> The “Corresponding Source” for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities.
This is disappointing, time to migrate from yet another was-open-source-now-only-technically-open-source software.
At work the trend is only adopt open source on a paid plan, adopting or relying on open source otherwise is just waiting for the rug to be pulled. Essentially treat open source as closed source.
Oftentimes the alternative is migrating to one dying or unstable zero-funding open source project to the next.
Which path you go down depends on a lot of factors of that particular piece of software.
Can I trust FreeCORE to be a stable, secure project that hangs on to my NAS data if it’s an unfunded fork? Or would it be better to give up some openness for something like TrueNAS SCALE?
When it comes to stuff like Proxmox and TrueNAS I think a lot of what keeps those good products for homelab use is the fact that they have commercial customers paying for the development. I can count on TrueNAS being around and getting updates because they have a proven revenue stream.
That said, it’s a balance, because sometimes that commercialization becomes a rug pull. A lot of the viability of those projects for homelab use revolves around the companies that put those products out recognizing that they have a customer base that will not pay and accepting that fact without deciding to try to squeeze out every last person who just might pay by compromising the free edition.
I like the way Proxmox and OpnSense handle this: you get the community patch channels which are “not for production use” which scares the corporate businesses into paying but has zero negative impact for home users.
To be clear I’m also not saying FreeCORE specifically is a fly-by-night project or unstable or anything. I’m genuinely unfamiliar with it. I’m just saying, sometimes projects like that end up not being the right choice for certain use cases.
Regardless the estimated quality of the new thing, there is a point and a value in penalizing the old thing when it fails to adhere to principles.
One must have principles onesself to see this as imortant though.
A lot of people use words like "practical" or "pragmatic" to avoid inconvenience.
I've already been using zVault for a year which is a fork of the last truenas core, but it never got any further updates and it's about to become impossible to run up to date software in jails on the old host. The pkg repos that support the host version are about to go eol and the binaries will be gone. So I've already gone through the "what if this doesn't go anywhere?"
My point: It's a pain but I won't relieve that pain by running trunas scale.
I'd rather try switching to xigmanas or dump the whole concept of the front end and just run plain freebsd and learn a few commands to manage storage and file services directly and a maybe a seperate dedicated jail manager like bastille. At least that way I can never be stuck in this position again where I can't update my own software because someone else stopped updating their software.
I totally understand why you’ve made those choices. It just might be worth questioning how much of your own time these principles are worth.
This entire time you’ve been swapping between NAS distros fiddling with your NAS box my TrueNAS SCALE installation has required approximately zero of my time.
Did they violate some principles? I guess so; they deprecated build scripts, but the source code of the product is still available. The practical impact to the end user is effectively zero. I’m not even sure I would personally need the source code to be available when it comes to a NAS solution — I did consider unraid as a potential alternative before settling on TrueNAS.
But, at this point, AI makes code cheap enough that I would just fork the open source software and maintain it, instead of paying for a third party. Heck, even with saas, I may clone internally.
Knowing a little about the maintainers and their model, this isn’t a private-equity cash grab or Mr Monopoly twirling his moustache; it’s a business trying to stay sustainable in open source. There are no billionaires here: they staff developers who contribute to projects like OpenZFS, and developers who build enterprise features that fund all of it. Their business is being challenged by “global” shops taking advantage of the platforms openness - and this isn’t the virtuous or even legally tenable type of competition. It’s “violating terms and not legally accessible” competition, and they are taking measures. It sucks for everyone.
If TrueNAS/iX whatever turns around and sells to Bending Spoons, then I’ll of course owe you a beer. Until then, their leadership is just trying keep the project alive - and funding other open source efforts at the same time.
Why is it not enough to give you all of their code for free? Why do they also have obligations and expectations beyond that? The entitlement never ends!
> At work the trend is only adopt open source on a paid plan, adopting or relying on open source otherwise is just waiting for the rug to be pulled.
"Won't do free labour for our commercial enterprise to profit off of forever" is a rug pull now? It's incredible that you can type these words with a straight face.
> "Won't do free labour for our commercial enterprise to profit off of forever" is a rug pull now? It's incredible that you can type these words with a straight face.
