The New NTFS Driver Looks Like It Will Finally Be Ready With Linux 5.15

[doublez13]

I wonder if Linux can be installed on NTFS partition with this driver. That would be quite funny.

I would assume a fair bit of work would need to be put into grub to support the different NTFS features.

[dragon321]

I wonder if Linux can be installed on NTFS partition with this driver. That would be quite funny.

Not a big fan of NTFS but It's very nice that we'll finally get this driver in Linux kernel. While exFAT is enough in many cases where interoperability with Windows is needed, it lacks some features like journaling so having good support for NTFS would help here. Yeah, there is ntfs-3g but kernel driver brings some advantages compared to FUSE file systems.

[kbios]

(Meanwhile I'm in an absolute minority of people who run their ext4 volumes with journalling disabled).

I hope you have a robust backup system

[birdie]

I agree. In that regard I've a question about Michaels final dig: "but these days NTFS is of declining relevance to companies now focused on exFAT and other file-system options, which has led [Paragon] to be open to upstreaming their code."

Given exFAT has advantages over FAT, in what ways does it or can it replace NTFS? I've always (perhaps mistakenly) thought all the FATs were somewhat toys compared to journaling filesystems such as NTFS, EXT4, HFS, and their ilk. (And where does f2fs fit in?) My question is what use cases do Paragon's paying customers have that previously required NTFS but now can at least get by with exFAT, use cases that with earlier FATs could not?

For the record, I'm really looking forward to Paragon's new NTFS3 for my own development and testing efforts split between Windows and Linux. I'm using ntfs-3g on a shared SDD, but have had to duplicate much of my effort onto a linux-only ext4 partition simply because fuse is too slow.

exFAT is (good) for removable storage and file archiving. It can't and won't work for storing anything which requires MAC, so NTFS is as relevant as ever.

Also exFAT doesn't support any form of links (software/hardware) which is really a bummer and I wonder why Microsoft didn't implement them. I guess they presumed that exFAT is mostly used by non-technical people who would have utterly confused by links (and what that reason I guess Google eliminated the Google Photos/Drive integration which was an absolute bummer for millions of people who were robbed of a feature which cannot be replaced by anything).

The absence of journalling is actually a good thing for SSD drives where this filesystem is primarily used.

(Meanwhile I'm in an absolute minority of people who run their ext4 volumes with journalling disabled).

[CommunityMember]

My question is what use cases do Paragon's paying customers have that previously required NTFS but now can at least get by with exFAT, use cases that with earlier FATs could not?

I think the number one use case is large files. FAT had a 4GB limit per file, which is rather constraining in today's world. For the ubiquitous flash drive exchange(s) between consenting adults, exFAT is now the common choice for one file system that can be read/written almost everywhere.

[pipe13]

So I believe the real answer was that it was about money. It is always about money.

I agree. In that regard I've a question about Michaels final dig: "but these days NTFS is of declining relevance to companies now focused on exFAT and other file-system options, which has led [Paragon] to be open to upstreaming their code."

Given exFAT has advantages over FAT, in what ways does it or can it replace NTFS? I've always (perhaps mistakenly) thought all the FATs were somewhat toys compared to journaling filesystems such as NTFS, EXT4, HFS, and their ilk. (And where does f2fs fit in?) My question is what use cases do Paragon's paying customers have that previously required NTFS but now can at least get by with exFAT, use cases that with earlier FATs could not?

For the record, I'm really looking forward to Paragon's new NTFS3 for my own development and testing efforts split between Windows and Linux. I'm using ntfs-3g on a shared SDD, but have had to duplicate much of my effort onto a linux-only ext4 partition simply because fuse is too slow.

[perpetually high]

So I believe the real answer was that it was about money. It is always about money.

Yup. True for many things. Money makes the go round, and that's not changing anytime soon

[CommunityMember]

Great, still don't understand the reasoning why the open source it, but am very happy with it. Making the world a better place

I believe there were a couple if things that helped the process. The first is that Microsoft joined OIN, which removed the IP issues with use of certain technologies in linux (it has always been unclear what, if any, patents exist on ntfs, but no one was comfortable being a test case). The second was that it seems likely that maintaining it (out of tree) for Paragon's customers was a cost without strong benefit for the number of sales they were seeing for the product (bring it in-tree means Paragon no longer needs to create per-kernel version modules, and also means customers using secure boot technologies could more easily use the driver without needing to install additional keys per machine).

So I believe the real answer was that it was about money. It is always about money.

[jarekZ]

Great, still don't understand the reasoning why the open source it, but am very happy with it. Making the world a better place

maybe NTFS is losing its popularity among the device manufacturers so it was not so profitable for paragon to continue selling this driver as a proprietary product. On other hand, they're getting a lot of (good) PR from this story. Also, looks like the person sending these patches isn't just a developer, but an owner of this company. So he's doing it simply because he can, why not?

[Slartifartblast]

Excellent news