Untitled Linux Show 248 Transcript
Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.
Jonathan Bennett [00:00:00]:
Hey, this week we're taking a look at the pine Time Pro KDE 6.6, the newest Kali Linux release. Ubuntu's sort of controversial changes to Grub and a whole lot more. You don't want to miss it, so stay tuned.
Rob Campbell [00:00:16]:
Podcasts you love from people you trust.
Jonathan Bennett [00:00:21]:
This is Twit. This is the Untitled Linux show, episode 248, recorded Saturday, March 28. It's like a crayon. Hey folks, it is Saturday and as you know, that means it's time for Linux. It's the Untitled Linux Show. We're going to talk software and hardware. I'm going to lead off with some hardware here in just a minute. We're going to talk about Everybody's favorite stuff.
Jonathan Bennett [00:00:45]:
OSes and desktop environments, systemd and the ongoing age verification stuff. It's going to be fun. And let's get into it. So I've got my normal cadre of co hosts, we've got Rob and Jeff and Ken, and we're going to, we're going to talk about some stories. Rob is going to lead us off actually, and he's got one of those, one of those companies, one of those open source groups that we enjoy the stuff from, has a pro version of one of their very old offerings. Rob, what is up with Pine?
Rob Campbell [00:01:20]:
And now it's time for another segment of the Something Jonathan Should Buy show. So apparently nearly seven years after the original Pine Time launched, Pine 64 is getting ready to try again. Which actually really surprises me because I know me and Jonathan both bought one of the original Pine times a few years ago, I don't know, three, four years ago. And at that time I thought that, I thought that it just came out then, but I know we haven't been around for seven years, so maybe, maybe there was some other big story that led us both to get at that time. I don't know, but I guess it's been seven years either way. Jonathan made use of his, but mine collected dust. It was fine, but I don't know, I thought it was kind of chintzy or cheap in my opinion. But to be fair, it didn't cost very much either.
Rob Campbell [00:02:20]:
So what, what, what, what was I to expect? I just didn't feel like wearing it in public. But now in a community update from Fosdom 2026, the company revealed the Pine Time Pro, an updated version of its fost friendly smartwatch. The new model is set to include an amyloid display built in, gps, blood oxygen sensing and rotating digital crown with an extra button which sounds a lot more premium and maybe somewhat comparable to some of the more popular smartwatches today. It's at least clearly an upgrade from the original that is going to be more capable while still appealing to the same open source crowd. Pine 64 always says the Pine or also says the Pinetime Pro will use a custom chip, though technical details are still limited. One of the key improvements is that individual components can be powered down independently, which should help battery life. Paired with the AMOLED screen, that could make the new watch more efficient than its predecessor. And I mean, I hope so.
Rob Campbell [00:03:43]:
Efficiency is kind of a thing that usually comes with progress, so. But there are still plenty of plenty of unknowns. Pine64 has not announced a price or release date yet and it is still deciding how to handle fundraising. But Whatever happens, the Pinetime Pro is unlikely to match the original sub$30 price, though it could still be a very strong value. I mean, yeah, I think it was $20 or something somewhere under $30 US dollars. And I don't know, it's so cheap. I'm like, yeah, let's try it out, I'll buy it. And I did nothing with it.
Rob Campbell [00:04:22]:
But another open question is software in Infinitime, the open source firmware that helped keep the original Pine Time relevant, may or may not run on the new model. So not sure what they're going to have on there for sure yet. So for now, Pine64's message is simple. The Pine Time is back. This time with more features, more ambitions and plenty of questions still to be answered. If this turns out to be a decent watch, all we really will need is a decent Linux phone to go with it.
Jonathan Bennett [00:05:02]:
So I've seen some of these different devices come and go. I do a lot more embedded work these. And so I've got some examples here of different watches. And I will say one of the things that the original pinetime really had going for was its size. It was tiny. It actually worked as a watch. So you've got like some very cool things. It's a lily go watch, but it's huge.
Jonathan Bennett [00:05:24]:
And this is one of the things that is sort of the. Yeah, I won't tell you what people have described this as being similar to because it's not very nice. But just think about it a little bit and I'm sure you'll come up with it. But it's, you know, I mean you can pack a lot of capability and a lot of battery into something that size if you're willing to go that big. But it Then it, it's just, it's become bigger than the size of watch that most people would be willing to wear. And looking at these two pictures, I'm kind of afraid that the pinetime Pro is going to kind of stray into that same problem. It's just, it looks like it's, it's fatter, it's bigger and I don't know, I think, I think that's really the biggest advantage that the original had going for it is it was so small.
Rob Campbell [00:06:07]:
Yeah. What I do remember from the original is at least when it was off, it looked like a decent premium watch. Like the Apple Watch or anything else. Though when it turned on, you just had like a square screen that was noticeably separate from the, the outside. You know, when it was off, it looked like the whole thing was a screen just like an Apple Watch. But yeah, just had that little odd screen that, you know, it could have been worked with, but I don't know I really have time for it. It was just. I, I donated my $30 to their, to their cause initiatives.
Rob Campbell [00:06:52]:
To their cause.
Jonathan Bennett [00:06:53]:
Yeah.
Rob Campbell [00:06:53]:
Yeah.
Jonathan Bennett [00:06:53]:
One of the, one of the other things, one of the other things I see pointed out is that apparently Pine is having. Because the price of RAM trying to keep some of these devices manufactured and there's questions about what that's going to mean for this one. And I will say probably not as much because this is based on an NRF, which is an all in one. It's like an SoC, a little tiny SOC from Nordic. I'm very familiar with the nrs that they're using here. So the RAM prices have not hit that yet. Although I will tell you, I have heard rumblings that going forward in the future, the, the availability on even some of these little MCUs is going to be affected by the various shortages we're going through.
Rob Campbell [00:07:33]:
Yeah, you guys could tell me more, but I can't imagine whatever RAM is in the tin, the tiny little RAM is in that watch, it's is the same as the same production lines necessarily
Ken McDonald [00:07:44]:
so much as that it's getting a production line to do that instead.
Rob Campbell [00:07:48]:
Well, right. What I was saying is I don't imagine it's the same production lines.
Jeff Massie [00:07:53]:
A lot of times those little socs, they just design the RAM as part of the chip. And a lot of those, they're not running the big cutting edge stuff. So the price of RAM has nothing to do with those little socs other than if they're contracting out. But usually they're not cutting edge because they don't have the sheer count of transistors and they're not volume. So they're running the probably in a fab with 8 inch wafers. And so it's not even holding up the huge manufacturing lines.
Jonathan Bennett [00:08:26]:
So this is a Nordic NRF 52 which that's like two generations old from the new NRF 54 that Nordic is really trying to push right now. So it's based on the arm cortex M4, it's 64k of RAM. So yes, we're not talking about very much RAM. And so I was sort of saying that tongue in cheek. However, I am hearing from my MCU guys that even little tiny things like this are starting to be difficult to get in volume.
Ken McDonald [00:08:51]:
Is that because the volumes decreased over time for, especially for the older ones.
Jonathan Bennett [00:08:56]:
It's just because everything is growing so big and there's so much competition for fab time. For everything.
Jeff Massie [00:09:03]:
Yeah, so it kind of, it kind of pushes everything down. Right. So for example, let's just come up with like tsmc. They're booked with the latest Nvidia, Apple, some Intel stuff. Well okay then who's going to make that? Well, okay, Samsung picks up some, intel picks up some. But some of the lower end chips they were making now gets pushed to more mom and pop shops. You know, some of the 8 inch, 5 inch, 6 inch fabs, but the stuff they were making then gets pushed down. And so it, it's just everybody's pushing wafers through as hard as they can.
Ken McDonald [00:09:39]:
Yep.
Rob Campbell [00:09:39]:
You know I, I heard some, I was reading, I didn't hear, I was reading some of the comments in, in one of the articles and a lot of the commenters were not very positive about this. Basically magic about how they've been burned on things like the pine phone, on how you know, they release their pine phones and then dropped the ball and just kind of didn't do anything with them, didn't go anywhere with them, didn't make them keep improving or make them useful and that that was a good number of the comments on there I saw.
Jonathan Bennett [00:10:13]:
Pine has sort of gained the reputation of building something that's hardware and then acting like it's spaghetti and just kind of throwing it against the wall and see if it sticks. See if anybody will build good software
Jeff Massie [00:10:22]:
for it, more firmware. Personally, I'm just waiting for the, for the good Linux phone.
Rob Campbell [00:10:30]:
Me too, whenever that comes up. So I find a little odd that they throw a watch out there.
Jeff Massie [00:10:34]:
But I'm a bit of a Luddite. I'm all about the old school Analog dial.
Jonathan Bennett [00:10:39]:
Nice.
Ken McDonald [00:10:41]:
Well, I got so used to not wearing a watch because every time I'd get to work when I was in the Air Force, I'd have to take it off.
Jonathan Bennett [00:10:49]:
True.
Rob Campbell [00:10:50]:
I like to be able to see when my meetings are and when I have, have a maps going down the road, it'll vibrate and say, oh, you gotta turn so I can have this thing. I could have the phone muted so I don't have to hear the dumb, you know, voicing turn. Right.
Ken McDonald [00:11:08]:
Which voice did you select?
Jonathan Bennett [00:11:11]:
So apparently talking about the pricing thing again, apparently this week, Intel, AMD, STMicro Allegro and probably a couple of others have all come out and said, we're going to have to raise prices. It was a lot of different people that were affected by again, this crunch on the, the fab side of come out and, you know, we can't, we can't do what we're doing. We can't leave prices where they're at. And so they're gonna, they're gonna increase prices pretty much everywhere. So, you know, you got that to look forward to for this year, the rest of this year.
Ken McDonald [00:11:47]:
Quickly, I wanted to touch on the firmware. You said Infinity Time may or may not be the default. Is it possible Pine's thinking about going with the Pine Time Light, which is a fork of Infinitime?
Jonathan Bennett [00:12:01]:
I mean, it's hard to say. I was actually going to take a look quickly and see if I could see what. Yeah, it looks like they have an NRF52 port, so it wouldn't be difficult to make Infinitime work on the new device. It's all NRF52. So it's just a matter of porting over the display driver and then remapping the GPIO and then adding drivers for whatever new, new, new hardware you've got on there and then adding for, you know, whether they're. They say they can turn on and off the individual pieces of hardware. So they either have that control with just GPIOs or they've got like a power controller switch on there. And either way it's pretty straightforward to get support for that going.