I explicitly stated in my comment that we pay for open source (via vendor support). That is not the issue.
It is now a common business strategy to commercialise an open source project. FreeNAS was commercialised as TrueNAS, which is now having some of the source pulled.
TrueNAS doesn’t exist in a vacuum. It’s built on source projects.
> I explicitly stated in my comment that we pay for open source (via vendor support). That is not the issue.
Your statements clearly took issue with the fact that your company has to pay open-source devs whose work it uses for reliability. You presented the options as "pay them, or else get rugpulled", and then went on to say "how did we end up in this state of affairs [where we must pay to not get rugpulled]". There is no possible way to interpret your comment other than lamenting the "state of affairs".
> TrueNAS doesn’t exist in a vacuum. It’s built on source projects.
As long as it isn't violating the license of those projects, they still don't owe anyone anything. If it is violating the license, complain about it on those terms. But your complaint was about open-source in general, not specifically only this project.
They're allowed to stop publishing new work whenever they want to. That's part of the free, open-source software game, wherein: We're all free to provide as much as we wish, and also free to consume as much as we want. Nobody necessarily owes anyone else anything and it's really very beautiful in these ways. This is enabled in large part due to the very low cost of distributing copies of intangible software (which can be very close to zero).
But when money gets involved, then things become different. Contracts can be had. Support requirements can be determined. If a contract requires 5 years of support, then it can be reasonably expected for support to continue for at least 5 years.
It's OK to want that, and to never use software that doesn't involve money. It's also OK for an author/company/org to not want that, and to never agree to those terms.
These are different goals, but one party doing it one way doesn't detract from another party doing it some other way. It's all OK.
They do not give all of the code. If you have an object, and don't have everything needed to reproduce the object, then you have not been given the source to the object.
And with BSD licenced code this is their right.
And it's my and anyone else's right to observe all of this and criticize it.
Likewise, it is my right to criticise you for exercising your right to criticise them. Nobody said you aren't allowed to, but just as anyone is allowed to whine about not being given enough free stuff, I'm allowed to call out that they're acting like an entitled child in public. Just because you can legally do something doesn't mean that it is socially acceptable to do so.
> They chose to base their enterprise on the work of others, only to pull the ladder up behind them.
This is literally why BSD exists in the first place. To allow companies compete freedom with the code, including freedom to not release changes.
I hate to break it to you: basically every commercial storage vendor on the planet bases their appliances on either BSD or Linux and not a single one is sharing their proprietary code with you. There’s nothing shameful about it, they built a company to make money.
You didn't deserve to be downvoted to hell and back. People deserve to be paid for their work, and without copyright law only mega-cap companies would ever be able to capture enough value from their work to survive.
People act like you're saying Disney should be able to hold a copyright on Mickey Mouse for over 100 years.
Calling it a "rug pull" when the creator of an open source project decides to stop giving away future work for free is a delusional level of self-entitlement.
(1) The previous releases are still open source. If you don't like it, fork the last free release and maintain it yourself.
(2) People deserve whatever the market is willing to pay for their work. See above.
Frankly, if D. Richard Hipp decided to stop giving away future versions of SQLite for free, I would be the first person to congratulate him for finally having the courage to put his foot down and say "my fellow committers and I deserve at least a few million dollars for 20 years of work making something that billions of machines rely upon."
Open-source is an inherently radical anti-capitalist philosophy. In your eyes everyone "deserves" what the market is willing to pay, but open-source developers would rather donate their time and energy to create or contribute to something that is freely accessible to anyone and everyone. Also, believe it or not, but some people are not motivated by money, and actually derive value from the process of creating software.
The issue with rug-pulls like this is that they are deceptively masquerading with this philosophy as a calculated business strategy to adopt users. Then in this case they start using deceptive tactics like removing build files so they can continue appearing to be fully open source, even though they are violating the spirit of that term. Finally, they drop the facade when it is deemed valuable enough. Even if you are a strict capitalist, there is nothing virtuous or meritocratic about these kinds of business strategies, they rely on tricky, concealed intentions, and rightfully deserve criticism.
I'll also add that there is nothing "courageous" about deciding to extract more value from your labor at the expense of public accessibility. True courage is found in acts of selfless sacrifice.
