Paragon Looks To Mainline Their NTFS Read-Write Driver To The Mainline Linux Kernel

[ferry]

Why don't Linux kernel devs not use GitLab like normal people? Sending patches to mailing lists is stupid.

Do you mean it is finally possible to hold a linux repo on gitlab?

[oiaohm]

Now I'm curious how they have managed to release ntfs3 under GPL considering a huge number of patents and copyrights related to NTFS.

The realities with copyright is that it does not come into play as long as the code is independent development. Patents don't come into play until you start distributing and using the code. Yes GPL does have some implied patent protection. Please be aware Samsung exfat was released under GPLv2 for a long time since 2013 the patent issues were only sorted out with a OIN agreement for Linux kernel end of 2019. Submitting the code to mainline to come included in the OIN define of what Linux is triggers the Patents problem to be addressed if they exist.

Something else that is easy to miss birdie. NTFS 3.1 file system design was in fact released October 2001. Yes no new features since October 2001 have been added to the NTFS file system used by Microsoft. Patents last max of 20 years so the end of NTFS patents is either October 2021 or October 2022 depending on the jurisdiction. It could in fact take until end of NTFS patents for the NTFS driver to be accepted as merge-able quality. So patents may not even come into play.

[oiaohm]

Windows from Vista onwards has separate small boot partition. Technically you need only that to be NTFS, actual root file system could be ReFS.

https://docs.microsoft.com/en-us/windows-server/storage/refs/refs-overview

You did not read above. The answer is absolutely no due to missing features.
The following are missing REFS feature and effects
1) Extended attributes Yes current versions of windows does require these on the real root partition.
2) Short names need for legacy application support. Funny enough some of the windows bundled applications are legacy as in expect the short file names of particular things.
3) Page file support installer would have to create another partition for this.
4) File system compression this is in fact under Windows used to reduce the install size. Not absolutely require but not having is going to be horrible.

That is basically in order of need to be implemented. Extended attributes functional is a bare min to have a Windows 7 onwards root file system that requirement is why you could install Windows Vista on fat32 and Windows 7 will not. The reality is the first desktop windows with ReFS is Windows 8.1 so the support has been long gone for installing on a partition without extended attributes at that point.

Please note I am not saying some future version of windows will not install on ReFS as root file system but no current version of windows can because the feature/s required is not implemented in ReFS yet.

[CommunityMember]

Surprised this comment coming from an account since 2010.

Age has never implied that wisdom has been acquired. Age simply eliminates one common excuse for rejecting the knowledge available (the "young and stupid" defense).

[CommunityMember]

Now I'm curious how they have managed to release ntfs3 under GPL considering a huge number of patents and copyrights related to NTFS.

You can have GPL code implementing IP protected algorithms (go look at FFmpeg as a well known example). USING that code may, or may not (depending on jurisdiction and IP protection status in that jurisdiction), require a license. OIN provides various rights and protections for those companies that have signed that agreement. But not all companies (or for that matter all distribution(s)) are members, making things possibly complex going forward for some.

[birdie]

First of all, huge thanks to the company for going open source with their kernel NTFS driver which means NTFS could become a first class citizen in Linux and can even be used for hosting your root partition (it has all the required features). Also, to be honest NTFS-3G is horrible: it's very slow, it has caused file system corruptions for me several times (that's why I stopped using for downloading torrents - this type of activity kills it with ease), it doesn't have a proper bug tracker or development workflow. It more looks like "eat what you're given", it hasn't seen new released for three years now (!), IOW it's no longer maintained.

Now I'm curious how they have managed to release ntfs3 under GPL considering a huge number of patents and copyrights related to NTFS.

[skeevy420]

Surprised this comment coming from an account since 2010.

I'm not. Everyone has their own preferences and it's only a matter of time before someone at random says "Why not X instead of Y?".

I don't really care for the GitLab interface all that much so I chuckled at the original comment. It's starting to grow on me since more and more projects are using it so I find myself there more often, but I still prefer the older GitHub interface because their new setup isn't very intuitive to me after using the old one for so long.

[ksec]

Why don't Linux kernel devs not use GitLab like normal people? Sending patches to mailing lists is stupid.

Surprised this comment coming from an account since 2010.

[aht0]

https://docs.microsoft.com/en-us/windows-server/storage/refs/refs-overview

That is a think. ReFS is not able to be bootable with windows. So Workstation/Enterprise editions windows is still installed on NTFS. Yes NTFS is still the default. If you read above ReFS is still missing lots of features. Yes ReFS missing features difference between XFS vs ZFS look like minor issues.

Windows from Vista onwards has separate small boot partition. Technically you need only that to be NTFS, actual root file system could be ReFS.

[kpedersen]

I don't send mails to mailing lists because mailing lists are stupid.

What would you recommend? They "chill out" on consumer Slack?