Approved: Fedora 33 Desktop Variants Defaulting To Btrfs File-System

[k1e0x]

Of cource it provides "zfs set copies=N <volume>"
https://docs.oracle.com/cd/E19253-01...vpg/index.html
Try to find something like that on any Linux FS

Hey one thing I've noticed. Don't link to the Oracle docs, because they are often out of date or just wrong or don't have the features of OpenZFS. If you're doing a search try adding "ZFS FreeBSD" to it - Usually that gives good, up to date and proper docs for OpenZFS.

The Oracle docs are all over and they are not always right.

[k1e0x]

Hea, it's starshipeleven Here to tell me how irelvent and wrong I am. Go check this out. https://www.phoronix.com/forums/foru...58#post1193958

[starshipeleven]

Thank you for the confirmation that only way to have dedup on btrfs is offline

I guess when I run duperemove on my mounted btrfs it's just lying to me. Is it all in my head, none of this is real, right? (it should also work on xfs, never used it on xfs though)
https://github.com/markfasheh/duperemove

Here the btrfs wiki also explains more and lists some other tools to do that on a MOUNTED btrfs https://btrfs.wiki.kernel.org/index.php/Deduplication

[starshipeleven]

The same thing is wrong with BTRFS that is wrong with most of Linux today.

It's hackish and it has no core direction.

Of all good reasons to shit on btrfs you always come up with nonsense from your own imagination.

Btrfs was routinely shouted at because it is breaking the mold by not integrating with many kernel subsystems, and you decide to claim it does that, like other linux kernel features.

I can give you an example in Linux containers.

Completely unrelated, also wrong.

BTRFS reuses and recycles a ton of things

Like for example? It's not using MD for RAID, it's not using LVM for volume management. How about you tell us what is it "reusing"?

You can look at it like food.

More unrelated examples.

[Britoid]

because zfs doesn't provide redundancy on single device(btrfs does)
i don't care about license. and i mean it, i'm not freebsd idiot who hates gpl. but i wouldn't use some unsupported out of tree crap and license has indirect relation to that. but even if zfs was relicensed by oracle and merged to linux, i'd have no use for it, because btrfs is clearly superior. just like i have no use for xfs, it's ridiculous to even consider using non-resizeable fs

XFS can resize up, but not down.

[starshipeleven]

Regarding vi and nano, I disagree.

I guess the vi I'm accustomed to is busybox's vi that is more basic than that.

Then again, the only usecase I have in mind for a terminal text editor isn't writing latex, but editing config files on a headless system, if you use it to do more you are a neckbeard just like the people using EMACS. So that's probably where the disagreement starts.

I can do most of that anyway in the same timescales by opening multiple terminals (or ssh sessions) and copy-paste, and I don't need to memorize a set of commands to control a text editor that I'm only using sparingly, which is for me far more important.

For any important text editing job I'm using a GUI text editor or even Libreoffice, because I'm not a savage, we are in 21st century. And no, terminal vs gui speed is irrelevant, most of the time is wasted by me thinking about what to write anyway.

numasan this post is also for you

[k1e0x]

because zfs doesn't provide redundancy on single device(btrfs does)
i don't care about license. and i mean it, i'm not freebsd idiot who hates gpl. but i wouldn't use some unsupported out of tree crap and license has indirect relation to that. but even if zfs was relicensed by oracle and merged to linux, i'd have no use for it, because btrfs is clearly superior. just like i have no use for xfs, it's ridiculous to even consider using non-resizeable fs

Oracle doesn't own OpenZFS. The contributors hold their own copyrights. Oracle would need to get their permission to change it. (Good luck! lol Most of them are former Sun engineers that wrote the dammed thing who would rather delete it than give it to Larry.) The Linux Kernel works the same way btw. Oracle does have OracleZFS but that version was never ported, hense Oracle has no version of ZFS that could run on Linux even if they wanted it to. The ported code is copyright to OpenZFS. "Thanks former Sun dudes!" (and FreeBSD dude's/ets who did it first)

The problem really isn't with the CDDL either.. The CDDL states it can be mixed with other licences. (and often is in other OS's) .. The problem is the GPL's odd definition of works.. it's the GPL that is incompatible.. not the other way around. but.. that has shown to be a non issue because even if the licences are on paper incompatible, one can't file a lawsuit because you can comply with both licences at the same time without causing any harm to either of them.. therefore there is no case.

ZFS has also taken the position of being a multi-os filesystem.. supporting many platforms. So the fact that it isn't in the Linux kernel isn't really a big deal because it's not in MacOS's or Windows kernel either. (Rumour is mac's can even boot of it and use it as root, even with all the weird things apple does on file systems, so the support is very good there.)

Distros can bundle it and support it.. just like Ubuntu has done. Lots of things are done this way.. Do you own an Nvidia card? That is how it's driver works.

And pal666, I work in IT. Trust me, I know Linux a lot better than most and am *well* aware of where all it's warts and scars are.. I deal with them daily.

[kloczek]

all of them are out-of-band, i.e. you run some program and it does dedup. ones which use btrfs-specific functionality(there are some cross-fs dedupers), are very efficient. btrfs send/receive provides only newly written extents and btrfs keeps all page checksums, so there's no need to do rescans or checksum calculations etc

Thank you for the confirmation that only way to have dedup on btrfs is offline

[k1e0x]

btrfs send/receive provides only newly written extents and btrfs keeps all page checksums, so there's no need to do rescans or checksum calculations etc

: face palm :

ZFS does the same thing. I wonder where BTRFS got the idea. heh..

[k1e0x]

because zfs doesn't provide redundancy on single device(btrfs does)

??? So what you are saying is.. if I format my disk with BTRFS, and then I smash it to bits with a hammer.. BTRFS still has my files? Neat! Do I just set the broken bits in the bottom of the case to use this feature?

No file system provides real redundancy on a single device, if you believe that you are a fool.... however.. ZFS has a copies feature flag you can set on a dataset so it will keep X number of duplicate blocks in different locations on the disk... is that what you are talking about in btrfs? ZFS has that ability too. (not that I would rely on it... but it might be useful if you were paranoid)