What a shame I didn't discover this sooner -- I finally sent my 11-year-old TrueNAS system[1] through the one-way Linux door about a year ago because I gave up on any hope that something like this would come along and I didn't see another way to keep it current.
The migration path went smoothly but I preferred the FreeBSD setup if only because it forced me to do something a bit more interesting in my spare time. I do quite a lot of Linux work in my day job and it was nice to occasionally work on something that was surface-level similar but subtly and not-so-subtly different.
[1] Over 11 years I have replaced every component (even the CPU/motherboard) at least once except for the case. Migrating from MBR to UEFI was "fun".
FreeBSD with ZFS tends to be less chaotic once setup. Linux distro in storage appliances can be hit or miss, but ZFS/F2FS/CephFS for /home and ext4+Overlayfs for OS / are reliable these days.
Usually, for JBOD file archiving people want to avoid cutting-edge anything. Sometimes Linux kernel less-popular fs updates add or remove data-corrupting fringe filesystem bugs and features. People usually discover this on their own after the first SSD/HDD starts to wear-out, upgrade breaks compatibility, and the solution is re-install/copy-over everything.
Accordingly, I was surprised people now deploy btrfs on linux / for some distro, as your options when things go awry was limited at one time. Would be cool if it was finally production ready, and actually reliable on degraded hardware.
Too late for me, started my own mini vanity NAS/automated-testing project last year with a few built in features similar to https://guacamole.apache.org and Proxmox. If the OS build stops sucking over winter break, than I may share the build-script someplace for people to pull off parts they find useful, and a USB-bootable portable rtx GPU image for test-drive/recovery scenarios. =3
Bigger issue was that native encryption doesn’t (didn’t?) have a maintainer, so even though most bugs got fixed, it’s not a great thing to build on. Some of the key storage is also a bit outdated.
IMO use LUKS unless you really need the features of ZFS native enc.
Part of the problem is replicating fs bugs that take months/years to occur is nearly impossible... people still sell USB to SAS/SATA adapters that cause large file corruption in hardware too.
ZFS does have bit-rot checksum detection/repair, but the encrypted file feature was a bit of a YOLO feature for awhile. Not sure where it is at these days, as ddr5 ram prices skew away most of ZFS best key features. =3
In ~2017 I ditched FreeNAS when my FreeNAS Mini died from the Atom C2000 hardware failure. iXSystems was great about getting me a replacement board, but I decided when I rebuilt the system to just use FreeBSD and get used to command-line only administration. That directly lead to me learning pf, jails, bhyve, etc. Yeah, I don't have any fancy web UI, but I've had a rock solid system for almost 10 years with Samba + NFS running without issue.
I've been using things like Samba and NFS since the 1990s. They're not hard to figure out with useful resolution.
When I finally became determined to give ZFS a serious go around 2016, I looked at all of my options -- including throwing some hardware together for FreeNAS, and running FreeNAS in a VM.
I decided to just use an existing Linux box that was already doing storage stuff and add ZFS to it.
It's been fine ever since. Keeping ZFS going isn't deep voodoo at all; of all the things this box does, ZFS has probably been the most hands-off aspect of it.
This is what I do, to try out ZFS in a "safe" and easy way. An HP Microserver (my home server) with a USB3 disk attached formatted as ZFS. The server still runs Debian stable and uses DKMS ZFS. The recent pain points have only been the time taken to rebuild the kernel module for kernel updates and there have been a lot of them over the past few months (due to all the CVE's found and fixed). The ZFS side has been plain sailing so far but if I hit any ZFS issues I may have trouble (but, as I say, it's a test anyway).
That's a great test environment and it really helps drive away some of the misconceptions about ZFS.
It's a lovely filesystem and it can work in very stodgy ways with zetabytes on Serious Hardware, but it can also work with relatively small devices on silly hardware.
Another example: I've got a ZFS scratch disk that has seen dozens of terabytes written. It's an old ~120GB Intel-branded SATA SSD that is connected with a very cheap SATA to USB3 adapter cable. I've been using it this way for years.
All of the Internet lore has told me that this must result in resolute failure. Whether because USB is broadly not a serious data bus, or because ZFS doesn't like non-serious things, or because 1 copy is 0 copies, or because SSDs are fragile creatures that hate write cycles, or because I've created a portal to hell, or something else: The lore states that it must result in failure, and usually stipulates that ZFS would amplify this concern.