Jonathan Bennett [00:12:44]:
So they could, they could do it with, they could do it through Infinity. And with this, if the guys at Pine want to do it is they may fork Infinitime, build the support there and then slowly push it back up. That's a pattern you see pretty often. So it's hard to say. This kind of makes me want to go pull my Pine Time back out and flash the newer Infinitime to it because I know they've been working on it and this thing has continued to get updates and they've made it better and they've added features, so I kind of want to go try it out again.
Ken McDonald [00:13:21]:
Did they add the ability to run KDE Plasma on it?
Jonathan Bennett [00:13:26]:
So if you hadn't picked up that is a segue to Jeff's story about KDE Plasma 6 and there is some interesting stuff going on over there. Jeff, take it away and tell us all about it.
Jeff Massie [00:13:42]:
Well, I have two articles in the Show Notes today and both are benchmarking. No, not the new Intel CPUs though I will cover those when the Linux benchmarks are out there. No, this week we're looking at GNOME and KDE on both AMD and Nvidia hardware. It's been a while since we compared the two, so let's get started. All the testing was done on Ubuntu 26.04 beta with Gnome 50 on Wayland and KDE Plasma 6.6 with both Links X11 and Wayland sessions. The GPU on the first story was AMD Radeon 9070 XT with Mesa 26 and the second story where it's testing with Nvidia hardware, It was a GeForce RTX 5080 with the new Nvidia 595.58.03 Linux driver. All the software was how it comes out of the box, you know, from the Ubuntu 26.04 beta repository. So no settings were changed, just the default was how it was rolled.
Jeff Massie [00:14:45]:
Between the two articles there's some games which don't match, and that's because the AMD Radeon graphics still had some hard freezes with some of the games that prevented some from actually being tested on the AMD hardware. And there's some missing x11 benchmarks in there as well because some things just wouldn't run. So any missing data is because something locked up or couldn't run at all on the AMD GPU. When looking at the overall geometric mean, Plasma 6.6 Wayland came out on top. A little bit behind was Plasma X11 and a large step back was Gnome 50. So on AMD hardware KDE is performing better than Gnome. In the second article looking at the Nvidia GPU, there's a little change as there are no X11 benchmarks. Trying to get the X11 working on Nvidia was not working so well, so Michael had to abandon those results.
Jeff Massie [00:15:35]:
So on the Nvidia Hardware, it's just wayland only roundup. Other than that though, things didn't really change. KDE Plasma came out on top by a decent margin over gnome 50, pretty much very closely matching the AMD GPU test results. And on both of them there was a few places where GNOME was a little bit better than kde, a few places where they were identical, and several where KDE came out on top, and there was a couple where KDE was almost double the speed of gnome 50. Bottom line, if you want the fastest desktop between KDE and Gnome, you need to be on kde. Now, one thing that didn't show was resource usage. You know, I did take a look around the Internet a bit, but I didn't find anything recent. And a lot of times there was just some numbers published by somebody, but not a lot of context of software and revisions and how they tested.
Jeff Massie [00:16:28]:
They kind of just threw up some. Here's what I did. Well, I don't put a lot of stock where I can't see the methodologies and revisions of everything, so for now it's going to remain an unanswered question. You can have a look at the articles linked in the show notes and see what you think for yourselves. And let us know if this might change the default desktop you're going to be using in the future.
Jonathan Bennett [00:16:54]:
Yeah, this is super interesting. Worth pointing out that Ubuntu 2604 is just now in beta. Just had the beta release of that and April 23rd I think is what I saw the full release date is going to be. Yeah, April 23rd. So coming up soon, Resolute Raccoon. Install it somewhere 7.0. Yeah, I have to install 2604 somewhere just to be able to say I did.
Rob Campbell [00:17:20]:
It may be the last Ubuntu worth installing, but we'll talk about that later.
Jonathan Bennett [00:17:25]:
Wow, that is quite the statement.
Jeff Massie [00:17:28]:
Throwing some shade around.
Rob Campbell [00:17:29]:
Goodness, did I hear you say that he couldn't get Nvidia to work with X11, only with Waylon.
Jeff Massie [00:17:37]:
Yeah, but I mean, it's beta, so
Rob Campbell [00:17:41]:
I mean, that's sure a big change though, from, you know, the past and what everyone's talking about.
Jonathan Bennett [00:17:46]:
No, you don't. Don't you remember the feature we talked about about a year ago? That on Wayland under Nvidia, you had to wiggle your mouse back and forth to get the screen to refresh.
Rob Campbell [00:17:57]:
Right, I remember that now. Now it's the opposite now.
Jeff Massie [00:18:00]:
X doesn't work well and it was also the Ubuntu beta and it's using the brand new, you know, that 5.9 5.8.03 driver. So there's a lot of cutting edge stuff. And I have a story later that could go into maybe why it didn't work. So there's some revision requirements for X11 that maybe it wasn't meeting yet.
Rob Campbell [00:18:25]:
If you want cutting edge, you need Wayland. I hired it.
Ken McDonald [00:18:29]:
At this point, I seem to remember years ago you had problems with Nvidia, period.
Jonathan Bennett [00:18:38]:
That's still true.
Jeff Massie [00:18:39]:
Sometimes they clean it up a lot.
Jonathan Bennett [00:18:42]:
It's better now. Yeah, it's a lot.
Jeff Massie [00:18:43]:
They've cleaned it up a lot because yeah, it used to be you had to have a few settings or you'd get a black screen when you first booted in. You had to force some modes and things like that to get it to work. Once you did that, usually things were pretty okay on X11. Now fast forward to say the last year, Nvidia is really leaning into the AI. Well, AI is heavily into Linux so they've really tried to buddy up to Linux. So they, they've been throwing some extra resources into making the GPU side work a lot better.
Rob Campbell [00:19:19]:
I mean it is heavier, heavy in Linux, but not necessarily for the, the graphics and display part of it that we tend to care more about.
Jonathan Bennett [00:19:28]:
Yeah, but yeah, you get, you get the trickle down, you get some trickle down economics going on.
Jeff Massie [00:19:35]:
They're trying to get rid of the stigma.
Rob Campbell [00:19:37]:
Political here, Jonathan.
Jonathan Bennett [00:19:39]:
Yeah, so one more thing about this particular story with it being on Ubuntu 26.04, I want to tease real quick. The Ubuntu conference, the release conference is going to be in London end of May I believe, where they're going to celebrate this, have a bunch of talks. Looks like I'll be there for that. So we'll have more news about that as we get closer to it. Links and all of that. But last I heard my talk idea was accepted and so hopefully we can get over there and let you guys see it. All right, so here in a minute Ken is going to put on his hacker hat. Didn't realize he had one of those, but apparently he does and he's going to talk about Kali.
Jonathan Bennett [00:20:24]:
Linux will let him do that. Read for this. All right Ken, tell us about Kali.
Ken McDonald [00:20:29]:
Well, Jonathan, as you said, we've got some information coming this week from Bobby Borisov, Marcus Nestor and Jack Wallen. They all wrote about the latest release of the Debian based GNU Linux distribution for penetration testing and ethical hacking. Kali Linux 2026.1. According to Bobby. It features a new annual theme, a Backtrace inspired desktop mode, updated tools, and a kernel upgrade. This year's visual refresh includes a new boot splash screen, a new bootloader theme, and a brand new desktop theme for the flagship XFCE edition. It also includes updated artwork for the graphical installer and the login and lock screens. This year marks the 20th anniversary of Backtrace Linux Backtrack Backtrack Linux Excuse me, that's my mind wandering.
Ken McDonald [00:21:35]:
The predecessor to Kali, which celebrates this milestone by adding a backtrack mode to Kali Undercover. This mode transforms the desktop to recreate the look and feel of backtrack 5 with the same wallpaper colors and window themes. Jack even provides more historical background in this article. According to Marcus, the Kali NetHunter app for mobile devices has been updated as well with a new kernel for Android 16 on Redmi Note 8 devices, support for using the internal wireless Firmware of Samsung S10 series devices in Kali, Chiru WI Fi, packet injection on QCAL D 3.0, and various bug fixes. I'm going to let you read Bobby, Marcus and Jack's articles to get the list of new tools added to the repositories, but definitely check out that history section that Jack gives.
Rob Campbell [00:22:44]:
Yeah, he's like KD3 or, or just Trinity Desktop to get that old KDE look.
Jonathan Bennett [00:22:50]:
No, it's, it's just, it's just theming.
Rob Campbell [00:22:52]:
This is go all in, go all retro.
Ken McDonald [00:22:58]:
You can do that with Cali 66 or KDE 66.
Jonathan Bennett [00:23:02]:
Yeah, I'm now looking at the history of Backtrack. It was the merger of Auditor Security Collection and Wax. I think it was sort of before my time even. I do remember using Backtrack back in the day, but Wax was a slacks based Linux distro for Security and Auditor Security Collection was a live CD based on Knoppix. Oh there's. Yup, I've not heard that name in a long time, but I do remember Navix.
Rob Campbell [00:23:33]:
Yeah, I wouldn't say they're before my time, but I Before you were using Linux? No, I don't think they're before I was using Linux either. I mean nopix is came after I was using Linux. It's just I. I think they were just never in the public eye enough that I heard of about them or I was.
Ken McDonald [00:23:51]:
Yeah, I think there's a reason for that.
Jonathan Bennett [00:23:55]:
Yeah. Well when I say before my time, I mean before I was invested in Backtrack and now Cali. Yeah but yeah man, nopix brings back some memories though. Was that. That was one of the first live cd.
Rob Campbell [00:24:09]:
Yeah, I think it was the first. That was their thing.
Ken McDonald [00:24:12]:
One of the first ones that allowed you to do penetration testing.
Rob Campbell [00:24:16]:
Knoppix.
Jonathan Bennett [00:24:17]:
I mean you could off of Knoppix but the backtrack was really the one that was.
Rob Campbell [00:24:22]:
Yeah, I don't think not made it easy. I don't think Knoppix had anything special about it for pen testing.
