Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Your point is?
My guess is that he was trying to make a rather humorous point about Hans Reiser, the "father" of Reiser4. He murdered his wife, thus the "yes" on the "Will murder your wife"-column in the picture he posted.


I found it terribly funny, literally.
 
Two questions:

What exactly are the advantages of ZFS?

If ZFS is not used in OS X nor in Windows, then what is the point of its existence? Why did Sun make a complex project and then not let anyone use it?

Sun has their own operating system - Solaris. It's for high-end server computing, and so is ZFS. It's very good though, and since I don't have time to explain why right now, you'll have to google it or believe me when I say that if it was included in Windows Server and not Mac OS Server, I would switch immediately.
 
What they really need to do is not necessarily rewrite from scratch what they have in HFS+ but rather to add the features that benefit their needs.
They have been doing that for the last 10-15 years. There is a point where it just isn't worth the hassle.

Snapshots really aren't necessary if you have a solid Time Machine. Preventing "silent corruption" with checksum really isn't that necessary. What is necessary is the fs taking on some of the tasks of storing data that that today is strewn across the OS.
Snapshots makes stuff like Time Machine easy and cheap to implement. Ever heard of copy-on-write?

Who here looks at .ds store littered around and thinks "geeze this is archaic".
Those files are still needed on filesystems that don't support their needs. Exactly the places where you are most likely to notice them (FAT32 thumb drives).
 
Too bad they couldn't spend some of that $30+ billion in the bank to "overcome any licensing issues" ;)
 
They probably could before Oracle bought Sun. Oracle doesn't want ZFS to succeed, cause it'll never make them any money. They're already invested in BtrFS, and have planned to support that for the foreseeable future.
 
I'd rather have Apple custom developer their own Filesystem.

ZFS is nice but many of the features it offers are for Enterprise. Apple need only support the really good features and keep their FS light and for speed and functional for their area of computer (consumer, small biz)

The Consumer and the Enterprise are merging.

You can count on the next file system to read ZFS filesystems but extend Apple only technologies that work in distributed networks.

This is exactly what we talked about doing during the merger with Apple, while at NeXT.

They probably could before Oracle bought Sun. Oracle doesn't want ZFS to succeed, cause it'll never make them any money. They're already invested in BtrFS, and have planned to support that for the foreseeable future.

Your argument makes no sense.

They will make BtrFS compatible with ZFS so they can have mixed clusters of Solaris and Linux.

They have two channels of different licensing schemes that gives Oracle more options to their clients.
 
While I see the benefits of ZFS and was disappointed it did not make it to Snow Leopard, how is it beneficial to shut down the open source project?

Even if Apple could come out with its own version with a easy to use GUI perhaps built into Disk Utility, what's the harm in having the open source project alive to share in its development and possibly come up with new solutions?
 
Your argument makes no sense.

They will make BtrFS compatible with ZFS so they can have mixed clusters of Solaris and Linux.

They have two channels of different licensing schemes that gives Oracle more options to their clients.

Yeah... MY argument makes no sense...
 
The main and really good features (volume management, filesystems, snapshot management) are the very useful for enterprises. If you have any things that are only useful for enterprises, please enlighten me.

It sure would help out their Time Capsule product - versioning built into the OS Filesystem would make all the little issues that TC has a thing of the past. Not to mention, regular users would benefit from versioning in so many ways..
 
Filesystems are a difficult one. I would think that if Apple are indeed planning a "new" filesystem for 10.7 it would have to be some feature rich abstract extension to HFS+ rather than something completely new.

The big problem with massive filesystem changes is that it may require you to reinstall the whole OS...

Will Apple really want to go down the route of XP and Windows 7... which requires a complete backup and reinstall to "upgrade"?
 
Developing it themselves doesn't mean it won't be open source when it's done.

True. But it does mean that only Apple engineers are involved in the process of architecture and design. In addition, it wouldn't likely achieve anything more than marginal interest in the *nix community. Boy howdy, having a single FS that most Linuxes as well as OS/X could agree on as a default FS would be tremendous!

It just feels to me like they'd have a lot to lose if they aren't connecting with the Open Source community as a whole. There are a lot of smart people who have made File Systems their life who could have a lot to contribute.
 
Well, how about if Apple made it an option to first take a full TM backup, then make the new FS, then roll the backup back onto the drive? Any new features like versioning and such could then ALSO be put on the new FS, since that's the main feature of TM.
This wouldn't necessarily be possible only at install, but a choice in SysPrefs at any time.
 