But it's been great. Really. Zero issues -- except when I unplugged it on accident while it was hot...but even then, it came back fine. (Of course it did. ZFS is intended to be robust in the face of unplanned failures regardless of cause.)
And because ZFS: If it ever has any issues, then they'll be reported and logged either in the normal course of doing things, or during a periodic scrub. Every read is checksummed (even in normal operation), and every discrepancy is reported. It's been clean.
It is my intent to keep it as a scratch disk until it becomes problematic, and my further hope is to be able to analyze those problems once they occur.
It won't last forever, but it's lasted for quite a long time so far. (A little bit of inquiry with smartctl and some math just now shows that it has endured writes equivalent to ~43,000% of its capacity.)
Yeah, I just run bare Debian myself. It's really not hard to setup LVM and Samba. And all of the major frontier models are well-versed in those things, so I used their help to get all of the various bits setup to make access from MacOS fast.
Why does anyone bother with these ephemeral meme distros when you can accomplish the same with vanilla FreeBSD/Illumos/Linux using standard commands and documentation?
If you need your hand held to file share, that's why Windows Server still exists and is the gold standard for SMB.
Because I don't want to spend that much time an all systems i'm running?
you also get a whole community testing the stack, fixing it too.
And don't be that pissi... when I started 20 years ago, stuff like this would have been the gateway to linux world and is for sure still today for plenty of people...
If you have coworkers or other priorities it's sometimes preferable to have the option to have a hand held at some tasks or incidents rather than making sure someone competent read all the documentation pertaining to some irreversible command lines. You can try to treat a LLM as that hand holding but it can be a psychophantic member of group think while a checkbox stays a stable checkbox.
I used TrueNAS in the past but last touched it several years ago.
I gather now there are two flavors: CORE which continues to use FreeBSD under the hood, and SCALE (aka Community Edition) which is based on Debian Linux. If I understand correctly, the former is only receiving security and stability updates, and iXsystems is developing all new features only on SCALE.
If you were to install fresh today, would you prefer this (FreeCORE), BSDnas, or the official SCALE version - and why?
I migrated from TrueNAS CORE to SCALE, and found that Linux had better performance. I don't see why you'd use FreeBSD if you don't have a very specific reason (hobby, jails, etc)
And for those still confused, CORE is FreeBSD-based and has been discontinued, while SCALE is Debian-based and the version that will be developed by TrueNAS going forwards.
You can integrate a fleet of Linux storage servers with a collection of K8s clusters more easily than a you can BSD storage servers. Getting BSD and Linux talking smoothly needs Real Unix Skills.
Colleagues in K8s shops tell me that their k8s guys come for help with complex tasks like.... deleting files. They know to construct and deploy containers, but not how to delete a file without using a GUI (meaning Windows or at best WSL).
I am told that some companies are now demanding "bash" among the list of skills for would-be container-herders. And here, "bash skills" means you should know to to create, edit, and then delete a file -- over an ssh connection.
Nah. Doesn't really matter when you are building the hardware to do that job. And in any event, FreeBSD is primarily a server OS, and its support is quite good in that department.
Example: OpenBSD and NetBSD both will install X11 and a window manager by default from the installer, if you want. FreeBSD doesn't. They're working on it, but they've been working on it for years.
I did the migration to Linux and it actually was painless. I didn't really notice any degradation or hiccups in any way. But I guess this is a good option if FreeBSD is important to you.
Who is behind this? I've been burned once with nas, and I watched the long Plex to Jellyfin saga. What assurance do I have that this will continue to be developed and will become good?
This seems nice. FreeBSD always seemed nicer than Linux, though now that Linux is being developed by LLMs and Torvalds is vibe-coding, I wonder if they'll be able to keep up.
As far as I understand, Torvalds said that he is not going to oppose all AI-generated improvements to Linux, not that he's vibe coding new kernel modules for fun.
Linux is definitely going to speed up as it starts to get vibe coded. In a few years, I bet that it's all going to be built with agentic loops, and nobody will be looking at the code at all.