Jonathan Bennett [00:24:29]:
No, no, I don't think so. But yeah, it was. Man, that was cool. Cool times. I remember that.
Ken McDonald [00:24:33]:
Why am I thinking of watching a Screensavers episode where they talked about using Knoppix while they were for Drive by Wi Fi.
Jonathan Bennett [00:24:47]:
Very possibly war driving. That was the thing.
Jeff Massie [00:24:52]:
Well, I haven't heard that term for a while.
Rob Campbell [00:24:54]:
Yeah, it'd probably be beneficial at the time just because you could boot into it and have a live boot session and then just do whatever, which. But you probably, probably won't be any different than if you just installed a
Ken McDonald [00:25:07]:
wardrobe and then just pulled it out of the laptop when somebody asks you to check it out and they find out it's Windows.
Jonathan Bennett [00:25:17]:
There you go. Fun times. I've not installed Kali Linux for a while. I need to. I need to get another bootable, bootable Flash drive going again and get the newest, newest Cali on it and all that good stuff.
Ken McDonald [00:25:30]:
Is that one that you could put on a Ventoy and boot from.
Jonathan Bennett [00:25:35]:
I'm sure. I'm sure you could. You may have to grab the newest Ventoy update. I've had times where a new ISO wouldn't boot without the newest Vento before, but I'm sure it is something that you could, you could get going there pretty easily.
Rob Campbell [00:25:47]:
Cali or Parrot? Or just put them both on there.
Jonathan Bennett [00:25:50]:
Put them both on there. You might as well. Flash drives are cheap these days and big.
Jeff Massie [00:25:56]:
Big.
Rob Campbell [00:25:57]:
They're still.
Jeff Massie [00:25:58]:
Yeah.
Jonathan Bennett [00:25:58]:
Oh, oh, oh, big.
Jeff Massie [00:25:59]:
Okay.
Ken McDonald [00:26:00]:
Size wise, storage wise.
Jonathan Bennett [00:26:04]:
I thought you were telling us that we needed the big Linux distro and I was trying to figure out what you were talking about. No, I'm not familiar with that one, man.
Rob Campbell [00:26:11]:
I think. I think there is a big Linux though.
Jonathan Bennett [00:26:13]:
Probably there is.
Ken McDonald [00:26:15]:
I want to say it's from a South American country. I can't remember which one.
Jonathan Bennett [00:26:21]:
Offset. Well, I mean, we have all kinds of weird names. We've got Grub, which is. Yeah, they're big Linux. It's a Brazilian Manjaro. But Ubuntu's Grub is making some changes. Rob actually has a story about this and it's dropping file systems. Oh no, don't we need those?
Rob Campbell [00:26:41]:
Somebody thinks so. So Ubuntu may be headed toward one of its more controversial changes in years, or maybe just months. I mean, shocking Ubuntu creating controversy again. That's kind of what they do these days, isn't it? So this time it centers on Grub, secure boot and what users may have to give up in the name of security. Canonical is proposing to strip down the signed Grub bootloader for Ubuntu 2610. Grub's many parsers and features have been a steady source of security problems. So reducing what signed Grub can do would shrink the attack surface in the pre boot environment. From a security standpoint, that makes sense.
Rob Campbell [00:27:35]:
And you know, security and usability is always a balancing act. It's always been that way. But the trade off could be huge for some people. Under the proposal, signed Grub will lose support for butterfs, XFS and zfs, along with lvm, most MD RAID modes, Luke's encrypted dry disks, and even some theming support. In practical terms, secure boot systems would be pushed toward a very basic setup. Basically an unencrypted ext4 boot partition. That's, that's all you can have if you want Ubuntu now in. It's not the end of the world.
Rob Campbell [00:28:23]:
But if you wanted the advanced stuff back, you would need to use an unsigned Grub build and give up the secure boot. So for Ubuntu, security is only important for unencrypted ext4 and isn't, you know, I mean, isn't, isn't kind of a lack of encryption also its own security issue? I mean, if you really want a secure system, don't you also want encryption? I know, maybe, maybe boot drive, it doesn't matter as much. But anyway, things like butterfs, you know, as you know, it's not a fringe file system anymore. It's. It's one of the most popular Linux file systems out there after ext 4, probably. Especially with users who want to use snapshots and rollback features. And I know Jeff has had some experience using it lately with Cashios. Then there's zfs or for you guys, cross Pond or in Canada, zfs, which makes this especially awkward for Ubuntu and why, and that's because Ubuntu has long been the only major Linux distro willing to support zfs.
Rob Campbell [00:29:39]:
ZFS en route, out of the box. Nobody else was willing to, willing to cross that licensing line. So if you wanted ZFS, ZFS, you had to use UbaTub. You know, that's been a unique selling point. This proposal would, it wouldn't remove ZFS from UBA2 entirely, but it would remove it from the signed Secure Boot path. So your boot path could be ext4 and everything else could be ZFS. But you know that it's, it limits, it limits your, your, your options of what you really want. Maybe you want, you know, all the benefits of ZFS or butterfs in, in, you know, on your, on your route.
Rob Campbell [00:30:23]:
Because you know, if you want to have snapshots, that's kind of one place you want to be able to snapshot. So anyway, technically, you know, all these things will still exist. You know, for users who want Secure boot, it's basically just gone on root. Critics have also pointed out that some of the features Canonical wants to cut are, are not necessarily niche, such as RAID 1 boot setups. Those are fairly common. Ubatu's own installer has leaned on LVM for quite some time now. Ubatu's encryption setup also depends on LVM today. So this is not just removing obscure options nobody uses.
Rob Campbell [00:31:06]:
It risks breaking with the way Ubuntu itself has told people to install systems. So once again Canonical is going out on a limb seeing if they could push the limits of whatever they want to push this week. Better security, I guess what they think by making Sign Grub much smaller and simpler. But in doing that, UBA 2 could also lose some of its key strengths. I mean, I know a lot of Linux users are just used to disabling secure boot anyway, but that's also kind of been one of the. One of the benefits to UMA too is you actually could just have Secure Boot on and use that feature. But. There you have it.
Jeff Massie [00:32:00]:
You know, that really surprises me because I could see maybe ZFS because it's not built into the kernel.
Ken McDonald [00:32:07]:
Take a step back here. Which partition are they doing this with?
Rob Campbell [00:32:12]:
It's just for the boot up process for Grub. So it only affects the root partition
Ken McDonald [00:32:16]:
once you've got booted up. And then your root can be on a ZFS or butterfs, correct?
Jonathan Bennett [00:32:23]:
Yeah, it's not changing the root partition, just boot. I want to hear what Jeff was thinking though.
Jeff Massie [00:32:29]:
Well, I could see though that maybe zfs you could make an argument because it's not built into the kernel, but betterfs.
Rob Campbell [00:32:38]:
The.
Jeff Massie [00:32:40]:
Well, basically the other ones are built in. It's like, yeah, it makes no real sense to Me, I think, I mean, I think I hear what they're saying. But then the comment would seem to be, well, why not just use System Debut? If you want to just really streamline
Rob Campbell [00:33:00]:
booting process, you can get rid of Grub altogether. But yeah, you know, I, I think Butterfest should be there. I mean, that's, that's like, that's huge these days for the competition itself. Well, I mean, that's how you do snapshots. If you want to snapshot your entire system, you know, even your boot, you're gonna def. You're gonna want to snapshot that, because when you do update, you want to be able to roll back your whole boot and kernel and everything.
Ken McDonald [00:33:29]:
But if you've got a copy of that boot under your root, the boot partition itself can be anything while the copy is still under the. But beyond Butterfs, correct, you can, you
Jonathan Bennett [00:33:44]:
can mount it there, but that does not give you the snapshot shotting you
Rob Campbell [00:33:49]:
want to actually be snapshotting.
Jonathan Bennett [00:33:50]:
So in the, in the thread here, one of the guys pushing back on this is Conan Kudo, which. That is the, that is the username for Neil Gampa, a friend of mine, and he makes some really interesting points about why part of this is a bad idea, including and how that you want your shot tied to a particular kernel version. That's sort of point, part of the point. And then also apparently some of the different flavors of Ubuntu use those images which they are looking to try to turn off. I, I must say though, that I get the point that all of these different flags and things inside of Grub is a NSC put, as Julian K. Puts it, a constant source of security issues. And Grub has had some security problems trying to do really solid, secure boot. And so I get why they want to do this.
Jonathan Bennett [00:34:52]:
The challenge you have with something like this is every time you list something and say, we're going to remove this, someone on the Internet somewhere is going to go, no, that breaks my workflow. And of course there's an XKCD comic about this. Can't you just add an option to heat up again? Go check it out, you'll get the joke. But it's hard to take features away from users. As someone who has run open source projects, it's hard to do. Sometimes you have to. Sometimes you got to swallow the bitter pill and say, no, this was always a missed feature and it's going to go away. I'm sorry, I don't know if this is.
Rob Campbell [00:35:27]:
I don't think all those Files systems. If we just stopped using ext4 altogether at some point and then I don't know.
Jeff Massie [00:35:35]:
Well, I mean to me too Grub has got such leg legs under it. I mean it's been around for quite a while now. It's
Ken McDonald [00:35:44]:
for Secure Boot.
Jeff Massie [00:35:46]:
Oh yes, yeah. Yes. But I mean if you're. I. I still say if you. You would be better off if they just said okay, we're going to go System D or we're going to do something else rather than try to tweet. They really want to that worried about it.
Rob Campbell [00:35:59]:
You know I think systemd boots are more of a future. Well maybe not with systemd's current. I don't know.
Jonathan Bennett [00:36:11]:
Is systemd boot UEFI only I don't remember. I think it might be and you would immediately go oh well that doesn't matter. Everything is uefi. Well no, everything hardware is uefi but you get into virtualization and cloud computing and a lot of those are actually not booting UEFI because the, least the last time I checked the tools to virtualize UEFI were still very immature and there was a lot of old school BIOS booting.
Rob Campbell [00:36:41]:
You know, one other option could be they could maintain maybe two or three or I don't know, they could decide what other features are really important and maybe they've to them they've already decided that but you know, they could have a Butterfs only one a ZFS only one. You know what, whatever ones they decide are and I'm sure it, it probably cost a minute, I don't know to get each one signed or something.