I don't think it's possible for Apple to "shut down" an Open Source project. :rolleyes:

Ignoring that ZFS is a Sun technology, not an Apple one, it's Open Source! Yeah, maybe Apple shut down the code repository, but anyone with a copy of the last snapshot can start a new one on SourceForge, Google Code, etc.

Even if Apple bought ZFS and decided to take it to proprietary, the project can simply fork at the last GPL version and keep going on it's own. There's plenty of people interested in seeing ZFS through whether it's on Apple's platform or not. ZFS is not a "Mac thing".

Sensationalist headline is sensationalist.
 
This is not Sun's fault. Apple never put in the resources to make ZFS a viable part of OS X. They also put a lot of effort into improvements to HFS+, and that appears to have won out at least for now in Apple's internal planning.

I disagree slightly here.... the last revision listed on macosforce was zfs-119. This was an older build of ZFS for OS X

This build was NOT the last development carried out on ZFS for Snow Leopard. Snow Leopard 10a286 contained the latest (and last public) ZFS kernel extensions. The binaries from 10a286 work just find in 10.6.1!!! You can create, read and write! The only unusual behaviour is if you have an external ZFS hard drive which you unplug without first exporting the zpool. In fact the binaries are so stable I can boot into full 64bit mode and they still work just fine!

Based on this, Apple developers were right on making ZFS work for read and write in Snow Leopard.... i'm certain it was something specific that made them stop.
 
Snapshots makes stuff like Time Machine easy and cheap to implement. Ever heard of copy-on-write?
To visualise it even better. With ZFS it would be possible to make backups while you aren't even near your backup drive. And when you do you can just sync to it by simply transferring the changes. This way you can never miss a backup moment, do block-level backups (no more big transfers because one byte changed), and more efficiently transfer and use your backups.
 
Jfs

Why doesn't Apple pick up the phone and ring IBM for a non-GPL JFS?

JFS2 is used in AIX, its performance is good under a wide variety of conditions, and it is respectful of system resources so it could be used on svelte mobile devices.

Apple also has some decisions to make regarding exFAT support, since certain high capacity memory card formats (such as SDXC) use it.
 
Someone mentioned about Apple working on OS X working on Intel processors from the beginning or close to it. An insurance policy of sorts. Very true. Apple is beholding to no one. They'll work with you while it is to their advantage, but are prepared to jettison if they have to. Smart.

I'm with the majority here that feel that Apple has had an ace up their sleeves on this one. Of course, Steve Jobs is THE NEGOTIATOR. He's not going to let Oracle bully him. Yes, they have something that Jobs wants and yes, they are made entirely out of cold hard cash. But Jobs is never afraid to say "no". This is all part of the negotiations process.

That or Apple will write their own kick-ass file system.
 
All I really want is a more 'modern' filesystem (basically something more than FAT32) that is out of the box compatible and stable with both read and write operations across all three of the major operating systems. Right now its a major pain to have to deal with trying to create removable storage drives that can handle large files when moving between a Mac laptop and PC / Linux Desktop. Not to mention the university lab machines.

Really that is all I want :(
 
They have been doing that for the last 10-15 years. There is a point where it just isn't worth the hassle.


Snapshots makes stuff like Time Machine easy and cheap to implement. Ever heard of copy-on-write?


Those files are still needed on filesystems that don't support their needs. Exactly the places where you are most likely to notice them (FAT32 thumb drives).

Agreed on 1

Snapshots are not equivalent to Time Machine. In a storage device the snapshot features take a point in time picture of the whole system making it easy to roll back and fix problems. Apple has attempted to be more granular with Time Machine allow you to return an individual file rather than a backing up to a state as modern day snapshots do.

I agree files are needed but developers shouldn't have a handful of areas in which to read/write metadata.
 
From a practical point of view -- Who cares? Solaris is free. Anyone who wants ZFS can get Solaris for free and the ZFS implementation of Solaris is mature. People who need ZFS are running servers and Solaris is a better server OS anyways.

I've been waiting for Apple's version of ZFS only because I wanted to see how Apple would handle the user interface. ZFS is complex and there is no way the average Mac user could understand it. No matter how pretty you make the point-and-click user interface the concepts are over the heads of most users. So I was waiting to see how Apple could solve this problem. How would they hide ZFS from the user?

I think Apple, maybe, could not figure this out and maybe even concluded that it could not be done. Likely some one at Apple said "There is no way we can ship this, even if it works."
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.