What do you call people who vibe code? When he says that he sees a 10x improvement in productivity with AI code gen, what makes you think he's going to put the tool down? Do you think he's stupid or something?
He'll keep ramping up his vibe coding, because he can see it's the future.
[0] https://forums.truenas.com/t/clearing-the-air-on-build-scrip...
> The “Corresponding Source” for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities.
At work the trend is only adopt open source on a paid plan, adopting or relying on open source otherwise is just waiting for the rug to be pulled. Essentially treat open source as closed source.
How did we end up with this state of affairs?
Which path you go down depends on a lot of factors of that particular piece of software.
Can I trust FreeCORE to be a stable, secure project that hangs on to my NAS data if it’s an unfunded fork? Or would it be better to give up some openness for something like TrueNAS SCALE?
When it comes to stuff like Proxmox and TrueNAS I think a lot of what keeps those good products for homelab use is the fact that they have commercial customers paying for the development. I can count on TrueNAS being around and getting updates because they have a proven revenue stream.
That said, it’s a balance, because sometimes that commercialization becomes a rug pull. A lot of the viability of those projects for homelab use revolves around the companies that put those products out recognizing that they have a customer base that will not pay and accepting that fact without deciding to try to squeeze out every last person who just might pay by compromising the free edition.
I like the way Proxmox and OpnSense handle this: you get the community patch channels which are “not for production use” which scares the corporate businesses into paying but has zero negative impact for home users.
To be clear I’m also not saying FreeCORE specifically is a fly-by-night project or unstable or anything. I’m genuinely unfamiliar with it. I’m just saying, sometimes projects like that end up not being the right choice for certain use cases.
One must have principles onesself to see this as imortant though.
A lot of people use words like "practical" or "pragmatic" to avoid inconvenience.
I've already been using zVault for a year which is a fork of the last truenas core, but it never got any further updates and it's about to become impossible to run up to date software in jails on the old host. The pkg repos that support the host version are about to go eol and the binaries will be gone. So I've already gone through the "what if this doesn't go anywhere?"
My point: It's a pain but I won't relieve that pain by running trunas scale.
I'd rather try switching to xigmanas or dump the whole concept of the front end and just run plain freebsd and learn a few commands to manage storage and file services directly and a maybe a seperate dedicated jail manager like bastille. At least that way I can never be stuck in this position again where I can't update my own software because someone else stopped updating their software.
This entire time you’ve been swapping between NAS distros fiddling with your NAS box my TrueNAS SCALE installation has required approximately zero of my time.
Did they violate some principles? I guess so; they deprecated build scripts, but the source code of the product is still available. The practical impact to the end user is effectively zero. I’m not even sure I would personally need the source code to be available when it comes to a NAS solution — I did consider unraid as a potential alternative before settling on TrueNAS.
But, at this point, AI makes code cheap enough that I would just fork the open source software and maintain it, instead of paying for a third party. Heck, even with saas, I may clone internally.
If TrueNAS/iX whatever turns around and sells to Bending Spoons, then I’ll of course owe you a beer. Until then, their leadership is just trying keep the project alive - and funding other open source efforts at the same time.
Maybe AI fucks all of this up, who knows ….
> At work the trend is only adopt open source on a paid plan, adopting or relying on open source otherwise is just waiting for the rug to be pulled.
"Won't do free labour for our commercial enterprise to profit off of forever" is a rug pull now? It's incredible that you can type these words with a straight face.
I explicitly stated in my comment that we pay for open source (via vendor support). That is not the issue.
It is now a common business strategy to commercialise an open source project. FreeNAS was commercialised as TrueNAS, which is now having some of the source pulled.
TrueNAS doesn’t exist in a vacuum. It’s built on source projects.
Your statements clearly took issue with the fact that your company has to pay open-source devs whose work it uses for reliability. You presented the options as "pay them, or else get rugpulled", and then went on to say "how did we end up in this state of affairs [where we must pay to not get rugpulled]". There is no possible way to interpret your comment other than lamenting the "state of affairs".
> TrueNAS doesn’t exist in a vacuum. It’s built on source projects.
As long as it isn't violating the license of those projects, they still don't owe anyone anything. If it is violating the license, complain about it on those terms. But your complaint was about open-source in general, not specifically only this project.