Jeff Massie [00:37:08]:
But well, and you know they could even do like cache. When you first go into it and you're loading it up, they ask you what bootloader do you want? You want System D, you want Grub? I think there's a couple others they could just go do you want willing to take the risk with full open Grub or do you want the lockdown basically ext 4 grub?
Rob Campbell [00:37:27]:
No. I wonder. I didn't even pay. I don't pay attention. This cache is signed.
Ken McDonald [00:37:34]:
Yeah, well that's the question I had too. Are you still able to install Linux distro with using an unsigned Grub build?
Jonathan Bennett [00:37:46]:
You can, yes. You have to go into your BIOS settings and tell it that you have to turn off Secure Boot. Essentially that's. That's the point of having assigned Grub
Rob Campbell [00:37:57]:
is that I mean for years a lot of people are instantly told that. Like when Secure first came out, people Are told, well, when you install Linux, you got to disable secure boot and things like that. I mean, that's kind of mostly a thing from the past.
Jeff Massie [00:38:12]:
Well, and realistically, okay, there's security issues in Grub. How many of those are like you have to have access to the local machine versus remote access.
Jonathan Bennett [00:38:24]:
They are almost entirely. In fact, I would say that probably every one of them are. You have to access, have access to the local machine. The reason why that does matter though is particularly like in corporate environments. You'll have setups where the drives are encrypted and Grub will get an encryption key from like a local TPM and use that to.
Rob Campbell [00:38:46]:
How are we going to encrypt them with Lux? Trust that they're taken away?
Ken McDonald [00:38:54]:
Actually, no.
Jonathan Bennett [00:38:55]:
Well, so there's a couple of different ways to do drive encryption. Lux is not the only one. This scheme also does work with an unencrypted boot if you have working secure boot, because it doesn't have to be encrypted to check that security signature on your actual kernel. And then the part that's encrypted is the rest of the hard drive, right, like your root partition. And so this is a thing that you actually see. In fact, it's required, unfortunately, in certain use cases that you've got to have some sort of solution set up like this to where you can have an encrypted, an encrypted root partition and then some way to automatically decrypt it at boot. And the way that this is done in many cases is to talk Grub. There is a bug in Grub run, you know, whatever.
Jonathan Bennett [00:39:54]:
If you can run kernel image or an unsigned initial init, ramfs, something like that, well, then you can decrypt your root volume and do whatever you want to on it. You can escape out of that sort of sandbox environment, that encryption sandbox that is intended for the machine to use. So like in certain, in some of these circumstances, it is actually a really big deal to keep Grub 2 really secure. And so that's the sort of thing that they're getting at.
Rob Campbell [00:40:27]:
Well, you know what they could also have, they could have that. They could just have two. They could have the same Grub that they have today signed and then they could have their ultra secure simplified Grub.
Jonathan Bennett [00:40:40]:
I think you run into questions there about whether Microsoft would be willing to sign the two different versions of it. Isn't it Microsoft that signs all of these for Linux? Linux distros?
Rob Campbell [00:40:50]:
I think so. But they're like buddy Buddy, they're really close friends.
Jonathan Bennett [00:40:56]:
Because Microsoft loves Linux.
Rob Campbell [00:40:58]:
Well, especially Canonical. I mean, they were the first ones on wsl and they're buds, basically.
Ken McDonald [00:41:05]:
Is this proposal primarily aimed at enterprise users?
Jonathan Bennett [00:41:14]:
You know, it's. It's in the proposal. It's not necessarily aimed at anyone, but I'm just saying the group of people that really care about Grub security, that's where they're at. They're in the enterprise. There, there. It's the group of people that also care about security compliance and FIP certification, blah, blah, blah.
Rob Campbell [00:41:32]:
I know. I think I saw something about them getting a new fips recently.
Jonathan Bennett [00:41:36]:
Yeah.
Ken McDonald [00:41:37]:
So unless you're actually wanting to play around with the system using a Secure boot, then you could just
Jonathan Bennett [00:41:49]:
turn on Secure Boot and Boot, whatever. Yeah, yeah.
Rob Campbell [00:41:52]:
I mean, you know what? I'm just gonna keep doing what I'm doing and not use Ubuntu.
Jeff Massie [00:41:59]:
If you can turn off Secure Boot.
Jonathan Bennett [00:42:01]:
Secure Boot. Are there any desktops, are there any motherboards out there that you can't.
Ken McDonald [00:42:08]:
The one in the HP system, I've got.
Jonathan Bennett [00:42:11]:
Can you not turn it off?
Ken McDonald [00:42:13]:
No.
Jonathan Bennett [00:42:13]:
I found. I have found a way. On every. Every machine I've ever touched, I've eventually found a way to turn off Secure Boot. I don't think I've ever worked on one that I couldn't. But that is not to say that that doesn't exist. I would absolutely believe that it would, especially with hp.
Jeff Massie [00:42:31]:
Well, some of the newer stuff, it's just kind of defaulting to that. That's what you want to run.
Jonathan Bennett [00:42:38]:
Well, yeah. I mean, everything defaults to Secure Boot on, and sometimes they really work to hide it, make it difficult to turn it off. You know, I've. I've. I've seen times where it's like, you know, you have to go into a completely different BIOS menu and then it'll reboot again and it'll put a code on the screen. Like, if you really want to turn off secure boot, type in 053 and press enter. And then it'll reboot again and finally turn it off for you. All kinds of fun stuff that seems
Rob Campbell [00:43:03]:
like it'd be very limiting if there's actually something, you can't turn it off at all.
Jonathan Bennett [00:43:08]:
Yeah, indeed. All right, well, let's move along and let's talk Nvidia again. And guess who? It's Jeff that has this story as well. Somehow we're not surprised. Give us the scoop. Sell us on Team Green.
Rob Campbell [00:43:23]:
Tell us how bad they are. Go for it now.
Jeff Massie [00:43:28]:
So in the last article story we talked about Nvidia hardware using the 595.58.03 driver, and if you've not heard of this one, that's because it just came out a few days ago. So the new driver marks the first stable release of the 595 production branch, and it has some meaningful improvements for Linux for Wayland and Vulkan and Linux gaming overall. Something new is that the Nvidia DRM space mode set equals one is now enabled by default, which is some of the historical stuff we were talking about. You'd have to set manually set your mode set and your group your grub boot options. But now that's set by default, which is important for proper Wayland support. In Gnome and kde, Plasma updated the driver to allow Nvidia SMI to reset GPUs while Nvidia DRM is loaded with the mode set equals one parameter enabled as long as no other processors using the GPU. And they also fixed a regression introduced after the 470 release series that could cause four 4K monitors driven as separate x 11x screens on one GPU to fail mode set. So they put a lot of work into mode set for this driver.
Jeff Massie [00:44:56]:
They have added a Vulkan extension VKEXT descriptor heap, which allows games, especially those running through Proton with DXVK or VKD3D Proton to store buffers and images directly in the GPU accessible memory. This reduces CPU overhead and lays the groundwork for future performance gains and Windows to Linux translation layers. While users shouldn't expect immediate FPS boosts, this is a foundational change that will matter more as Proton and Wine integrate it more. Another extension added is the vkextpresenttimings which gives applications detailed timing information about frame presentation and lets them target specific presentation times. This helps smooth out gameplay and video playback, reducing stutter and improving frame pacing. The new driver also adds DRi3 version 1.2 support, bringing DMA fences to Nvidia's stack. This is an improvement for modern Linux gaming, helping prevent tearing and improving synchronization between GPU and display servers. There is for those that like number crunching, there's a new application profile.
Jeff Massie [00:46:10]:
There's a CUDA no stable performance limit which allows CUDA based applications to reach the GPU's highest performance state P0, which can significantly improve compute workloads or that were previously capped. Now, while researching things for the driver I also found some people are now saying that HDR now works without Needing the HDR layer or Enable HDR WSI though they didn't officially document it, but it seems like those were a little more inner. Those were temporary fixes until other groundwork got laid and it sounds like now you possibly don't need those anymore. So they, they've streamlined the driver a little bit. While there's a lot of goodies and bug fixes in this release, I. I should add it does raise some system requirements. So you need to have Wayland version 1.2 or 1.20 along with glibc 2.27. And this could be what was going on in Ubuntu 26.04 beta and not running X is you have to have Xorg 1.17.
Jeff Massie [00:47:25]:
So take a look at the article linked in the show notes for more details of upgrades and bug fixes. And you know, there was a lot of fixes for freezes and stutters in games, you know, and things that running better now that I didn't even cover. So there's. There's a lot still in there. The articles also have a link to the official Nvidia release notes, so happy reading.
Jonathan Bennett [00:47:49]:
Apparently the Nvidia release notes have an issue that they. They have things in the changelog that were for the previous version. So it's sort of a pain to read in some cases.
Rob Campbell [00:48:00]:
Hey, it's new Nvidia, of course it's not going to be good.
Ken McDonald [00:48:05]:
So you take and compared the previous changelog to the current one with diff.
Jonathan Bennett [00:48:10]:
Yeah, you diff it. And the difference is the ones that are actually new, they were really trying
Rob Campbell [00:48:14]:
to stuff it to make it look like they did more than they did.
Jeff Massie [00:48:19]:
Well, you know, the thing is though, is as much as people hate Nvidia, they're like 95% of the market.
Jonathan Bennett [00:48:26]:
Oh yeah.
Jeff Massie [00:48:26]:
As much as everybody says, oh, I hate Nvidia, you know, why are people buying AMD more?
Rob Campbell [00:48:32]:
You know. You know, as much as people say they hate Windows, it's still like 95% of the market. Come on, Jeff, that's not a good reason.
Jeff Massie [00:48:41]:
Well, it is because. Because on Windows you could at least argue you're locked in because of software.
Rob Campbell [00:48:46]:
I think that's you have a reason to hate them. You got to hate the big guys, you got to love the underdogs.
Ken McDonald [00:48:54]:
Rob?
Rob Campbell [00:48:55]:
Yeah.
Ken McDonald [00:48:55]:
Did you know just Nvidia still supporting x.org
Rob Campbell [00:49:03]:
yeah. So they're stuck in the past too. Gotcha.
Jonathan Bennett [00:49:08]:
Oh, goodness. All right, well, we're going to talk about something not at all controversial next system D and a fork. We're going to do it right after a quick break. We'll be right back.
