Concerns Raised Over The "New" NTFS Linux Driver That Merged Last Year

[sinepgib]

Find in Torvalds Git and lkml mailing list.

1. That's not how forums work. You post a comment where you communicate something to others. If I need to not only go myself to a mailing list, but to assume I need to, then don't post in the forum, as those who would guess will already be following the mailing list and the comment is pointless.
2. I still don't know who you mean. There's an article for today that points to the PR for the NTFS driver with most commits made by the official maintainer, so either "they" is not Paragon or you're wrong. You mean independent collaborators? The distros?

Effectively, regardless of what the actual facts are your post remains unclear.

[timofonic]

Who? You quoted nobody and gave zero context.

Find in Torvalds Git and lkml mailing list.

[sinepgib]

Lies. They don't help at all. Their credibility is becoming below zero.

Who? You quoted nobody and gave zero context.

[timofonic]

Lies. They don't help at all. Their credibility is becoming below zero.

[dskouras]

A few comments on this interesting but sometimes odd discussion.

I use ntfs3, paragon ntfs for Linux retail version and ntfs3g extensively on several systems. In my my opinion, none are adequate as file system drivers, as that is a software category which requires perfect stability, reliability and maturity.


All 3 have problems that make them unnaceptable for anything more than hobbyist use. ntfs3g and ntfs3 have both at times left mounted volumes in a dirty state or led to the OS being unable to unmount them during shutdown. You would never accept that from ext4, btrfs or any file system that the OS claims to support.

Ntfs for Linux is only supported on old versions of the kernel (I think 5.11) although paragon support will send newer unreleased versions if requested. Whether you would use them in a production system is another question. On the other hand older kernels have issues with newer hardware especially AMD zen/epyc as well as 12th generation intel platforms. The commercial version has a full set of file system utilities but again the above issue is a deal breaker IMO. And it's reliability is not 100% either, though it is definitely better than the other two.

Paragon support used to be good (I also use their products for MacOS which are 100% rock solid) but lately response times have gone from 1-3 days to over a month. This is for the Linux products, I have not requested support for MacOS. This started before the Ukraine conflict.

So we simply can't safely share drives between windows and Linux and of course Microsoft doesn't care. MS is a listed corporation, thus required by law to generate profits for its shareholders. If they give away the value of their intellectual property, they can get sued for it. They open source code only as part of a wider strategy to generate profits not because they love OSS. They can't love anything, they're a legal entity, not people.

Paragon is I believe privately held company so they can do as they wish with their products. I think it would have been ideal if they had thoroughly documented the code they open sourced so others could easily take over it, though I suppose the technical skills to do so are scarce.

Also, as far as I know, ntfs3 is supposed to be a different driver from the netfs for Linux product, though I can't imagine how different they can realy be, given they both come from the same developer(s). If that is indeed the case, I think parsgon shou,d have open sourced the commercial product instead (I mean how much money could they possibly make from it?!).

But as with everything else in Linux it is what it is. Volunteers, individual contributors and small projects can only go so far.
Like others have said, I wish we had more central coordination of the core OS development, beyond the kernel, so key projects like file system drivers could have proper resourcing, funding and management. But we are clearly a long way away from such a situation and until then, the best we can hope for is major distributions like ubuntu, SUSE or fedora or indeed the corporations behind them, taking on some of the bigger projects.

[bemerk]

does the Paragon driver include support for reparse records?
Win10 has some weird special features

[linuxgeex]

I've had swap disabled both in Linux and in Windows for more than a decade now. People who insist on SWAP presence even if you have plenty of RAM (I rock 64GB on my desktop and 16GB on my laptop) don't understand why it's needed. I've never used hibernate either because I've got SSDs all around and I don't want to write gigabytes of data to save a few seconds of boot time. Then I'm not even sure hibernation with an SSD is faster than a cold boot. It might not even be the case.

Swap can improve performance even if you have more memory than you need, but only if you have long-lived processes which have large amounts of seldom-touched anonymous memory. If you swap that out then there's more room for page cache, which speeds up disk access. That was a lot more important before SSDs.

I have 24G on my laptop, 32G on my desktop. Both have ZRAM swap size 100% with one volume per core (not thread) and both have /swap files of 8GB, because on the very rare occasion that I manage to exhaust ZRAM swap, it's good to have that extra breathing room rather than having 15 minute lockups and OOM-killer acting with zero wisdom. I'd rather not lose my work. 8GB out of 1.5TB is not a large tax to pay for the security it adds. 13.3 cents per year over 6 years.

Hibernation is enabled on both. It preserves my work without consuming power to preserve state. So my desktop hibernates outside work hours, saving me $$ on the power bill. The laptop hibernates only when travelling. Suspend to RAM holds for 12-48h depending on state of charge. Triple the stock RAM is part of the reason for that short suspend life. Hibernate OTOH holds indefinitely. Does it cause extra writes to the SSD? Yes indeed. Over 3.5 years I've written a total of 18.5TB and 3.7TB to my 512GB TLC Silicon Power NVMe system disk (800TBW) and AData 960TB QLC data disk (200TBW). The system disk writes are dominated by Google Chrome's cache/session IO. Anyhow I will be long dead before they exceed TBW limits.

My buddy with the same laptop with 8GB RAM on Windows managed to kill the same system SSD thanks to swapping and using browser automation to harvest poker accounts 24/7. He didn't care that swapping made the UX choppy while the scripting runs. $70 SSD in the garbage felt like a win for him vs buying the RAM since he gets a new laptop every 3 years and gives the old one away.

[sinepgib]

For this exact reason, either Ubuntu (very likely) or SuSE (enterprise one) will step in and takeover the project. As SUSE is based in Germany, more likely.

Exactly what reason? Ubuntu and SuSE seemed very happy to ship ntfs-3g for decades for their users, and it doesn't even get installed by default. They are more likely to just do that if the kernel driver gets abandoned (which apparently wasn't the case).

[Nsane457]

Konstantin Komarov finally responded:

Hello Linus, Kari and all.

First and foremost I need to state that active work on NTFS3 driver has never stopped, and it was never decided to "orphan" NTFS3. Currently we are still in the middle of the process of getting the Kernel.org account. We need to sign our PGP key to move forward, but the process is not so clear (will be grateful to get some process desciption), so it is going quite slow trying to unravel the topic.

As for now, we can prepare patches/pull requests through the github, and submit them right now (we have quite a bunch of fixes for new Kernels support, bugfixes and fstests fixes) -- if Linus approves this approach until we set up the proper git.kernel.org repo.

Also, to clarify this explicitly: in addition to the driver, we're working of ntfs3 utilities as well.

Overall, nevertheless the NTFS3 development pace has been slowed down a bit for previous couple of months, its state is still the same as before: it is fully maintained and being developed.

And finally, we apologize for late reply; I allowed me short vacation after most restrictions because of covid ended up this month in Germany.

Thanks.

[root-core]

They have 200+ employees all over the world, the NTFS driver was not the majority of their profits; that would be their backup software.

They have like 20 employees according to reports of people working there.. they are like my neighbours.
No idea if this is only germany or worldwide tough.

NTFS-3G has corrupted all of my NTFS drives by updating stuff from Steam.
I have not tested NTFS3 enough to make a judgement, but it seems to be more stable.