Phoronix: Is OpenSolaris About To Be Forked As Illumos?
There are still a few weeks left before the deadline that demands Oracle appoint a community liaison for their OpenSolaris operating system that is capable of communicating their future intentions to the OpenSolaris community (like where the hell is OpenSolaris 2010.1H) or else the OpenSolaris Governing Board will return control of the community back to Oracle. However, some OpenSolaris community developers have already had enough: they've begun work on a new project...
Oracle seem very confused when it comes to Solaris and OpenSolaris, clearly it isn't a good investment long term, nor short term as they are competing with Linux and their own Unbreakable Linux offering specifically. They will never be able to match in terms of investments and code that produced by the Linux kernel.
It is merely sucking money away from other projects that do bring value to Oracle and the wider Open Source community.
It may be cruel, but it would be best for everyone long term if the project was responsibly dismantled, existing support contracts fulfilled and then entirely disbanded.
A nice start would be to relicense the IP and code they have so it can be ported to Linux which would offer customers a migration path to Unbreakable Linux if adopted upstream.
The problem with that is that there are a lot of major customers that have a significant investment in Solaris. These investments come in the form of money, time, and programs that are targeting Solaris. Many of these programs would of been made in-house or through contractors and that sort of thing. Even if they purchased software that runs on Linux migrating over probably would require a re-licensing, which can be very expensive.
And it's not just OSes, it's architectures that are different. They would be designed for a different CPU and would be targeting large machines with a lot of CPUs and whatnot that simply do not exist unless you go to another high-end architecture like you see from IBM or HP.
The cost of migrating away from Solaris and Sparc would be very prohibitive. Linux would require a different architecture and to be cost-effective they would have to port the application to a cluster configuration, which would be extremely expensive for anything that makes heavy use of threads.
I am not saying that is going to happen to everybody, but it's going to require Oracle to support Solaris and Sparc for the foreseeable future pretty much regardless of how much it costs. Don't worry. they will pass those costs to the customer with a hefty premium. Support contracts of this type should be very lucrative for Oracle.
So what they will probably do is continue to improve Solaris in some ways, but nothing significant. They may release new Sparc systems to use up the current hardware IP they have. Then downplay Solaris x86-64 and do what they can to discourage anybody doing anything new with Solaris.
Thus it will die a slow and draw out death over the course of the next decade, or two.
Solaris will not die. IBM has offically said that they are phasing out AIX to be a Linux company. This is official and outspoken. Hence, AIX will die. On the other hand, there are no plans that Solaris will be killed off.
Larry Ellison said officially he bought Sun to get the two key assets: Java and Solaris. That is why he invested several billion dollars. I dont think he invests billions dollars to get technology, that he later kills? That is bad business, and Larry is not a bad businessman. He is actually one of the richest in the world.
Oracle database runs on Solaris, more than any other OS in enterprise companies. In fact, Solaris is the recommended reference OS to run Oracle database - this was outspoken long before Oracle bought Sun. Solaris and Oracle database plays together extremely well, and as Solaris is for Highend and Linux for Lowend (according to Larry) I dont see why Larry would kill Solaris. He needs Solaris to run Oracle database as good as possible.
OpenSolaris source code will be the basis for next gen Solaris 11, which will probably be announced next month at Oracle Open World. I dont see how Oracle can scrap all OpenSolaris source code, if Oracle wants to release next gen Solaris 11. The OpenSolaris source code is rapidly developed with lots of new exciting functionality, which you can see the checkins in the repository.
Solaris will continue to be the best Unix out there. With ZFS, DTrace, etc it keeps the top spot and magnifies the lead. According to a study from HP just recently, Solaris is the High End Unix most used, and has the "brightest future"