Ken McDonald [00:49:20]:
Jonathan, thank you for that long break. Now everybody should be ready for me to talk about Saurav Roodruf latest article where he wrote about a developer, Jeffrey Satheran Sardina and I do apologize if I'm mispronouncing that, but about his response to the systemd project. Merging a pull request, adding a birthdate field to its user records. I think we covered that recently, didn't we? Well, the field is optional, can only be sent by an administrator, and systemd itself does nothing with the data. What was this, Jeffrey? Jeffrey's response. He forked it, creating liberated systemd to strip out what he considers surveillance, enabling code, keeping everything else intact and stay in sync with upstream. As it develops. The fork changes 12 files across five commits, all focused on scrubbing out everything related to the birth date edition.
Ken McDonald [00:50:31]:
That means not just the field itself, but also also the option to set a birth date via the home ctl. The relevant man page entries display code and test. According to Sourov, forks like this are meant to ignite conversation rather than ending up as a significant open source project. Now, you can find more information about this in Sirv's article. Jonathan, what's your take on this?
Rob Campbell [00:51:04]:
Ridiculous.
Jonathan Bennett [00:51:06]:
I do think it is kind of ridiculous. I understand why people do not like the age verification stuff, but it's just a birth date in a database. The systemd thing, you can make an argument that it's a good thing to have in systemd, even if there's not any, any laws regarding it. So I, I don't know. I, I don't really care for these forks that are just performative art. I, I think it is, yes. I think there's enough real work to be done that we don't need this
Ken McDonald [00:51:39]:
sort of thing with what it describes. Jeffrey did do a lot of work with this.
Rob Campbell [00:51:45]:
Yeah.
Jeff Massie [00:51:45]:
But looking at, looking at the code, there's just some stuff removed. It's not even that much. I mean it's seems pretty small. Now the thing too, when you talk performative art, anybody thinks, well, I'm going to jump on this. The article talks about it was 37 commits behind as when the article was written.
Jonathan Bennett [00:52:06]:
So
Ken McDonald [00:52:08]:
he's working well on keeping it up to date.
Jonathan Bennett [00:52:13]:
Yeah. All right, let's go ahead and move right into. Because I think these two go really, really go together. Go right into the systemd interview Rob has. Tell us, tell us the other side of the story.
Rob Campbell [00:52:26]:
Yeah, so you know, we've been talking about this for weeks and Ken even brings up another bonkers story about it. So there's more to, there's more to systemd than, than just the, you know, all the, the community gone bonkers over adding a feel to it. There's actually a human side of the story too. The developer at the center of this all is Dylan M. Taylor, a longtime open source contributor with work tied to the Arch Linux installer Nixos packages and other FOSS projects. What pushed him into the spotlight was seemingly a small change. I think, I think most of us agree. I'm not sure about Ken, but the optional, optional birth date field in systemd user base meant as a lightweight and optional way for Linux distributions to deal with the emerging state age verification.
Rob Campbell [00:53:36]:
I mean it's out there. Things had to be done. You can't just, I don't know. Anyway, Dylan is very clear that this is not, not what many people seem to assume it is. In his own words, a common misconception about this change is that it introduces age verification into Linux. You know, and we've been trying to debunk this for weeks, you know, he says it doesn't. He points out that there are no ID checks, no facial recognition and no third party verification services. Involved is just a simple addition of a birth date field as was pointed out every week here for the past month almost in fact, he says you can enter any date you want, even January 1, 1900.
Rob Campbell [00:54:34]:
You could be 126 years old. He also pushed back on the fear that this is the start of OS level surveillance. Dylan said, quote, moving towards OS level surveillance is definitely not the intention and argued that if anything, a simple self attested signal like this is relatively weak compared to the far more trackable data data most of you out there are using like browser cookies. He even made the point that not reporting a common value could make users more fingerprintable rather than less. But the story gets darker after that and this is where people really get bonkers. According to the interview with Dylan, he says the backlash went far beyond criticism. I mean, okay, you can criticize for. You shouldn't have, you should have gave it and put that on there.
Rob Campbell [00:55:35]:
It's acceptable. But he describes harassment, doxing death threats, hate mail and abuse. Abuse severe enough that he had to disable issues and pull requests requests across his GitHub repositories. How does this help anything? When asked how he he was handling it, his answer, you know, he says, quote, honestly not super well. He says people were posting his Personal information, harassing him across repos, placing takeout orders with his information and even sending Mormon missionaries to his house. There we go. Anyway, you can oppose these laws, you can hate the patch. You could believe Lennox should resist this kind of compliance entirely.
Rob Campbell [00:56:30]:
Dylan himself acknowledges the deeper concern and, and even says we may eventually see a split between more compliant corporate backed distributions and more freedom first independent ones. He argues that corporate backed distros may feel they have no practical choice even if they dislike these laws. I mean the laws are there, they're a corporation. If they don't follow them, bad things can happen. So you might as well, you know, have it there for them. But none of that justifies what happens to them. As a community, we're better than this. And you know, as techie nerds, as many of us Linux folks are, I kind of thought we were smarter than this.
Jonathan Bennett [00:57:24]:
Yeah, very, very well said. You read the details here. Apparently people have been putting his Social Security number, address and cell phone number onto pay spins and things like that. And it's just dumb. It's legit dumb. You may not agree with the actual pull request.
Rob Campbell [00:57:43]:
We all hate the law, I think. I mean not the law in general, but I mean these specific age verification laws.
Ken McDonald [00:57:49]:
Yeah, privacy invading laws.
Jeff Massie [00:57:55]:
Right, Yeah. I do have a bone to pick with you though, Rob. You could say you're 126 years old. No, a true Linux UNIX geek would say they're 56 years old. They were born on January 1, 1970.
Rob Campbell [00:58:10]:
Yeah, well there you go. He's the one who said you could put 1-1-1900 in there. So.
Jeff Massie [00:58:16]:
But you didn't bring it up. That shameful.
Rob Campbell [00:58:19]:
You, you can.
Jonathan Bennett [00:58:20]:
Apparently it's not Unix time based. That's. If you were to be angry about anything in this, it should be that, that it's not based on unix.
Rob Campbell [00:58:28]:
Why did you not use UNIX time? This is a UNIX based system. I am curious.
Jonathan Bennett [00:58:34]:
Yeah, yeah, yeah, this is, this is actually really disheartening that, that he's getting this kind of stuff.
Rob Campbell [00:58:43]:
I don't know if I keep, if I keep calling him stupid. I wonder if I'm gonna start getting some hate mail.
Jonathan Bennett [00:58:52]:
Rob, I've, I've, I've actually had this, had this thought before because we do live stream. This does go out on the Internet and I've, I've gone to my local police force and I've had a conversation with some of their guys before about like what do I do? I live stream. What do I do to make sure you guys don't show up at my door one of these days where somebody has tried to swap me. It's kind of a scary thing.
Rob Campbell [00:59:16]:
Maybe a little bit better off with my name being fairly common. I know of more than one person in my area with the same name.
Jonathan Bennett [00:59:25]:
So
Rob Campbell [00:59:29]:
just remember that if anyone has any ideas, I'm not the only one. So just, just don't do it. Just don't do it anyway, but don't do it indeed.
Jonathan Bennett [00:59:41]:
Yeah, I'm real disappointed to see that this is, this is what's going on with this. I think it's a huge overreaction to the technical issue here and I get why people don't like the law. Although looking around at how many different governments are doing this, it looks like it's going to happen.
Ken McDonald [01:00:01]:
Whether you like it or not is in some. It's going to be from end user perspective, especially if you're using a free and open source software. Unenforceable.
Jeff Massie [01:00:18]:
Well, you know, there's a lot of different initiatives out there trying to solve this problem and ideally how do you prove that you're a certain age without giving away a whole bunch of other private information? And there's people working on that and I. It probably is going to come down to something like that eventually just because, you know, and talking in America and a lot of the other world you have to be a certain age to go buy cigarettes or alcohol or things like that. So there's going to be that kind of. How old are you on the Internet? The thing is, you show your ID at the liquor store, the guy looks at it and goes, oh, okay, you're old enough. You're not. You put it on a website. You now have possibly exposed it to millions of people depending on how they keep their security databases and whatnot.
Rob Campbell [01:01:11]:
So in the future you just had to bring your Linux device when you buy those things and say here I see, I'm old enough.
Ken McDonald [01:01:18]:
Yeah, everybody's got a portable Linux device, don't you? Everybody would qualify as one.
Rob Campbell [01:01:23]:
I. My laptop, I think I remember I was scrolling at. I didn't really look at it too much but I saw somebody post a question on like some iOS group. I think it was asking if people in other countries have to. And I don't think this person was in the U.S. i don't, I don't know. But if they got, if the recent iOS update, if they also had to put their birthday in just with the recent one. So I mean of course we all
Jeff Massie [01:01:52]:
know Brazil doing something like this.
Jonathan Bennett [01:01:53]:
Oh yeah, Brazil, yeah. Brazil has Also added that requirement.
Rob Campbell [01:01:58]:
Brazil seems to have quite a few Linux distros too it seems like. I keep hearing about this one in Brazil and that one in Brazil. It's like man, there's a lot of Brazil, isn't there?
Ken McDonald [01:02:09]:
So I, I say the easy thing is, is just assume everybody's a minor and treat them that way.
Rob Campbell [01:02:18]:
I try to treat you like a kid all the time, Ken.
Ken McDonald [01:02:21]:
Thank you.
Jonathan Bennett [01:02:22]:
I mean so I mean think about this. One of the things, one of the things that's part of this is like we're starting to see this has been happening for a long time now actually but you're starting to see things happen because of it. Studies about how bad certain social media sites and abilities are for, for kids and I mean you know you've had kids commit suicide result to seeing things or having conversations on social media. There was, there was just recently it was in the news that something with Meta where they were held for Google I think yeah it was like New
Jeff Massie [01:02:57]:
Mexico I think or Arizona. Yeah.
