Linux Today: Linux News On Internet Time.

NewsForge: 2.4 kernel: Contributor Ts'o's practical approach to Linux

Feb 22, 2001, 00:15 (1 Talkback[s])
(Other stories by Julie Bresnick)

"Thirty-three year old Theodore Ts'o remembers when the Linux kernel was small enough to print out and read cover to cover, when "it was about an inch thick." That was around 1991, when he first joined the then scanty ranks of kernel hackers. He was a hobbyist, working as a lead developer of Kerberos and as an information technology architect for MIT during the day and working on Linux at night, until VA Linux hired him, in 1999, to be a principal engineer...."

"I first saw Linux in September of 1991. Linus had just released version 0.09. I thought it was a really cool thing. I saw it was missing a whole lot of features but figured I could add them, so I started work and made some contributions. The first thing I ever added was called POSIX job control. That's something that allows you to run a program in the background by putting an ampersand at the end of a command line. Previous to that it couldn't actually do that. I had the POSIX specification, which is sort of the industry standards for how Unix was supposed to be done, and the job control specification was an optional part of the spec that Linux didn't actually support at that point. So I just sort of sat down and looked at the specification and said, 'yeah, I can do that,' and just started coding. It was probably two weekends worth of work, more or less...."

"The most interesting thing about that period is that Linux really is a child of the Internet. Back in 1990, that was when the Internet really became reliable enough that it was easy to use it as a means for collaboration. Previous to that you had projects such as the X Consortium and the original BSD work but that had a very physical focus of activity. The X Consortium was located at MIT and the BSD work was based at UC Berkeley. There were a few people that contributed things over the network but primarily most of the development happened in one central place for each subsystem. The Internet back then was much slower and much less reliable."

Complete Story

Related Stories: