OpenZFS 2.3-rc3 Adds JSON Output For Commonly Used Commands

[phoronix]

Phoronix: OpenZFS 2.3-rc3 Adds JSON Output For Commonly Used Commands

OpenZFS 2.3 continues working its way toward release as a big step forward for this open-source ZFS file-system implementation for Linux and FreeBSD systems...

https://www.phoronix.com/news/OpenZFS-2.3-rc3

[uid313]

Very nice!

This could be very useful for making things like a GTK application or a web interface. Maybe a web UI for a NAS. I hope Btrfs and Bcachefs also adds JSON output.

[Joe2021]

Should become a best practice for most cli tools...

[unwind-protect]

Should become a best practice for most cli tools...

In FreeBSD many already do. This is where this comes from.

/goes back to parsing old zpool output in Python...

[caligula]

It's actually funny how json also became a thing in command line. 55 years of Unix without any innovations in the department of easy to parse interchange formats.

[lsatenstein]

As a side note, "Can you please grammar check your posts as a general "pre-publish post" Check my bold text for insertions."

Example1,
The fast dedup (deduplication) support in OpenZFS 2.3 is a "major performance upgrade" compared to how the existing OpenZFS handles deduplication of file-system contents.

Example2.
OpenZFS 2.3 also incorporates many bug fixes, Linux 6.11 kernel compatibility, such as support for longer file names/directories of up to 1,023 characters, and other changes.

#I used "incorporates" to replace "brings". The word "brings" does not mean implemented or incorporated.

I realize that there are time constraints to produce contributions for posts, but your posts have international readership, and man are non-fluent English technical individuals.

[ahrs]

Very nice!

This could be very useful for making things like a GTK application or a web interface. Maybe a web UI for a NAS. I hope Btrfs and Bcachefs also adds JSON output.

BTRFS already has JSON output:

$ btrfs --format json fi df /
{
  "__header": {
    "version": "1"
  },
  "filesystem-df": [
    {
      "bg-type": "Data",
      "bg-profile": "single",
      "total": 1934663614464,
      "used": 1865146261504
    },
    {
      "bg-type": "System",
      "bg-profile": "single",
      "total": 33554432,
      "used": 294912
    },
    {
      "bg-type": "Metadata",
      "bg-profile": "single",
      "total": 64625836032,
      "used": 39429193728
    },
    {
      "bg-type": "GlobalReserve",
      "bg-profile": "single",
      "total": 536870912,
      "used": 0
    }
  ]
}​

(Yes, my filesystem is pretty full)

But it supports it inconsistently. Some commands have no JSON output at all:

$ btrfs --format json fi usage /
ERROR: output format json is unsupported for this command​

[stormcrow]

As a side note, "Can you please grammar check your posts as a general "pre-publish post" Check my bold text for insertions."

Example1,
The fast dedup (deduplication) support in OpenZFS 2.3 is a "major performance upgrade" compared to how the existing OpenZFS handles deduplication of file-system contents.

Example2.
OpenZFS 2.3 also incorporates many bug fixes, Linux 6.11 kernel compatibility, such as support for longer file names/directories of up to 1,023 characters, and other changes.

#I used "incorporates" to replace "brings". The word "brings" does not mean implemented or incorporated.

I realize that there are time constraints to produce contributions for posts, but your posts have international readership, and man are non-fluent English technical individuals.

Good luck. I don't believe Michael even uses a spell checker, let alone a grammar checker, despite the ubiquity of the first in nearly every text editor or processor while the latter is almost as easily obtained.

[archkde]

It's actually funny how json also became a thing in command line. 55 years of Unix without any innovations in the department of easy to parse interchange formats.

JSON is not that bad too parse and there are libraries for nearly all languages. Or do you mean that it's funny that this didn't happen before JSON even though there have been other easy to parse formats before?

[skeevy420]

As a side note, "Can you please grammar check your posts as a general "pre-publish post" Check my bold text for insertions."

Example1,
The fast dedup (deduplication) support in OpenZFS 2.3 is a "major performance upgrade" compared to how the existing OpenZFS handles deduplication of file-system contents.

Example2.
OpenZFS 2.3 also incorporates many bug fixes, Linux 6.11 kernel compatibility, such as support for longer file names/directories of up to 1,023 characters, and other changes.

#I used "incorporates" to replace "brings". The word "brings" does not mean implemented or incorporated.

I realize that there are time constraints to produce contributions for posts, but your posts have international readership, and man are non-fluent English technical individuals.

The lack of an editor and not using basic writing tools are the complaints I read the most in the comments. To me, the continual typos and grammar issues detract from how professional Phoronix appears. It doesn't appear professional when articles give off vibes that they were written using Notepad on Windows 3.1.

I come across typos and grammar issues in nearly article, and I feel like an asshole saying this out loud, but I used to report them all the time while nowadays I just don't bother reporting them unless I find them humorous. I have a bit of a tick where I can't shut off being a Grammar Not-See once I start seeing them ... like your ", and man are non-fluent" remove the comma before "and" and "man" should have been "many" ... but because of that I try damn hard to ignore typos and grammar issues unless they're that silly and funny to me ... like a post about grammar issues and typos having grammar issues and typos.