No way. NetApp will rabidly chase and kill whoever dares trying building commercial ZFS applications based upon OpenSolaris or *BSD (CDDL doesn't cover you against patents infringements, remember?).
I know a well-known privately-held extremely-high-end storage company who just canned its OpenSolaris/ZFS projects because they won't take the risk to be sued by NetApp.
> They'll all expire eventually. ZFS (like) file systems will eventually be very succesful.
Yes, that's all in the "like". In 4 or 5 years, btrfs and others will mature and ZFS will look less exceptional. That, and ZFS has a major culprit which will become more annoying with time : it's not cluster-aware.
I've read about NetApp /filing/ the lawsuits, but I haven't read about any results... Until it's over, it's just scary paper. have any of the lawsuits gone to court?
Yes, I'd certainly think twice about starting a company in that market right now. but my point is that to my knowledge, it's not over just yet for the companies that got sued... people keep talking like they are dead, and yeah, their future is in question, but "cold dead" ---? my understanding is that as of this moment, this is not true.
It won't be. Oracle will keep the CDDL, and use ZFS to market Solaris as a USP over Linux, until (and probably after) btrfs becomes mainstream on Linux.
Erik Trimble answered this: "With the various Linux
projects, Oracle was pretty much required to share back, so they played nicely. With Solaris, the mindset seems to be that "We own this, so let's make fat bank on a cool technology, and not let others steal our business"." http://mail.opensolaris.org/pipermail/opensolaris-discuss/20...
The whole post is extremely interesting BTW, and spot on. I know many a Solaris sysadmin, managing sometimes ancient machines (up to 10 years old or more) that were more than pissed of by the latest support changes. Be sure they (and many others) won't buy Sun/Solaris anymore if they can, and usually, they can.
It really isn't. They liked Linux because they could sell their database on white box hardware and not pay Microsoft (more money for Oracle less for a competitor). This move is designed to bring Solaris in line with their other enterprise software products. They have an OS and Server combination that can help them make more money.
Their Java move with Google is simple. IBM paid for Java, Google didn't. Oracle wants the cash.
Comments
If at least ZFS can be salvaged from the wreck...
No way. NetApp will rabidly chase and kill whoever dares trying building commercial ZFS applications based upon OpenSolaris or *BSD (CDDL doesn't cover you against patents infringements, remember?).
I know a well-known privately-held extremely-high-end storage company who just canned its OpenSolaris/ZFS projects because they won't take the risk to be sued by NetApp.
ZFS as open-source is cold dead. Sad, but true.
The only exposure that ZFS has is Patent Based. Some of the interesting ones are expiring in the next four to five years:
http://www.google.com/patents?as_q&num=10&btnG=Googl...
They'll all expire eventually. ZFS (like) file systems will eventually be very succesful.
> They'll all expire eventually. ZFS (like) file systems will eventually be very succesful.
Yes, that's all in the "like". In 4 or 5 years, btrfs and others will mature and ZFS will look less exceptional. That, and ZFS has a major culprit which will become more annoying with time : it's not cluster-aware.
I've read about NetApp /filing/ the lawsuits, but I haven't read about any results... Until it's over, it's just scary paper. have any of the lawsuits gone to court?
Building a business around the hopes that pending litigation will be dropped doesn't sound like a very solid plan...
Yes, I'd certainly think twice about starting a company in that market right now. but my point is that to my knowledge, it's not over just yet for the companies that got sued... people keep talking like they are dead, and yeah, their future is in question, but "cold dead" ---? my understanding is that as of this moment, this is not true.
It won't be. Oracle will keep the CDDL, and use ZFS to market Solaris as a USP over Linux, until (and probably after) btrfs becomes mainstream on Linux.
The btfs developer works at Oracle (and also works on OCFS).
It is a bit hard for me to reconcile Oracle's contributions to Linux with their seemingly mercenary attitudes towards Solaris and Java.
Erik Trimble answered this: "With the various Linux projects, Oracle was pretty much required to share back, so they played nicely. With Solaris, the mindset seems to be that "We own this, so let's make fat bank on a cool technology, and not let others steal our business"." http://mail.opensolaris.org/pipermail/opensolaris-discuss/20...
The whole post is extremely interesting BTW, and spot on. I know many a Solaris sysadmin, managing sometimes ancient machines (up to 10 years old or more) that were more than pissed of by the latest support changes. Be sure they (and many others) won't buy Sun/Solaris anymore if they can, and usually, they can.
It really isn't. They liked Linux because they could sell their database on white box hardware and not pay Microsoft (more money for Oracle less for a competitor). This move is designed to bring Solaris in line with their other enterprise software products. They have an OS and Server combination that can help them make more money.
Their Java move with Google is simple. IBM paid for Java, Google didn't. Oracle wants the cash.
Good point - that's one thing I wonder about - although nothing's been said, it wouldn't surprise me if Chris Mason left Oracle for IBM or Red Hat.
Why is that? Isn't ZFS part of Solaris? Or at least has the same licensing restriction as OpenSolaris?
ZFS is licensed separately and apart from Solaris, but it's still under the CDDL