They're allowed to stop publishing new work whenever they want to. That's part of the free, open-source software game, wherein: We're all free to provide as much as we wish, and also free to consume as much as we want. Nobody necessarily owes anyone else anything and it's really very beautiful in these ways. This is enabled in large part due to the very low cost of distributing copies of intangible software (which can be very close to zero).
But when money gets involved, then things become different. Contracts can be had. Support requirements can be determined. If a contract requires 5 years of support, then it can be reasonably expected for support to continue for at least 5 years.
It's OK to want that, and to never use software that doesn't involve money. It's also OK for an author/company/org to not want that, and to never agree to those terms.
These are different goals, but one party doing it one way doesn't detract from another party doing it some other way. It's all OK.
And with BSD licenced code this is their right.
And it's my and anyone else's right to observe all of this and criticize it.
Fuck that noise. They are the ones breaking the social contract, not the folks criticizing them for it.
Shameful of you.
This is literally why BSD exists in the first place. To allow companies compete freedom with the code, including freedom to not release changes.
I hate to break it to you: basically every commercial storage vendor on the planet bases their appliances on either BSD or Linux and not a single one is sharing their proprietary code with you. There’s nothing shameful about it, they built a company to make money.
People act like you're saying Disney should be able to hold a copyright on Mickey Mouse for over 100 years.
Calling it a "rug pull" when the creator of an open source project decides to stop giving away future work for free is a delusional level of self-entitlement.
(1) The previous releases are still open source. If you don't like it, fork the last free release and maintain it yourself.
(2) People deserve whatever the market is willing to pay for their work. See above.
Frankly, if D. Richard Hipp decided to stop giving away future versions of SQLite for free, I would be the first person to congratulate him for finally having the courage to put his foot down and say "my fellow committers and I deserve at least a few million dollars for 20 years of work making something that billions of machines rely upon."
The issue with rug-pulls like this is that they are deceptively masquerading with this philosophy as a calculated business strategy to adopt users. Then in this case they start using deceptive tactics like removing build files so they can continue appearing to be fully open source, even though they are violating the spirit of that term. Finally, they drop the facade when it is deemed valuable enough. Even if you are a strict capitalist, there is nothing virtuous or meritocratic about these kinds of business strategies, they rely on tricky, concealed intentions, and rightfully deserve criticism.
I'll also add that there is nothing "courageous" about deciding to extract more value from your labor at the expense of public accessibility. True courage is found in acts of selfless sacrifice.
It's like those SOHO router vendors that release incomprehensible tarballs of their GPL components; lotsa luck building it.
If I'm paying for closed-source ZFS, I'll go with the real thing and buy a Solaris license.
I think scale is also BSD licensed (atleast the gui and middleware)
Lets hope FreeCORE will make it longer.
Some comparison between Linux based TrueNAS SCALE and FreeBSD based TrueNAS CORE here:
- https://vermaden.wordpress.com/2024/04/20/truenas-core-versu...
Regards,
vermaden
The migration path went smoothly but I preferred the FreeBSD setup if only because it forced me to do something a bit more interesting in my spare time. I do quite a lot of Linux work in my day job and it was nice to occasionally work on something that was surface-level similar but subtly and not-so-subtly different.
[1] Over 11 years I have replaced every component (even the CPU/motherboard) at least once except for the case. Migrating from MBR to UEFI was "fun".
Usually, for JBOD file archiving people want to avoid cutting-edge anything. Sometimes Linux kernel less-popular fs updates add or remove data-corrupting fringe filesystem bugs and features. People usually discover this on their own after the first SSD/HDD starts to wear-out, upgrade breaks compatibility, and the solution is re-install/copy-over everything.
Accordingly, I was surprised people now deploy btrfs on linux / for some distro, as your options when things go awry was limited at one time. Would be cool if it was finally production ready, and actually reliable on degraded hardware.
Too late for me, started my own mini vanity NAS/automated-testing project last year with a few built in features similar to https://guacamole.apache.org and Proxmox. If the OS build stops sucking over winter break, than I may share the build-script someplace for people to pull off parts they find useful, and a USB-bootable portable rtx GPU image for test-drive/recovery scenarios. =3
This was a while ago so it may have changed but then changes can introduce new bugs...