Jonathan Bennett [01:03:03]:
A U.S. jury verdict against Meta and Google tee up a fight over tech liability shield. That's the Reuters Meta told to pay $375 million which is nothing for them but for misleading users over child safety. This sort of stuff is starting to work its way through the legal system. And so you know part of this is not just like, you know one side of this is different states are saying you've got to verify your age before you can go to someplace like Pornhub. Like that's, that's one side of it but the other side of it is a website like Facebook is going to be required to treat a user different if it is a minor versus a full on adult. And so you've got to have a way to verify that because there's now liability problems with interacting with minors. It's a, it's a big old ball of wax.
Jonathan Bennett [01:03:59]:
There's not, it's not just these simple freedom that, that I wish it were, I wish it was that easy. I don't want to have to verify my ID or my age or anything to use the Internet but I, I think that is the world that we are quickly going to be living in where you do have to verify for certain services, maybe not your identity but at least I would age because they're going to be.
Rob Campbell [01:04:24]:
I have seen it differently. I haven't seen an article on this but I've seen seen various people say that Facebook is one of the ones really pushing for these laws then I
Ken McDonald [01:04:36]:
think should include in these Laws, the requirement by any social media that once the law is in place to automatically close any accounts that have unverified. For unverified users.
Rob Campbell [01:04:53]:
Step two,
Jonathan Bennett [01:04:55]:
that would be. That would be. Yes, it would be.
Jeff Massie [01:04:58]:
So the reason Meta and other companies like that want to push this is because then it takes the burden off them. It all goes through the operating system then. So then they go, oh, we just have to have this little check and we don't have to worry about keeping people's data or how do we handle this? I mean ideally I would like to see some kind of government issued like blockchain kind of thing where it's just
Ken McDonald [01:05:23]:
like a Social Security number.
Jeff Massie [01:05:26]:
No, not so traceable. It's just a age bracket kind of thing that, okay, when you're over, you know, you're between X and Y age, you get this one. Once you're over like 21 in the U.S. you get this other one and there's no other identifiable information to it. It's not tied to your name, your address, your, you know, frankly
Rob Campbell [01:05:52]:
the way the laws are written today, having you manually input your age into your device in some ways is probably more privacy focused than like Facebook having to control that. Because like you said, taking the burden off Facebook. Now Facebook just uses whatever kind of API system gets created in the future to ping your device. Hey, how old are you? All right? They're old enough. You don't have to worry about it. And we're not going to ask who you are. Of course they know everything about you already anyway.
Jeff Massie [01:06:23]:
But, well, and you could look at it as just even more secure from a odds point of view. Okay, I've got, you know, Microsoft with my age and that versus and I have to worry about them getting hacked, but versus like oh, you know, X and Facebook, every site you've got an account.
Rob Campbell [01:06:44]:
Well, and you can lie on your computer too if you want. And that's, that's like your choice so far. It's not, that's not part of the law saying you have to tell the truth. So so far.
Jonathan Bennett [01:06:53]:
So so far. So far, yeah, I think that's the, I think that's the scare thing people are scared of is that it is going to turn into full on verification, like actual verification. The other side of that, which I had not thought of that until just now but like one could make an argument that it is much better to do that with your. As opposed to doing it with every website. Because you think about it, how do you do that with a website? You stand in front of Your webcam and you hold up your ID card like that is way worse on the privacy side than simply having it attested in your, in your os. So yeah, this is, this is unfortunately, again, it's unfortunately the world that we live in now. And it is not a simple problem that's going to be solved overnight. And if you're telling us suck for
Rob Campbell [01:07:35]:
a while, if you're telling everyone how old you are, then it's out there on social media, so many sites and one of them is going to get breached and your age is out there. If you don't want it out there, that's, that's not good. But if it's on your system, theoretically all, all they would have to say is are you old enough and not even necessarily get your actual age.
Jonathan Bennett [01:07:53]:
Yeah, yep, absolutely. All right, let's keep moving. We're actually going to take a very quick break and then we're going to talk gnome. We'll do that right after this.
Jeff Massie [01:08:06]:
A lot of times in the open source software, one of the biggest challenges isn't innovation, it's sustainability. GNOME over the years has had issues with their funding and we've done stories in the past how they've sometimes ran at a loss, funding wise. GNOME is now trying to take another step, trying to address that. The GNOME foundation has announced a new initiative called the GNOME Fellowship Program designed to directly fund contributors with working on the long term health of the project. Under the program, selected fellows will receive funding and when I say fellows, it's, you know, people not male will receive funding for a period of about 12 months with compensation typically ranging between about 70,000 and $100,000 per year. Now it depends on experience and your location, but the first round of fellowships is expected to begin around May 2026 and it will initially support a small number of contributors, potentially one full time fellow or two part time participants. A person can propose how much time they can dedicate to the project. So it's not like you're just saying I going to be, it's full time or nothing.
Jeff Massie [01:09:18]:
It's, you know, depending on the number of people and then time requirements, they can have a little flexibility in the support numbers. Now this, now this next statement is where I'm, I'm hoping it goes well. But they say rather than assigning strict deliverables, the program is designed around trust. Fellows are expected to identify high impact problems themselves and to work on improving the project's overall sustainability. For the first round, they want to focus. They, that's what they, exactly what they want to focus on is sustainability, infrastructure, testing, systems documentation and maintenance. You know, they don't often attract the same attention as new features, but they're essential to keep a project running. You know, they're the needed background tasks to keep things rolling, you know, and other areas like, you know, system builds, continuous integration, pipelines, developer tooling, accessibility and, you know, basically reducing technical debt.
Jeff Massie [01:10:19]:
Overall, it's work that's often critical, but rarely prioritized in a volunteer driven environment. One of the goals of the program is when you donate, the money is going to fund the code and not some other initiative. Sometimes, you know, when you, you give your, you know, it's kind of in a big bucket. This is, this is going straight, you know, you know where this is going.
Jonathan Bennett [01:10:46]:
I don't want to give the money that pays for the executives to get their first class tickets off to the in Bali.
Rob Campbell [01:10:53]:
How did I get in that bucket?
Jeff Massie [01:10:57]:
The initiative is funded through donations to the GNOME foundation and it particularly comes from its friends of NOME supporters. That funding model reflects now that it reflects a broader shift in how open source projects are thinking about sustainability, moving from purely volunteer driven contributions to more structured financial support. Now, applications for the initial inaugural fellowship are now open, targeting experienced contributors with a proven track record in a GNOME project. So that's where, because they're giving trust, you're going to have to have already really proven yourself to GNOME to be able to get on this, because to give trust, you know, you kind of had to earn some of it. Ultimately, the success of the program will depend not just on its first cohort, but on continued donor support and if sustained, could become a model for how major open source projects invest in their own future. You know, funding the work that keeps, keep this, keeps keeping everything else possible. Take a look at the two articles in the show notes for more details. And while I just put the two articles in the show notes, there are a ton of other articles on this all talking about this initiative.
Jeff Massie [01:12:09]:
And I personally will say this, while I love kde, I don't want to see anything happen to gnome because while, you know, GNOME and KDE are two of the biggest players out there, they feed off each other and other desktops for good ideas and it grows the ecosystem as a whole. So the more desktops out there doing new innovative things, the better, because you see what works and what works good together and what doesn't. And you know, people borrow from each other and it helps everybody. So, any particular thoughts from my co hosts?
Jonathan Bennett [01:12:45]:
I really like this, actually find the engineers that are actually doing the work and give them a reasonable wage to do it.
Jeff Massie [01:12:53]:
It's great. Yeah. And it really, you know that we've talked about it before. It's the support stuff that really is core to a project that's hardest to keep going. It's not flashy, it's not really cool, but it's. It's the daily grind that you got to have to keep everything rolling.
Rob Campbell [01:13:13]:
You know who apparently has a lot of money. Nobody's doing this story but Canonical gave. Is given $150,000 a year to the Rust project. I think it is. So
Ken McDonald [01:13:31]:
you're right. None of us covered that story this week, but it's out there.
Jonathan Bennett [01:13:35]:
Good for them.
Jeff Massie [01:13:36]:
Yeah, I like to see them given back.
Ken McDonald [01:13:39]:
But. Basically it comes down to. I hate to put it this way, but it's the only way I can think of phrasing it is following your heart on this.
Jonathan Bennett [01:13:55]:
No, you mean those people that apply
Ken McDonald [01:13:57]:
for it or that help to support it.
Jonathan Bennett [01:14:02]:
Oh, yeah, yeah. I mean if GNOME is. If GNOME has made a difference for you, then absolutely think about support this. The. The one. Jeff. Jeff sort of touched on this. The one thing that could become a problem.
Jonathan Bennett [01:14:15]:
Hopefully it doesn't. But they, they talked about not having deliverables and you. Someone will have to watch it to make sure that the people that are getting paid what is essentially a full time salary are actually putting an appropriate amount of work in and that can get dicey. It can be. I mean anytime somebody is salaried, that is a challenge to make sure that they're actually, you know, clocking in and doing the work as opposed to just phoning it in from home or whatever.
Rob Campbell [01:14:48]:
Yeah, I like, I like Gnome, but really these days cosmic is my DE of choice.
Jonathan Bennett [01:14:56]:
Oh, interesting.
Ken McDonald [01:14:57]:
Well, actually I. Back when I covered it, I started donating a little bit towards the open source lab when they were asking for help.
Jeff Massie [01:15:10]:
KDE and CodeWeavers.
Jonathan Bennett [01:15:15]:
Yep. I give a little bit towards our door in their. In their monthly thing. And they get their. Their arrangement is kind of like this now. Fewer people work on it and, and they are full time. I think there's two people that work on it full time and our door takes care of their salary. But it's the same thing.
Jonathan Bennett [01:15:31]:
People pay or donate to be able to get into that. And then they, they pay a wage to those guys that are doing the program.
Jeff Massie [01:15:37]:
I should clarify too. Codeweavers is a. Is a commercial product, but it. The work they do does a lot of feeding of wine. So the wine is kind of the trickle down of code weavers. So it.
Ken McDonald [01:15:50]:
Yeah, actually I think it's kind of both ways, isn't it?
Jonathan Bennett [01:15:54]:
No, I mean it definitely goes both ways.
Jeff Massie [01:15:57]:
Code weavers doing the majority of work because they. I want to say they fully support or hired. Was it Alexandria Juilliard? Isn't that the top developer of Wine?
Jonathan Bennett [01:16:09]:
Yeah, yeah, yeah.
