At that same WWDC Apple announced Time Machine, a product that would record file system versions through time for backup and recovery. How were they doing this? We were energized by the idea that there might be another piece of adopted Solaris technology. When we launched Solaris 10, DTrace shared the marquee with ZFS, a new filesystem that was to become the standard against which other filesystems are compared. Key among the many features of ZFS were snapshots that made it simple to capture the state of a filesystem, send the changes around, recover data, etc. Time Machine looked for all the world like a GUI on ZFS (indeed the GUI that we had imagined but knew to be well beyond the capabilities of Sun).
Of course Time Machine had nothing to do with ZFS. After the keynote we rushed to an Apple engineer we knew. With shame in his voice he admitted that it was really just a bunch of hard links to directories. For those who don’t know a symlink from a symtab this is the moral equivalent of using newspaper as insulation: it’s fine until the completely anticipated calamity destroys everything you hold dear.
So there was no ZFS in Mac OS X, at least not yet.
Somewhat related: the history of Microsoft’s WinFS.
I had my suspicions that oracle killed the prospects for ZFS with OSX. I get why Oracle doesn’t want to share, but it is such a shame really.
Bullshit! ZFS/OSX fiasco happened many years before Ellison’s acquisition of Sun.
1) ZFS was supported in Mac OS X Leopard (2007)
2) Time Machine was released also in 2007 using HFS+. (Sun’s CEO was Jonathan “Pony tail” Schwartz and McNealy the chairman)
3) Oracle bought Sun in 2010!!!
Oracle and ZFS/OSX fiasco are not related at all. Pure bullshit.
If you have counter-evidence against what’s written in the ARTICLE, then please link to it.
It wasn’t about Oracle or Sun.
http://www.zdnet.com/article/netapp-claims-suns-zfs-violates-its-pa…
Edited 2016-06-16 00:58 UTC
That was 2007.
The ARTICLE then goes on to explain events after that.
Yup. And that’s when Apple ditched ZFS.
And if you read the ARTICLE, it goes on to talk about events that happened AFTER.
What is hard to understand about “more things may have happened than the version of events I’m aware of”?
Edited 2016-06-16 05:06 UTC
I wasn’t bashing the entire article, I was bashing a quote against Larry Ellison.
ZFS was dropped from Mac OS X many years before Ellison’s acquisition of Sun Microsystem.
Wow, @kwan_e is never wrong, is he?
Facts. Learn them.
The events are closer together in time than you’ve indicated. Oracle’s acquisition of Sun was definitively agreed upon in April 2009.
It’s clear apple wanted ZFS and managed to get it out on some of it’s products, perhaps before the legal issues were settled. We also know OSX ZFS deployment was rolled back in 2008 for legal reasons. Apple’s ZFS port was officially “discontinued” in October 2009, after the acquisition…
http://daringfireball.net/linked/2009/10/23/zfs
You say the interest in ZFS during the oracle acquisition is “pure bullshit”, but timing-wise ZFS still made sense and quite frankly I think it would be much more perplexing if apple had not attempted to negotiate ZFS with oracle before giving up. Even if ZFS was a long shot, in all probably the events likely did take place as reported. I see no reason to dismiss the report. Do you have any evidence to the contrary?
Edited 2016-06-16 08:48 UTC
I researched it pretty extensively, and came to the same conclusion. I’m sure Oracle being Oracle, probably told/asked Apple to pay up as well.
I deal with certain Oracle software originally created by Sun. It is being used widely at Apple, one of largest product customers, for many years now. With that in mind, it sounds strange that Jobs and Ellison were refusing doing business together.
They couldn’t have stopped them anyway. They can’t stop OpenIndiana, Illumos, FreeBSD, or (now, apparently) Ubuntu.
OpenSolaris was Open, and it had ZFS.
Don’t need to stop. Just FUD users away. Oracle is a good Corp, but needed of urgent business re-engineering.
You’re missing an important point: OpenSolaris, FreeNAS, Ubuntu, etc are all free to download. OS X on the other hand was, at the time of the zfs happenings, a paid product. This alters the legal landscape in a lot of ways, both subtle and overt. Ianal, and I’ve not read the license for zfs and/or Oracle’s distribution terms, but I suspect there are relevant parts of these documents which would pertain to paid products and not to free software.
