History of software forks, the successes

Many derivative software completely overshadow the original. But this is not necessarily always the case.

With the announcement of Devuan, a fork of Debian, the question should be asked, this version will it substitute for the original and become the basis for many other derived distributions such as Ubuntu, Mint, SteamOS, etc. or, conversely, will it get lost in limbo... and ridicule?

To find out, let's review the main forks of known software, and the cause of separation, to see if Devuan can follow a similar path...

SystemD de ce coté!

Devuan / Debian (2014)

Reason to fork: Refusal of SystemD by some developers and users.

Some developers complain that decisions like going under SystemD are taken by the community rather than a smaller and more involved circle, and that SystemD is unsuited to servers. They then created a fork just to stay under SysV init.

OpenBDD / NetBSD (1995)

Reason to fork: The leader of the development team had divergent views with the NetBSD development group and was asked to resign.

He then decided to distribute its own OpenBSD version designed especially for safety. He opened the source code for all, which was not common at the time. According to all sources, OpenBSD is now more widely used than NetBSD, but I have only found accurate statistics from 2005 (that shows it twice used).

WebKit / KHTML (2001)

Reason to fork: Apple wanted to make its own browser.

Apple chose KHTML, the browser of Linux and KDE, because the code was lighter. It then developed WekKit which was also the basis of Google's Chrome browser, before the latter in turn creates a fork, Blink.

Wordpress / B2 (2003)

Reason to fork: Handling larger websites.

B2/cafelog, one of the first blog manager in PHP and MySQL actually had two successors, WordPress and B2evolution. The author of B2 is now contributor to Wordpress which became the most used CMS in the world.

Joomla / Mambo (2005)

Reason to fork: The development team did not agree with the new Mambo distribution license.

Joomla means "together" in Arabic. Mambo developers have created this new CMS after a highly publicized controversy. Joomla is now one of the three most used CMS with Wordpress and Drupal. As for Mambo... the foundation that distributed it closed in 2013, and the name is almost forgotten, but the code is still available on Sourceforge.

FluxBB / PunBB (2008)

Reason to fork: After the sale of PunBB, a forum CMS, developers wanting to retain their freedom of decision and left the project to found their own.

The head of the team decided to leave the PunBB project due to weariness and other developers then did the same to create FluxBB. The latter became popular while the original is no longer maintained.

LibreOffice / OpenOffice.org (2010)

Reason to fork: Lack of development will by Oracle after acquiring it from Sun.

In 2010, noting that Oracle does not intend to develop the software, programmers decided to create a new version with the same source code. Oracle then asked everyone involved in this competing project to resign, and LibreOffice development team has grown. It became the main open source office suite, and remained it even after Oracle has transferred OpenOffice.org to the Apache Foundation.


All these forks were successful. This is not always the case, for example, there are many fork of Node.js, but none has replaced it to date. All respond to the desire to further develop the software, or make it free. The case of Devuan is unique, it is the first time that a fork is made to continue using an older tool. Therefore unless SystemD is a failure and that its problems are completely unsolvable, it seems that it will be the fork and not the original that will go to limbo...

Updated December 6, 2014 : Just after the publication of this article, I have been informed of a new fork of Node.js, Io.js! Since it involves the leading developers of Node and the reason is to accelerate the development, it has more chance of success than the previous attempts... Unless there is fusion of the two projects.

Updated May 13, 2015: An agreement was finally reached between the Io.js and Node.js teams to merge the two projects in the Node Foundation. This is a success for Io.js insofar as their implementation must be Node.js 3.0.