IMO use LUKS unless you really need the features of ZFS native enc.
ZFS does have bit-rot checksum detection/repair, but the encrypted file feature was a bit of a YOLO feature for awhile. Not sure where it is at these days, as ddr5 ram prices skew away most of ZFS best key features. =3
When I finally became determined to give ZFS a serious go around 2016, I looked at all of my options -- including throwing some hardware together for FreeNAS, and running FreeNAS in a VM.
I decided to just use an existing Linux box that was already doing storage stuff and add ZFS to it.
It's been fine ever since. Keeping ZFS going isn't deep voodoo at all; of all the things this box does, ZFS has probably been the most hands-off aspect of it.
It's a lovely filesystem and it can work in very stodgy ways with zetabytes on Serious Hardware, but it can also work with relatively small devices on silly hardware.
Another example: I've got a ZFS scratch disk that has seen dozens of terabytes written. It's an old ~120GB Intel-branded SATA SSD that is connected with a very cheap SATA to USB3 adapter cable. I've been using it this way for years.
All of the Internet lore has told me that this must result in resolute failure. Whether because USB is broadly not a serious data bus, or because ZFS doesn't like non-serious things, or because 1 copy is 0 copies, or because SSDs are fragile creatures that hate write cycles, or because I've created a portal to hell, or something else: The lore states that it must result in failure, and usually stipulates that ZFS would amplify this concern.
But it's been great. Really. Zero issues -- except when I unplugged it on accident while it was hot...but even then, it came back fine. (Of course it did. ZFS is intended to be robust in the face of unplanned failures regardless of cause.)
And because ZFS: If it ever has any issues, then they'll be reported and logged either in the normal course of doing things, or during a periodic scrub. Every read is checksummed (even in normal operation), and every discrepancy is reported. It's been clean.
It is my intent to keep it as a scratch disk until it becomes problematic, and my further hope is to be able to analyze those problems once they occur.
It won't last forever, but it's lasted for quite a long time so far. (A little bit of inquiry with smartctl and some math just now shows that it has endured writes equivalent to ~43,000% of its capacity.)
If you need your hand held to file share, that's why Windows Server still exists and is the gold standard for SMB.
you also get a whole community testing the stack, fixing it too.
And don't be that pissi... when I started 20 years ago, stuff like this would have been the gateway to linux world and is for sure still today for plenty of people...
How long has it been around? Anyone tried it and can give an impression of what it's like?
I gather now there are two flavors: CORE which continues to use FreeBSD under the hood, and SCALE (aka Community Edition) which is based on Debian Linux. If I understand correctly, the former is only receiving security and stability updates, and iXsystems is developing all new features only on SCALE.
If you were to install fresh today, would you prefer this (FreeCORE), BSDnas, or the official SCALE version - and why?
No MacOS clients, so don't need SMB Spotlight.
Hence the "continued" part in the title.
You can integrate a fleet of Linux storage servers with a collection of K8s clusters more easily than a you can BSD storage servers. Getting BSD and Linux talking smoothly needs Real Unix Skills.
Colleagues in K8s shops tell me that their k8s guys come for help with complex tasks like.... deleting files. They know to construct and deploy containers, but not how to delete a file without using a GUI (meaning Windows or at best WSL).
I am told that some companies are now demanding "bash" among the list of skills for would-be container-herders. And here, "bash skills" means you should know to to create, edit, and then delete a file -- over an ssh connection.
Example: OpenBSD and NetBSD both will install X11 and a window manager by default from the installer, if you want. FreeBSD doesn't. They're working on it, but they've been working on it for years.
But he's very pro-ai, though, and is vibe closing other projects like https://www.phoronix.com/news/Linus-Torvalds-Vide-Coding
Linux is definitely going to speed up as it starts to get vibe coded. In a few years, I bet that it's all going to be built with agentic loops, and nobody will be looking at the code at all.
And why do you act like it's an insult? For most problems, reading and writing code by hand is dead. For the remainder, it will be soon.
> "Also note that the python visualizer tool has been basically written by vibe-coding"
He'll keep ramping up his vibe coding, because he can see it's the future.