Linux Gets A Proprietary, Read-Only ReFS File-System Driver

[phoronix]

Phoronix: Linux Gets A Proprietary, Read-Only ReFS File-System Driver

While Microsoft's ReFS file-system has been around for a few years to date there is no mainline Linux kernel driver supporting this file-system that's more advanced than NTFS. But there is now a read-only ReFS Linux driver and it's proprietary...

http://www.phoronix.com/scan.php?page=news_item&px=Read-Only-ReFS-Linux-Driver

[Veto]

I couldn't really tell from their web-page, whether it is a FUSE driver or a kernel module (GPL violating?). Does anyone know more?

[Michael]

I couldn't really tell from their web-page, whether it is a FUSE driver or a kernel module (GPL violating?). Does anyone know more?

I believe it's a real kernel driver, not FUSE based.

[doublez13]

I couldn't really tell from their web-page, whether it is a FUSE driver or a kernel module (GPL violating?). Does anyone know more?

Its only GPL violating if they we're to try to distribute it already in the kernel. You can legally do whatever you want to your own kernel after you install it. You just can't redistribute it if it contains non copyleft compatible code.

[bug1]

Its only GPL violating if they we're to try to distribute it already in the kernel. You can legally do whatever you want to your own kernel after you install it. You just can't redistribute it if it contains non copyleft compatible code.

The company who developed the module can legally do whatever they want with it, but they are violating the GPL if they distribute binaries to customers (because it obviously container non-copyleft code).

The GPL has been infringed when its distributed to the customer, not when the customer installs it.

[higuita]

if they distribute the code and you compile, they are safe... if they distribute a binary driver, ready to use, they are breaking the GPL, as they are linking the closed code against the GPL code. If they have a GPL layer that links to their close code, like nvidia... may look be a little better, but unlike the nvidia drivers that in theory is the same driver for all OS, with compatibility layer, this driver is build directly for linux and so, derivative code.

The way for then to be safe is to use fuse, it is there for this. Going for the kernel with close code, they are playing with fire.

Finally, for those wanting cross OS filesystem, there are one FS that everyone supports and it is always forgotten: UFS https://en.wikipedia.org/wiki/Universal_Disk_Format

This FS was build for DVD, specially DVD-RW, but have standard extensions for normal disk usage.
Format a HD with it and you can mount and read/write in almost all OS without any problem.

This FS have only 2 problems: most people do not know it, and as consequence, there are few tools to recover/repair it... but other than that, works very well.

So use it on DVD, HD, memory cards and usb pendrivers, if you need to use then in several OS

[starshipeleven]

The company who developed the module can legally do whatever they want with it, but they are violating the GPL if they distribute binaries to customers (because it obviously container non-copyleft code).

The GPL has been infringed when its distributed to the customer, not when the customer installs it.

No. Distributing blobs like that isn't a GPL infringement.

Paragon makes FAT and NTFS drivers for like the whole embedded devices of this universe, only thing that happens is that the kernel becomes "tainted", it sets a flag so that any bugreport can be clearly identified as coming from a system with binary drivers, so kernel developers can simply answer "no support for you".

[FishB8]

if they distribute the code and you compile, they are safe... if they distribute a binary driver, ready to use, they are breaking the GPL, as they are linking the closed code against the GPL code. If they have a GPL layer that links to their close code, like nvidia... may look be a little better, but unlike the nvidia drivers that in theory is the same driver for all OS, with compatibility layer, this driver is build directly for linux and so, derivative code.

The way for then to be safe is to use fuse, it is there for this. Going for the kernel with close code, they are playing with fire.

There is a shim layer in-between the binary and kernel code. It's not possible to distribute a direct binary module unless you only plan to support a very specific single kernel build.

[timofonic]

Its only GPL violating if they we're to try to distribute it already in the kernel. You can legally do whatever you want to your own kernel after you install it. You just can't redistribute it if it contains non copyleft compatible code.

They could make like Samsung with their XFAT driver, but STDU about it. I'm sure they are able to obfuscate source code quite wel, they a russian developers

WinFS seems like a bad copy of both ZFS and BTRFS. Anyway, it may happen to become mainstream on future Microsoft profucts and that's even worse than current exFAt situation.

I just hoped they were able to make WinFS a reality, maybe Btrfs would get some competition and become really usable in all ways right now.

[Daktyl198]

Can't we all just agree that whatever I install on my supposedly free Linux system is my own damn business?