Rob Campbell [01:16:10]:
Do you technically donate or are you just like a member? Like you have their product which. Which has a annual.
Jeff Massie [01:16:17]:
I have their product which is.
Rob Campbell [01:16:19]:
But you don't.
Jeff Massie [01:16:20]:
I don't really need it. But yeah, I don't really need it, but I'm. I like Wine, so I support it.
Jonathan Bennett [01:16:28]:
That's what makes it.
Rob Campbell [01:16:29]:
Officially you're just paying for the product, but since you don't use it, you're. It's a donation.
Jeff Massie [01:16:36]:
Yeah, well, because I use the downstream
Jonathan Bennett [01:16:42]:
now that we have cleared up that incredibly important detail, we're going to let ken Talk about OpenSUSE and the new Agama 19 installer. What's going on there? Ken?
Ken McDonald [01:16:54]:
Jonathan, as you said, we're going to be talking about what Marcus Neusser wrote about the OpenSUSE project releasing the Agama 19 web based installer for OpenSUSE, Tumbleweed, Slow Roll and Micro OS operating systems. This major release introduces several new features, small improvements and bug fixes. It introduces the ability to install some SUSE Linux distributions in so called installation modes such as Standard or Immutable, and adds the ability to use SSH public keys to authenticate the user. You now have the ability to install into an existing LVM volume group and add new physical volumes to an existing volume group, and even a new option to download the current installer configuration in JSON format. Under the hood, the Agema 19 installer introduces a major architectural revamp that allows the core of the installer to be controlled through a consistent and simple API while keeping the JSON based configuration format fully backwards compatible with previous versions of the Agema installer. As always, I recommend reading Marcus's article for more details about Agema 19 since I'm only taking time to cover some of the highlights.
Rob Campbell [01:18:31]:
Is it a Gamma or a gamma You've been saying both of them.
Jonathan Bennett [01:18:34]:
It's the Internet. Nobody knows how to pronounce anything.
Ken McDonald [01:18:38]:
It's the Internet. Twisting my words too.
Jonathan Bennett [01:18:41]:
Yeah, absolutely. So this is their new web based installer, right? Where everything, everything's HTML and JavaScript. Fun stuff. Yeah, I remember when that first became popular and people really hated on it. Like that's dumb.
Ken McDonald [01:18:59]:
Why?
Jonathan Bennett [01:19:00]:
It sort of makes sense. Just build everything the Same way. Looks like it works.
Ken McDonald [01:19:06]:
I'll let you know the next time I do an installation of OpenSUSE.
Jonathan Bennett [01:19:10]:
There you go. All right, well, there's one more story that's been hanging around this week that we desperately need to talk about. And we're going to get our AI fix for the day. And that's something going on at the Linux kernel which the Linux guys, the, the kernel guys have always been sort of the early adopters for tooling and they've been using AI. They've been using AI for a long time. Even before the current AI craze, there was LLM and inference and stuff that they were using, evaluating bug reports and evaluating pull requests and all of that. So Stephen J. Von Nichols sat down with Greg Kroll Hartman at apparently a kubecon in Europe this past week and they had a conversation about AI slop in the kernel and bug reports.
Jonathan Bennett [01:19:59]:
And so this is something that we've talked about here a few times. You know, we've talked about Daniel Stenberg over at Curl closing down their bug bounty program because of exactly this problem. Because it was so easy to convince an LLM or to ask an LLM to write a convincing bug report. Whether or not there was actually a bug there or not, the LLM could write a convincing bug report. And so there was this, there was this ongoing problem and it's still a problem. I even even really reading what they talk about here, I'm convinced that it is still a problem that it is so trivial to use an LLM to write a bug report even though there's nothing there to be found. Well, the kernel guys have actually, they've started seeing something different and that is that these LLM tools, these AI tools are instead of just finding slop, they are now finding real problems, real bugs. And they said about a month ago is when they really noticed the change.
Jonathan Bennett [01:21:02]:
And so it's, it's not just, it's not just the kernel that's seeing this. According to, according to Greg, it's a lot of different open source projects and these AI discovered bug reports are actually decent. Now I'm not seeing a whole lot of AI bug reports on the projects I'm in. We're still seeing a lot of AI pull requests and sometimes those are funny and take a lot of time and do not get merged in the end we shall say we also see something kind of funny. A lot of people are using LLMs to write feature requests. We got one the other day that it's like someone must have said, hey, LLM ask for this thing, but ask for it in a low key way because you know, the, the, the issue started out like hey fam, let me cook. And it's like, oh my goodness, that's the funniest thing ever. It was actually a decent idea, which is what's funny.
Jonathan Bennett [01:22:01]:
It's just the way that it was presented was you knew it was an LLM that was told to write it in this silly way. So the kernel is seeing this, that a lot of these bug reports are actually real and in some cases they can set an LLM on them and get reasonable code fixes too. And this is kind of what I've seen in various places. And what we've seen in the projects is someone that knows what they're doing can use an LLM to get really good results. And the problem is that it is so easy for someone that has no idea what they're doing to use an LLM to get not good results, but have those results look good at first glance. And so I think part of what is going on with the kernel is that to even be able to report a bug in a way that someone's going to look at it, you kind of have to have a clue because it's all email based and they have some really good spam protection tools. Like you've got to write the email in just the right way for it to make it through and someone to actually look at it. And that, that kind of adds just the right amount of friction.
Jonathan Bennett [01:23:10]:
I think that they are not being overrun with LLM stuff with, with the AI slot. So I don't know. Do any of you guys have your sort of fingers on the pulse of open source projects? Am I the only one that is really sort of running a project right now?
Rob Campbell [01:23:23]:
I think that anything big, I dabble in little things here and there but nobody, nobody ever downloads them or even goes to the page. And then after two years I'm like, well, nobody wants this. I guess I'm just gonna abandon it.
Ken McDonald [01:23:41]:
Most of my open source projects are ones I use myself. What I found interesting from this is that the Google's review system having become a Linux foundation project,
Jonathan Bennett [01:23:59]:
that one's interesting
Ken McDonald [01:24:00]:
means that that's going to make that available for all these open source projects to use.
Jonathan Bennett [01:24:06]:
Yeah, I'm hoping someone figures out a way to integrate that into GitHub. That would be, that would be very nice. And GitLab, I mean GitLab too, but let's be honest, maybe more open source projects on GitHub. No, I don't think so. I don't think that's going to be integrated into Git. It does not actually make any sense what age.
Rob Campbell [01:24:24]:
Let's put it.
Jeff Massie [01:24:24]:
GitHub.
Jonathan Bennett [01:24:24]:
GitLab.
Jeff Massie [01:24:27]:
And I think you really, you hit the nail on the head, Jonathan, when you said it's all about how the person uses it. Because you could have a good analogy. Just saying it's like a crayon. You give some people a crayon and they're going to come up with beautiful art. And some people, it's just a scribble.
Jonathan Bennett [01:24:44]:
Well, there's the show title. It's like crayon.
Rob Campbell [01:24:46]:
I also ironically funny. I don't know when you said that the Linux kernel has always been early adopters with tooling, yet they still use mailing lists.
Jonathan Bennett [01:25:00]:
Yeah, yeah.
Ken McDonald [01:25:02]:
If a tool works, keep using it.
Rob Campbell [01:25:04]:
I mean, we all know there's better tools right there for that too, though.
Jeff Massie [01:25:09]:
And, and I, I wasn't actually.
Jonathan Bennett [01:25:10]:
And they haven't used some of that.
Jeff Massie [01:25:14]:
Yeah, I was gonna say it. I wasn't actually even trying for a show title, but it's a good way to explain to people that don't really understand. It's like, well, it's good, it's bad. It's like, well, look, it's, it's the artist behind it. And if you use that analogy, it's one that a lot of people understand. And it's like, you can't just randomly go, here's the tool, go at it. You know, it's like, no, there's.
Ken McDonald [01:25:36]:
There's a sledgehammer. Go make me this. Fine skip.
Rob Campbell [01:25:39]:
I would say there is still some difference between a crayon, like for example, AI if you're making something really small and you get lucky, you can still make something that works well. And even without knowing what you're doing, it's the big projects that you're gonna struggle with.
Jonathan Bennett [01:26:02]:
It's an analogy, Rob.
Jeff Massie [01:26:04]:
I like analogies like fat kids like cupcakes.
Jonathan Bennett [01:26:10]:
Oh, geez, cut that one out, guys. We don't want that one as part of the show. That is not going to be the show title.
Jeff Massie [01:26:19]:
Actually. One of the nice things with AI is you can also have fun where you can say, review the work I've done this last week, but give it to me in a comedy roast format or. Or talk like a pirate.
Jonathan Bennett [01:26:32]:
Yeah. Yes, yes. Or be low key. Yeah, we've gotten some of those.
Rob Campbell [01:26:37]:
I do hate that sometimes when I ask to do something, it starts getting all flowery and wordy with things like Just, I don't talk like that.
Jeff Massie [01:26:47]:
Well, I had one. It was. And I wish I would have kept it. I. I was asking for something rather complex from AI and it replied back something like, I would pick up a cocktail and I would look at the person as I swirled my drink, and it's like, that has nothing to do with what I mean.
Rob Campbell [01:27:09]:
Like, I don't know what you're using, but all mine's saved. I can go back to forever. And everything I use.
Jeff Massie [01:27:17]:
Well, I was just firing a bunch off and I cleared it, and I
Rob Campbell [01:27:21]:
guess there's a few that I clear is like, okay, I don't need to save that one. But.
Jeff Massie [01:27:24]:
Yeah, because it was like, what the heck happened? Let me redo this. And then. Then it came back with what I needed, but it was just. It was like, I've never seen it derail that far before.
Ken McDonald [01:27:39]:
You've got to shut them down occasionally and start them up again.
Jonathan Bennett [01:27:45]:
If I'll board anything from reading sci fi novels about AI you absolutely can't just let them run forever.
Jeff Massie [01:27:53]:
Yeah, and you do, can. You have to reset them. Because one, it keeps the context. So it's looking back at the things you've already previously asked it. So it's trying to incorporate some of that knowledge. And then as it stores more, it gets more of those tokens. It builds up, and then depending what you're using, it starts getting too big and it.
