Fangtooth 25.04.2 Nightly is available with Classic Virtualization

Tbh, you can use timestamps in the YouTube video description and just listen to the parts that interest you.

But I still must say Truenas team is really tight-lipped about what their plans are.

1 Like

Another thumbs down here for the podcast format. I don’t have the time or the mental bandwidth to sit through videos on a regular basis to glean the stuff I need to know.

At best it’s annoying, at worst it’s discriminatory for those whose brains can’t cope with the format, and may miss important information.

And please don’t just provide a transcript of the podcast :face_with_raised_eyebrow:

4 Likes

What amazes me is that so many are still (seemingly) prepared to play along - continue to adopt and adjust, forget previous sins and time after time pull themselves through breaking changes and painful migrations, all while trying to decipher iX’s communication via social media and forums and read between the lines for forward guidance.

I happen to like this podcast but concede that isn’t always the case, there are plenty of tech-related casts I have no interest in/patience for.

So while I support having the podcast it should not be the only or even primary channel of vital information.

Information that is genuinely important should be available in written form in, among other places, the forum and possibly a blog (plus the newsletter for those who prefer that). Reddit is also a sensible channel to use in addition to the previouslly mentioned.

Given that there is a dashboard in TrueNAS I could also see there being something showing an rss feed of curated tidbits there.

2 Likes

Cmon… In that case everything written here is discriminatory against blind people.

3 Likes

I’m not sure how willing I am to believe that they have any plans at all.

2 Likes

You’ve heard of screen readers, right?

1 Like

Yeah, but if he doesn’t accept transcript then why should I accept screen reader? :smiley:

I think these are a bit unfair - the T3 podcasts do share their short term plans (i.e. next major release and interim releases) so we are hearing about GoldEye functionality.

However I think that there are several issues:

  1. Their actual plans are too short term - tactical plans rather than strategic plans;

  2. Their technology choices often turn out to have been mistakes e.g. choosing Kubernetes over Docker, choosing Incus - and changing again later results in more pain and effort from users - they need to give much more careful thought to adopting new technologies;

  3. These new technologies get implemented in a way that is more painful for users i.e. sudden death cut-overs rather than parallel running;

  4. Risks (both for TrueNAS and for its users) are not properly evaluated, leading to biting off too much and being over confident about success;

  5. Communication is often gung-ho rather than cautious (e.g. T3 because it is marketing material not technical documentation) - and technical documentation and good guidance is often lacking or hidden away rather than clearly pointed out at the appropriate times e.g. when users are actually about to upgrade, possibly with bad short-term outcomes;

  6. Refusal to admit mistakes and using Orwellian NewSpeak to describe a retreat back to mature technology instead as a move forwards to “classic” technology;

  7. Insular thinking - TrueNAS technical management could be too close to the coal face, and taking advice from experienced users might be a way to improve the quality of the planning particularly the above points.

So IMO we have poorly thought through and insufficiently cautious plans, rather than a failure to share those plans with users.

They could have:

  • Thought more carefully about whether the hot, new Incus technology was mature enough or compatible enough for inclusion in TrueNAS before launching into a full-scale technology integration
  • Limited the scope only to Incus LXCs and left virtualisation alone;
  • Done a technology preview separate from the 25.04 release to reduce risks
  • Done Incus VMs in parallel with libvirt so that new technology was optional
  • Not set themselves a fixed deadline of 25.04

Had they done this, perhaps there would have been a better outcome for FangTooth.

After all, the British Army has a motto about chopping and changing - “Order + Counterorder = Disorder”

AFAICT, TrueNAS relies heavily (or has relied heavily in the past) on users providing three things as good will (in recognition of getting to use TrueNAS for free):

  1. Testing - beta testing (of nightlies, beta releases and release candidates), and more beta testing as early adopters upgrade to the first couple of releases

  2. Community support - provided largely by a small group of more experienced users who give several of hours per week of their their time for free for zero recognition from TrueNAS

  3. Word of mouth recommendations

all of which have a financial worth equal to the budget that they would need to spend to achieve the same themselves with paid staff.

And there are several examples that can remind us that users should be valued: Sun Microsystems, Open Office, IRC Freenode, Redis, Terraform immediately come to mind.

4 Likes

iX (or whatever they’re calling themselves) are fortunate to have almost no competition in the “F/OSS ZFS-based NAS” market. XigmaNAS is out there, and zVault may yet fly, but both are based on FreeBSD, which is a dead-end OS. If you like Solarish OSs, napp-It is an option, though its UI is out of the '90s. Otherwise, the only competitor I’m aware of is OMV, and its ZFS integration is minimal at best.

1 Like

Except that this doesn’t even hint if the portion of information that I’m interested in even exists! Just wasting my time.

You missed Unraid.
It has some ZFS support nowadays.

I was just going to mention screenreaders, yes.

And reading a podcast transcript is even more tiring than watching a video as conversation is far less efficient at conveying data, particularly with umms and ahs, interruptions and pauses.

It’s far easier for people with a whole range of accessibility needs to access a chunk of well written and formatted text than it is almost any other media format, because accessibility tools to translate text into other formats are more advanced and widely deployed.

It’s the gold standard, really.

1 Like

Credit where credit is due - as a ZFS-based NAS I believe it’s a fine product and deserves recognition for what it does. The virtualisation journey has been and continues to be a train wreck though.

1 Like

Whether one “likes the podcast” or not is almost irrelevant. It’s still not appropriate as the only source of info.

Screen readers exist.

Not F/OSS, but I hadn’t been aware it did ZFS.

Didn’t I go into that in the following sentences?

Fair point.

That’s a change that’s being planned and evaluated for 25.10, not 25.04.2.

1 Like

It’s also an unambiguously horrible change.

3 Likes