My view of OpenSolaris and Illumos

December 2, 2010

My perspective is that I have always had a two-fold interest in OpenSolaris. One part of my interest was actually in Solaris source code (for which OpenSolaris was the closest available substitute), and the other part was an alternative to Solaris if Sun went crazy.

I never truly trusted OpenSolaris for two reasons. First, because I never fully trusted the depth of Sun's commitment to open source; it always felt like a tactical move, instead of a heartfelt conviction, and tactical moves are subject to revocation. Second, OpenSolaris as a usable operating system (ie as an alternative to Solaris) always relied on Sun's charity, because it required closed source binary modules in order to do useful and vital things like NFS locking (and it required Sun compilers in order to build). If Sun had ever wanted to kill OpenSolaris as an operating system, all it ever had to do was stop making those components available.

In short, in effect OpenSolaris was a BSD Net/2 release that happened to run on some hardware. It was the starting point for an open source operating system, but it was not an open source operating system by itself. In practice it was little more than bleeding edge Solaris that you got most of the source code for (cf an earlier entry).

Of course, in theory all of this could change (or could have changed); given the existing source code, the community could fork OpenSolaris and turn it into a real open source OS. But as long as Sun supported OpenSolaris this was never going to happen, because it was too much work for too little gain (and too much difficulty), and so the fundamental uncertainties of OpenSolaris were not going to go away.

In this respect, Oracle did OpenSolaris a favour by throwing it off the back of the truck and forcing people to create Illumos from the wreckage. It is far from certain that Illumos will succeed, but at least the ambiguity of the situation is gone; within not too much time either we will have a viable independent alternative to Solaris or it will be clear that Illumos cannot deliver and is dead in practice (regardless of what people say). As a result I feel much more positive about Illumos than I ever did about OpenSolaris.

Another way to put this is that I never felt Sun was committed to OpenSolaris. People seem clearly committed to Illumos and they have significant money on the line as a motive to make it work, so it is now much more likely that we will wind up with a genuine alternative to Solaris.

Comments on this page:

From at 2010-12-02 22:30:08:

Sun was committeed to OpenSolaris. They were also committed to going bankrupt because they had no clear plan how to gain revenue from OpenSolaris. Thus they are no longer a company.

Written on 02 December 2010.
« A modest idea on how to get people to use Fedora Rawhide
Why https was a mistake, but an inevitable one »

Page tools: View Source, View Normal, Add Comment.
Login: Password:
Atom Syndication: Recent Comments.

Last modified: Thu Dec 2 01:06:47 2010
This dinky wiki is brought to you by the Insane Hackers Guild, Python sub-branch.