Linus Torvalds Comments On The NTFS Linux Driver Situation

[sinepgib]

They get $1000 not because of ReFS but because server licenses are a priori much more costly. And if ReFS is much better than NTFS then Microsoft is making its desktop OS less competitive, and it has so much money because its history is rooted in its desktop OS.

Less competitive compared to what? Their server OS that costs $1000? These server licenses need incentives. One may argue that simply authorizing a number of concurrent users or stuff like that would do, but I'm not sure that wouldn't lead to lawsuits. Having different feature sets? That creates an incentive.
If all versions ship all the capabilities, why would you go for the expensive license?

[cl333r]

Less competitive compared to what? Their server OS that costs $1000? These server licenses need incentives. One may argue that simply authorizing a number of concurrent users or stuff like that would do, but I'm not sure that wouldn't lead to lawsuits. Having different feature sets? That creates an incentive.
If all versions ship all the capabilities, why would you go for the expensive license?

Which is why I'm saying that gerrymandering your product to create artificial value is a long term dead end. Compared to Linux or whatever where you don't need a paid license at all if you want to and you still have ZFS/Btrfs or whatever.

[sinepgib]

Which is why I'm saying that gerrymandering your product to create artificial value is a long term dead end. Compared to Linux or whatever where you don't need a paid license at all if you want to and you still have ZFS/Btrfs or whatever.

It seems to work just fine. How many desktop users does Linux have and how many Windows?
Quick question: how many regular users even know what a filesystem is? That's what will define whether that's likely to affect a customer's choice or not. How much of the desktop market share is OEM as opposed to retail? How much of the consumer software tailors specifically Windows (and maybe Mac)? All those factor are much more likely to affect whether a non-server user will buy your desktop OS. Then, you need the extra incentive for the server, and that's all these features here and there and the removal of artificial caps, etc.

[cl333r]

It seems to work just fine. How many desktop users does Linux have and how many Windows?
Quick question: how many regular users even know what a filesystem is? That's what will define whether that's likely to affect a customer's choice or not. How much of the desktop market share is OEM as opposed to retail? How much of the consumer software tailors specifically Windows (and maybe Mac)? All those factor are much more likely to affect whether a non-server user will buy your desktop OS. Then, you need the extra incentive for the server, and that's all these features here and there and the removal of artificial caps, etc.

Almost all users pick this or that OS based on which one they worked with before. But given enough time and degradation the tide will eventually swing towards the better OS, rather the one they already know.

[sinepgib]

Almost all users pick this or that OS based on which one they worked with before. But given enough time and degradation the tide will eventually swing towards the better OS, rather the one they already know.

So they'll keep adding stuff in a delayed fashion as they've been doing for decades. The NT kernel was around since the days of Windows 3.1, yet it made it to consumers about 10 years later. NT was the pricey option for professional workstations (yes, at the time of 3.1 it was also a matter of it requiring more beefy hardware than most people could pay, but that wasn't the case when Windows 95 got released). NTFS was designed for it when the rest of their customers were still using FAT32.
Fun fact: Windows NT is younger than Linux and its degradation still doesn't catch up for desktop users. It has a lot to do with Linux not really being tailored to consumers commercially, which leads to a lack of investment. The best filesystem will never ever even be a factor if the desktop experience in general sucks (and it sucks for most users who need to run their proprietary software). It will simply keep getting market share in non-consumer areas where actually having the fastest and most stable filesystem is relevant. And that applies to many of the areas where Linux shines. E.g. 40Gbps packet processing? Most people barely have a 100Mbps internet connection, so they don't care.

[hoohoo]

MS failed with embrace, extend, extinguish. Is the new attack straight up "infiltrate"?

[Old Grouch]

For all the people talking about using exFAT as a common filesystem between Linux and Microsoft Windows, a slight problem is that, for example, with ext4 the only forbidden characters are '/', and the reserved file names '.' and '..', whereas in exFAT you can't use any of U+0000 (NUL) to U+001F (US), / (slash), \ (backslash), : (colon), * (asterisk), ? (question mark), " (quote), < (less than), > (greater than), and | (pipe).

It is a minor problem, unless you are doing an rsync from an ext4 filesystem to an exFAT filesystem, possibly for a backup, and you have files with forbidden characters. You can use the --iconv option, but it is not always transparent/reversible: it will convert a filename into something acceptable 'on the other side', but it doesn't necessarily convert back in the opposite direction.

I got bitten by this rsyncing a volume with filenames that contained colons to an exFAT external USB drive.

Wikipedia says the following for NTFS:

Allowed characters in filenames -

In Win32 namespace: any UTF-16 code unit (case-insensitive) except /\:*"?<>| as well as NUL[8] In POSIX namespace: any UTF-16 code unit (case-sensitive) except / as well as NUL Trailing spaces are not allowed and will be removed[9]

[khanich]

About the "maybe Russia-Ukraine connection thing":

Russia has a pretty big operation ongoing there. You don't plan such things in like a month. The decision to invade was easily done a year ago, but stuff like that takes time to actually result in observable effects (especially in peace time).

It could very well be that Paragon's devs got drafted and from a time pov it would actually be possible.
The Russian army started to amass troops on the Ukrainian border around the beginning of this year.
Depending on the situation in the country and outside, competency of the military leaders, earlier experiences etc., basic training can take a few weeks to many months.
So three months sounds reasonable, especially because they were not in a war at that point.
So this would pretty much fall in line with when they stopped responding.

[rabcor]

Some have hypothesized the lack of NTFS3 driver maintenance may be fallout from the Russia-Ukraine war with Russian developers involved. However, this lack of upstream bug fix submissions and responding to other developer inquiries dates back to Q4 last year, well before the conflict began.

Translation: Some idiots decided to mix politics that have nothing to do with open source in the slightest, with open source and ousted russian developers. And now we're in trouble because we kinda needed those developers and the whole point of open source is that it's free. For everyone. Even nazis if nazi germany was still a thing shouldn't be treated any different than anyone else in the open source community because if they're contributing, they're contributing, this is good for everyone. If they're ousted, they're not contributing anymore, bad for everyone.

If this is actually what's going on and not just a thoughtless comment by Michael, it's a great loss for open source, as it would mean we've taken a step in the direction away from 'open' source towards 'limited' source, just a few more steps and we're in proprietary territory

Or as some others said, I suppose they might have been drafted but I find that a bit doubtful. All of them drafted? I don't think the Russian government is quite that desperate for more troops, they have always had a very strong army, they're more or less unbeaten even since ww2. They didn't lose that war for military reasons, but economical ones.

[tuxd3v]

Translation: Some idiots decided to mix politics that have nothing to do with open source in the slightest, with open source and ousted russian developers. And now we're in trouble because we kinda needed those developers and the whole point of open source is that it's free. For everyone. Even nazis if nazi germany was still a thing shouldn't be treated any different than anyone else in the open source community because if they're contributing, they're contributing, this is good for everyone. If they're ousted, they're not contributing anymore, bad for everyone.

I din't wanted to touch on the subject, but I am afraid that..you nailed it!
We may very well start to have nazism attitudes in the Linux kernel development now..

Or as some others said, I suppose they might have been drafted but I find that a bit doubtful.

A lot of people consume a lot of propaganda, they are easy targets...
No there are no emergency draft in Russia.
And by the Way, there are no Drafted military in the operation, only professional soldiers..