This is false. In fact, at the time Apple was creating Darwin installers which would install an entire OS that was open source, and was a subset of the open source packages that Apple provided.
You can still to this day download the OS X kernel (both the Mach and BSD components, and the glue in between) from Apple and compile it yourself. The core of OS X is, and always was open source.
That doesn’t really change anything legally. OSX is still a paid product. It matters not that is based on an open source core.
This might be better for Apple in the long run. ZFS is very robust, and supports a ton of features that aren’t necessarily great for single-disk systems, which is all Apple sells, and looks to be selling in the future.
Sure, one of those features is native support for a Fusion-type setup (smallish SSD and large spinner working together), but Apple had that already.
In the long run, there would really be two possible outcomes:
Either all those other features would have been a support burden for Apple in the long run, creating a fair amount of technical debt for them for use cases that Apple users will never come across,
or,
they would eventually have a divergent, incompatible version of ZFS.
I know ZFS isn’t changing that rapidly, but considering the length of time HFS+ has been in use…
Now, if Apple brought back a real MacPro again, that would be different, but somehow I don’t think that’s the case…
Remember that when ZFS was being peddled as the “saviours of Apple’s FS woes”, they were also selling OSX servers like XServe and and XServe RAID.
There was a notorious bug/limitation which effectivly meant that SQL performance fell off a cliff under heave load. This, effectively, killed their server business before it got off the ground in the enterprise.
http://www.anandtech.com/show/1778/5
ZFS potentially solved this issue(or people thought it would at the time), as well as making them able to compete directly with big boys thanks to other features it offered.
Adurbe,
Great points! It was common knowledge that HFS+ was obsolete and apple was seeking to replacement it, any in house FS would be many years away. ZFS was ready then and hit lots of checkboxes for big enterprise. This is just casual speculation, but the fact that apple was unable to get this killer feature may have been a contributing factor in apple’s decision to announce the end of XServe line of products in november of 2010.
http://www.macworld.com/article/1155483/xserve_discontinued.html
Edited 2016-06-16 14:36 UTC
That article is about OSX’s horrible threading in Tiger, though.
That, at least, couldn’t have been helped with ZFS.
From what I understand COW filesystems are not ideal for storing databases.
I always thought it was dropped as it has/had some pretty heavy resource requirements.
On a single drive client machine im sure users wouldnt be best pleased with the filesystem used 25% – 50% (1GB – 2GB) of a machine with 4GB (machines back then had even less).
If apple could just do checksuming in conjunction with TimeMachine i would be really happy.
I have a server at home which runs FreeNAS and ZFS purely to ensure that precious things like photos, home movies and other important data is kept secure and intact.
Congrats! REM2000
REMembering Us of right ways things should be done.
But a lot to Store Monetization Schemes:
ECJ has released ‘opinion’ on the Vereniging Openbare Bibliotheken case…
http://www.theregister.co.uk/2016/06/16/ebooks_same_as_printed_says…
Monetization Schemes better sensitize on sharing and reselling, or abandon the battle line.
On how wise tested Legislation is leveraged as framing to guide opinion making on new challenges.
…are you a bot? Or is there some other reason that 95% of your posts here are just seemingly non-sequitur replies to yourself?
BallmerKnowsBest,
This has been asked, and not denied… I also would like to know what the story is behind dionicio, some of his posts make sense, others seem completely random.
Dionicio, I have a few questions, if it’s not too much to ask:
Where are you from?
How old are you?
What do you do for employment?
Do you have your own website you could share here?
Is this scenario:
I buy a very good cotton paper [Dutch] Book, and once read, put it at my library.
Time will come when my daughter [or the son of my daughter] could consider giving copies from it to family and friends. Knowledge a treasure to be shared.
Copyright Digital Management simply doesn’t have a way to allow COPYRIGHT CADUCITY.
If Bibliotheken had bought that book on paper, will have no problem on allowing copying of a caducated CR.
And that problem is inherent to every Non open specification compliant CR layer out there [all of them?].