SFC Considers Combining ZFS With Linux A GPL Violation

[chithanh]

This happens in Windows once in 10? years.

Driver breakage happened to me when I went from Windows 98 to Windows 2000.
It happened again when I went from Windows 2000 to Windows Vista. I could literally no longer use any of my peripherals except keyboard and mouse under Windows. Scanner, printer, webcam, DVB-T receiver, didn't work any longer. The printer has since received an updated driver, the other devices not.
Not strictly driver breakage, but when I wanted to upgrade from Windows 8.0 to 8.1, Microsoft had suddenly decided that CMPXCHG16b would be required on x64, which left my system totally out in the rain. Luckily, I had mostly stopped using Windows except for games at that point, so I went back to Vista which had longer security support remaining than Windows 8.0 (go figure).

The only instance when a piece of my hardware stopped being supported on a Linux upgrade was my PowerVR Kyro II, which depended on proprietary drivers that only worked with kernel 2.4 and not 2.6. I have bought only ATI/AMD graphics cards which work with open source drivers since.

[aht0]

Installed Ubuntu 16.04LTS for a quick peek. Installer did not appear to be offering ZFS, just usual Linux file systems. But module is there
zfs 2813952 0
zunicode 331776 1 zfs
zcommon 57344 1 zfs
znvpair 90112 2 zfs,zcommon
spl 102400 3 zfs,zcommon,znvpair
zavl 16384 1 zfs

[chithanh]

debianxfce As I said, I had mostly migrated away from Windows by that time, and did not use it any more than necessary for launching the stuff that didn't run on Linux. But I used it once in awhile and my experience did not match with what drSeehas claimed about 10 year driver API stability.

[starshipeleven]

This happens in Windows once in 10? years.
This happens in Linux every? year.

Oh look, another idiot. Let me repeat myself.
This happens in Windows once in 3 or so years. Major releases happen every 3 or so years. Windows is a frozen release model, you cannot update its kernel from the development branches.
Plus for some lucky devices it happens random whenever arrives a magic software update.

This happens in Linux every time you decide to change the kernel AND are using closed source drivers.
If you use opensource drivers you get only random regressions, which aren't terribly frequent, just like random breakage on windows.

The frequency this happens depends from the distro you have chosen and your hardware.

The big issue here isn't that linux does not have a stable ABI, the big issue here is that many things don't work at all in linux

Who is the idiot here?

you, of course.

No need to use harsh language. But the Linux community are so used to the constant kernel breakage, they dont see it as a problem. Only if you use an other OS, you will see the problem of the constant breakage.

I'm not "not seeing it as a problem", I'm not seeing the horrible mess of breakage you see on any of the linux systems I used, period.

I've updated kernel from 3.something to 4.4 and i've yet to see any breakage. Why? Because I'm using opensource drivers.

But then again I'm on Debian stable with backports.

On what distros you see this mess? Don't tell me crap you read from uninformed fools clickbaiting you into reading their nonsense just to show ads.

[bridgman]

Oh look, another idiot. Let me repeat myself.
This happens in Windows once in 3 or so years. Major releases happen every 3 or so years. Windows is a frozen release model, you cannot update its kernel from the development branches.

I think the reality is somewhere in between your two viewpoints - major Windows releases come along every few years but in many cases MS makes a major effort to ensure that drivers from the previous Windows generation are supported and validated prior to OS release, so you end up with the effect of a stable driver ABI for a lot more than three years. If you want the new OS features that are driver related you usually need drivers that work with the new ABI though.

I'll leave it up to you guys to figure out whether that means both of you are idiots or neither of you are

[chithanh]

MS makes a major effort to ensure that drivers from the previous Windows generation are supported and validated prior to OS release, so you end up with the effect of a stable driver ABI for a lot more than three years. If you want the new OS features that are driver related you usually need drivers that work with the new ABI though.

Even if the ABI doesn't change in a way that outright crashes the driver, they often stop working properly on OS upgrade.
At work we had some old PCs with NVidia FX5000 series cards (the dustbuster generation). After Windows Vista to Windows 7 upgrade, Aero became a slideshow with the Vista drivers. No Windows 7 drivers were available from NVidia. Aero worked fine on Vista, it was not a new feature.
More recently with Windows 8.1 to Windows 10, many things suddenly stopped working for users, until vendors provided updated drivers. Internet forums listed many complaints.

[kebabbert]

Oh look, another idiot. Let me repeat myself.
This happens in Windows once in 3 or so years. Major releases happen every 3 or so years. Windows is a frozen release model, you cannot update its kernel from the development branches.
Plus for some lucky devices it happens random whenever arrives a magic software update.

This happens in Linux every time you decide to change the kernel AND are using closed source drivers.
If you use opensource drivers you get only random regressions, which aren't terribly frequent, just like random breakage on windows.

The frequency this happens depends from the distro you have chosen and your hardware.

The big issue here isn't that linux does not have a stable ABI, the big issue here is that many things don't work at all in linux

you, of course.

I'm not "not seeing it as a problem", I'm not seeing the horrible mess of breakage you see on any of the linux systems I used, period.

I've updated kernel from 3.something to 4.4 and i've yet to see any breakage. Why? Because I'm using opensource drivers.

But then again I'm on Debian stable with backports.

On what distros you see this mess? Don't tell me crap you read from uninformed fools clickbaiting you into reading their nonsense just to show ads.

"On what distros I see this mess"? Well, it is not that _I_ see that. There are lot of other people seeing this mess too. Even Linux kernel developers:

http://www.phoronix.com/scan.php?page=news_item&px=Linux-Kernel-Stable-API-ABI

Artem Tahskinov writes:
"It's not a secret that there are two basic ways of running a Linux distribution on your hardware. Either you use a stable distro which has quite an outdated kernel release which might not support your hardware or you run the most recent stable version but you lose stability and you are prone to regressions. This problem can be solved by decoupling drivers from the kernel and supplying them separately so that you could enjoy stable kernel version X with brand new drivers like it's done in most other proprietary OS'es. I've been thinking of asking Linus about this decoupling for years already but I'm hesitant 'cause I'm 99.99999% sure he will downright reject this proposal."

Maybe you should inform he that he is doing it wrong, and he is an idiot?

[drSeehas]

... my experience did not match with what drSeehas claimed about 10 year driver API stability.

You noticed my question mark behind the 10?

[chithanh]

drSeehas
It is not even close to 10 years.
Even ignoring the Windows 8.1 CMPXCHG16b SNAFU, the breaks were Windows 2000 (2000), Vista (2006), 7 (2009) and 10 (2015).

[drSeehas]

... the breaks were Windows 2000 (2000), Vista (2006), 7 (2009) and 10 (2015).

So we have 6/3/6 years driver API stability. And now compare to Linux.