspired by earlier timesharing operating systems. (Although the name Unix itself came from a pun based on the fact that the first version only allowed two users at a time.) Linux carries this timesharing tradition in its very genes, in everything from multiple ttys to the existence of separate home directories.

Timesharing and Security

This timesharing heritage is one of the factors that makes Linux more secure than Windows. The 9x series of Windows (which includes 95, 98 and ME) comes from a line of development going back to DOS and Windows 3.11. The 9x series was designed as a single-user operating system. Networking was added in later, almost as an afterthought. Because of this, it is basically impossible to make a 9x operating system secure in any real sense of the word.

The NT line, which includes Windows XP, is a bit more secure, and was designed with multiple users in mind. However, the NT line is designed to be backwards compatible with the 9x line, in order to work with users and programs used to working with the older version. Because of this, it is possible, although very difficult, to make an NT operating system secure. (Although you still have to worry about programming defects.)

By contrast, Linux from before its birth, was designed with security in mind. Like in the example of email, there is much in common between timesharing and networking, including security.

Jargon File entry

Forgot they had this one. I'll have to integrate it into the main article. Crazyeddie

Timesharing is the technique of scheduling a computer's time so that they are shared across multiple tasks and multiple users, with each user having the illusion that his or her computation is going on continuously. John McCarthy, the inventor of LISP, first imagined this technique in the late 1950s. The first timesharing operating systems, BBN's "Little Hospital" and CTSS, were deployed in 1962-63. The early hacker culture of the 1960s and 1970s grew up around the first generation of relatively cheap timesharing computers, notably the DEC PDP-10, PDP-11, and VAX lines. But these were only cheap in a relative sense; though quite a bit less powerful than today's personal computers, they had to be shared by dozens or even hundreds of people each. The early hacker comunities nucleated around places where it was relatively easy to get access to a timesharing account.

Nowadays, communications bandwidth is usually the most important constraint on what you can do with your computer. Not so back then; timesharing machines were often loaded to capacity, and it was not uncommon for everyone's work to grind to a halt while the machine scheduler thrashed, trying to figure out what to do next. Early hacker slang was replete with terms like cycle crunch and cycle drought for describing the consequences of too few instructions-per-second spread among too many users. As GLS has noted, this sort of problem influenced the tendency of many hackers to work odd schedules.

One reason this is worth noting here is to make the point that the earliest hacker communities were physical, not distributed via networks; they consisted of hackers who shared a machine and therefore had to deal with many of the same problems with respect to it. A system crash could idle dozens of eager programmers, all sitting in the same terminal room and with little to do but talk with each other until normal operation resumed.

Timesharing moved from being the luxury of a few large universities runing semi-experimental operating systems to being more generally available about 1975-76. Hackers in search of more cycles and more control over their programming environment began to migrate off timesharing machines and onto what are now called workstations around 1983. It took another ten years, the development of powerful 32-bit personal micros, and the Great Internet Explosion before the migration was complete. It is no coincidence that the last stages of this migration coincided with the development of the first open-source operating systems.

See also

• History of Hacking