Press "Enter" to skip to content

Linux turns 30: Success factors then and now

Congratulations on your 30th Linux! You saw the light of day on Friday thirty years ago, after Linus Torvalds had previously announced that you would appear on August 25th. You have retained some rough edges from your early days to this day. Don’t worry if someone holds them against you: critics often fail to realize that some of them are the reason for your triumph. On the anniversary, it is therefore a good idea to take a closer look at some of your characteristic properties.

The main factor for the success of the Linux operating system kernel is clearly the many helpers. By constantly improving Linux a little bit for their respective purposes and continuously improving it, it was able to gradually outperform its competitors and ultimately become so good that even Microsoft uses the Torvalds kernel for a few things.

Linus Torvalds, who was only 22 years old at the time, quickly attracted the first comrades-in-arms after he released Linux 0.0.1 on September 17, 1991. The brisk influx was mainly due to the fertile soil on which Linux fell: apart from the deliberately limited Minix teaching system, there was nothing comparable for Intel’s 386 processors, which were then popular.

Just like 30 years ago, Linus Torvalds still manages to attract many people to help.

Torvalds wouldn’t even have started Linuxif it’s the FreeBSD and NetBSD predecessor 386BSD would have already given. However, it didn’t appear until the spring of 1992 – about a year after Torvalds started using Linux.

The software world of today could therefore easily have looked very different. The question, however, is whether a kernel derived from BSD would actually have had the resounding success of Linux. Because while BSD offshoots mostly use a liberal license, Torvalds chose one that requires all modifications to be made in the case of further distribution.

Torvalds, however, made a misstep when he did Formulated license conditions for the first version. Although they were scarce, they still contained something that would have denied them the label “Open Source Software”.

Specifically, it was a clause that forbade taking fees for the redistribution of the code, as some service providers did at the time. Such a restriction, however, contradicts the basic conditions of the Definition of “open source”, which only arose years later in the wake of Linux’s success.

A clause in the license conditions initially used by Linux contradicts the basic idea of ​​open source.

(Image: screenshot)

Software that breaks such a hurdle finds people who are difficult to help today; the commercial environment in particular is holding back. Back then, the misstep was less frightening. Nevertheless, the first colleagues soon advised to switch to a more sophisticated license – namely version 2 of the GNU Public License (GPL).

The GPLv2 already knew the scene around free Unixoid operating systems from the many tools of the GNU project that were necessary to build an operating system with the Linux kernel – and which are still important today, even if their importance is waning. Torvalds agreed with the idea. With Linux 0.12 he announced the license change to GPLv2, which he carried out at the end of 1992 with version 0.99.

However, Torvald’s first license conditions already had one thing in common with the GPLv2, which is still used today: They obliged everyone to provide the source code used for construction when selling the operating system kernel. That was helpful in the early days, but not terribly crucial either. Because anyone who worked on Linux in order to make it better for his requirements, often transmitted these changes out of idealism to Torvalds, so that he could integrate them into new versions of his operating system kernel.

The aspect became more important in the following years: In the further course of the nineties, more and more companies began to earn money with products that contained Linux. Many of these corporations saw no point in submitting the changes with which they had drilled out Linux for its purposes to Torvalds for integration. No wonder, because Torvalds and his colleagues often demand improvements in the quality control carried out, which work causes the submitters. In addition, companies see competitive advantages at risk in “upstreaming”, because after all, competitors receive their in-house modifications with new Linux versions free of charge.

According to a study by the Linux Foundation, almost 84 percent of the changes made to the kernel from 2007 to the end of 2019 came from developers who, according to their own statements, are paid for their work on Linux. Employees from Intel and Red Hat make a particularly big contribution, but they too have to admit defeat to amateur developers.

(Image: heise.de, with raw data from the Linux Kernel History Report 2020)

This kind of thinking can still be found in some companies today – especially those who are just getting used to Linux or who operate in markets with small profit margins. However, the obligation to provide the source code for further distribution still helps the operating system kernel: Anyone with enough motivation can process the code and incorporate it into Linux.

This is exactly what hobby developers or programmers from other companies do from time to time. This often involves significant work and is therefore not entirely commonplace, but it happens often enough to give Linux an edge. It is mainly thanks to independent developers that Torvalds’ kernel supports many Raspberry Pi out of the box these days, because for a long time the foundation behind the microcomputers did not give a damn about integrating their modifications into Linux.

Article Source

Disclaimer: This article is generated from the feed and not edited by our team.