Rob Campbell [01:28:12]:
Yeah, I mean, falls apart. It. It doesn't go away, though. I mean, at least some of them, you can delete it and whatever. And you talk about something two years down the road, and then it brings up something you said two years ago. I was like, I. I don't want you to bring that into this. That's old news or whatever.
Jonathan Bennett [01:28:33]:
Well, so is. That's. That's not necessarily where it's saving your context. That's just where your previous conversation is now part of his training data. It is funny, though, when you think about it. Like, how many of the. How many of the sci fi novels talking about AI had that as part of their plot points? And now that we're using AI for real, it is a thing that we're seeing that, like, these things do get weird if their context goes on for too long. It's trying to keep too much in state at the same time.
Rob Campbell [01:29:02]:
Ask your AI Sometimes. All right, Sometime, what do you know about me? And see what it says. I've done it.
Jonathan Bennett [01:29:07]:
Oof. I'm sure. All right, let's. Let's move on to our Tips. We've talked about a whole bunch of new stories. It's getting late. Let's get into some tips. We're gonna let Rob go first, but first, last, last, first.
Jonathan Bennett [01:29:21]:
We're gonna take our first, last, last, first break.
Rob Campbell [01:29:25]:
All right, so last week I showed you, oh My Posh, the way to spice up your, your terminal. Well, and I told you that I would show you an easy way to install it and customize it. So that's what I'm doing this week. It's called OMP Manager or oh My Posh Manager. So with that, it is a cargo app. Check the show notes, install it via cargo, and then once you install that, super easy by running the cargo installation, OMP Manager, I think. But anyway, run OMP Dasher. And it will.
Rob Campbell [01:30:09]:
The first thing it'll bring you up, I mean, it'll bring up a menu where one of them is a quick setup. So if you just do that, it will go through the steps. Install My Posh complete. Install Nerd Font complete. Choose a theme and you choose your theme. And when it's complete, it's complete. Configure the shell, you pick yourself, bash zfs, fish, and then once that's installed, you can always come back and browse the themes and install other themes. So if I browse the themes, there are a lot of themes in here and they're all just different colors.
Rob Campbell [01:30:49]:
But for those watching, you can see all these different themes and you know, I'm only in the Bs here, but all the different themes where you can oh my Posh, your terminal
Ken McDonald [01:31:04]:
now.
Rob Campbell [01:31:06]:
You can install other fonts if you want. I actually haven't. Maybe you can't install their fonts. I guess they only have one font. I haven't even looked at that because I didn't really care about fonts. I'm not a font guy. Update All My Posh. You could reset to the fault if you.
Rob Campbell [01:31:21]:
You don't want to do it anymore. But anyway, that is, that's it. It's OMP Manager to easily set up All My Posh and customize your terminal. And you can do it once and you get bored. Next week you can do a different theme and do a different theme the week later. And you just keep that thing, keep that terminal interesting, fresh.
Jonathan Bennett [01:31:43]:
Keep it fresh, everyone. Nice.
Rob Campbell [01:31:47]:
Nice.
Jonathan Bennett [01:31:47]:
All right, cool. Up next, Jeff doing some music.
Jeff Massie [01:31:52]:
I've got Fluidsynth as my command line. Tip of the week. And Fluidsynth is a cross platform real time software synthesizer based on the Sound Font 2 specification. SoundFont for those who don't know, is a file format for sample based instrument sounds. So that's where you keep little samples of sounds and then the MIDI is which. So let me back up. Fluidsense supports real time MIDI effects controls. So when the MIDI is how you play those different sounds found in your sound font file.
Jeff Massie [01:32:31]:
I hope that makes sense. Fluidsense can also be used as a shared library for embedding in other applications. It can play MIDI files and has a command line shell. In fact, you know, a lot of other applications use FluidSynth for audio synthesis. FluidSynth itself though, is only a command line tool, though there are other GUI applications which can be used as a graphical interface for it. The simplest way to use it is. So the simplest way to use it to see about getting a sound file would be on the command line. You would do Fluid spins Fluid synth space your sound file so that, that's your, that's your sound font file and then space and then a MIDI file.
Jeff Massie [01:33:22]:
Now most of the time you're not going to be using a MIDI file to play your, play your sounds and music. Instead you're going to be using a program or an interface to pass MIDI events to the program. But this just lets you see that things are working and they do have a link to some test files you can use just, you know, so you can play with it a little bit. Take a look at the link in the show notes for the link to the GitHub page. And it also has a link inside the GitHub page to their wiki pages where there's a ton of documentation on setup and usage and getting the most out of it. And that's where you also find some of those demo files that you can play with and have a lot of fun with. So happy music making.
Jonathan Bennett [01:34:07]:
Yeah, one of the, absolutely. One of the really fun things you can do if you're into old school like DOS games, you can run a DOS game inside of DOSBox. And if it, if it wants to give an MIDI file output, you can actually route that MIDI around inside your system and so you can route it to a running Fluid Send and then out your speakers. Some, some very fun stuff that you can do with fluency and midi, especially
Ken McDonald [01:34:31]:
if you've got a MIDI keyboard to use.
Jeff Massie [01:34:34]:
And MIDI used to be really big because back in the, in the days you didn't have the storage and the memory. So you would just sample a few different sounds and then you could use, by playing those specific sounds over in a different order, you could have A whole big music sounding production. But it was a very small file because it only used, you know, eight notes or something like that.
Ken McDonald [01:35:01]:
Four notes?
Jonathan Bennett [01:35:02]:
Yeah,
Jeff Massie [01:35:05]:
yeah.
Jonathan Bennett [01:35:06]:
However many you wanted to use. It's only four chords.
Ken McDonald [01:35:09]:
Yep.
Jonathan Bennett [01:35:09]:
Every song is only four chords. Anyway, Ken, talk. Let's talk about meta. Meta information, probably in graphics, if I understand this exactly.
Ken McDonald [01:35:19]:
I'm actually talking about the exiftool is the command line tip I'm bringing this week. It's basically a command line interface to the image colon colon EXIF tool used for reading and writing meta information in a variety of file types. The file is one or more source file names, directory names, or can be the standard input or the dash when you type it on the keyboard or on the terminal. But the metadata is read from those source files and printed in readable form to the console or written to an output text files with a dash W. Basically you can use EXIF tool to read a photo's current metadata or to write updated metadata to the photo. It even allows you to copy or export the metadata into a comma separated value file format, HTML or JSON. Let me go ahead and bring up my console here and I'll give you an example of using it with a ping file I have. For those of y' all listening, I've got XF2 followed by the name of the ping file.
Ken McDonald [01:36:38]:
I'm not going to bother reading that because it's a long one, but it comes back, gives you the version number for the EXIF tool I'm using. It gives you the file name, the directory I'm accessing it in relation to the directory I'm accessing it from. In this case it's just a dot. The file size, file modification, access date and time and the file inode change, date and time. It was interesting when I was doing this that the original file had a different inode date compared to the last. The initial modification date. But it also goes down, gives you file permissions. Type, type extension, the MIME type, image width and height, bit depth.
Ken McDonald [01:37:35]:
In this case it's 8 bit color type, which is RGB. With alpha compression, it's deflate, inflate. Since it's a PNG file, the filter is adoptive. The interlace is non interlaced SRGB rendering an image size and mix of pixels. Now what you can do with XF tool is say you've got a group of files and you want to add comments to the metadata. You can do that by typing in the XF2 either before or after the file name. You can put in quotes. In this case, I'm putting dash comment equals sunset.
Ken McDonald [01:38:21]:
That's just to add the comma value to it. And now we go back and read that file and you'll see you've got a comment line in the metadata. This has the value of sunset.
Jeff Massie [01:38:37]:
Nice.
Ken McDonald [01:38:39]:
So it's a way you can go in and add additional information to some of your graphics files. You can even do it with videos
Rob Campbell [01:38:48]:
comment that you did in there. That isn't a special field. Like you could have said comments or my stuff or anything you wanted there. Or is comment or dft.
Ken McDonald [01:38:56]:
Yeah.
Rob Campbell [01:38:56]:
Okay.
Ken McDonald [01:38:57]:
You could say dash comment or you could say dash my comment.
Jonathan Bennett [01:39:03]:
Yep, nice.
Ken McDonald [01:39:04]:
Equals. And then the value is going to be cool.
Jonathan Bennett [01:39:09]:
All right, good to see. I've got a. I've got a bit of a demo here that we're going to try. And I'm trying it as a demo because I intend to run this as a demo at the Ubuntu conference. And so you guys get to get it live, at least part of it. And that is we're going to do some Grafana work. So if you're not familiar with Grafana, it is basically a toolkit for making dashboards, for visualizing data. So I've done this on the desktop behind me.
Jonathan Bennett [01:39:40]:
I've not done it on this machine yet. So this is Fedora specific. Although these are going to be similar for basically any distro. Um, we're just going to install it. And of course, you know, this is not. Not always a great idea. It makes it a little easier on typing stuff in. So we're going to grab the install.
Jonathan Bennett [01:40:02]:
We get to see how fast my Internet connection is. As soon as it's done, we can use System CDL to start it. And systemctl start Grafana Dash Server and systemctl enable Grafana Dash Server to get it to auto start each time. Hold on. We're running the Selinux script, which that's going to be important here in just a second. You're on an Selinux system like Fedora.
Rob Campbell [01:40:28]:
Don't you think that's a little blasphemous running DNF at a Ubuntu conference?
Jonathan Bennett [01:40:36]:
So I have thought about doing this on, but we'll see. So Grafana Server. All right, it is started. And then I'll click run a status. It is there. Storage Starting. Starting. It's doing its thing.
Jonathan Bennett [01:41:03]:
All right. And I believe that's 127001 8000. Nope, I can't remember it. Let's See, I've got it in my notes. 3000 colon 3000. Can't type while live. Here we go.
Ken McDonald [01:41:28]:
I copy and paste.
Jonathan Bennett [01:41:30]:
Let me replace with the tab. All right. Welcome to Grafana. The username and the password is just admin by default. And yes, it's offering to add to LastPass. No, I don't want you to add to LastPass. And then of course Grafana, the first thing it does is it says update your password. This is not secure.
Jonathan Bennett [01:41:54]:
And I know I'm going to hit skip so it brings you right in. So here is the initial view of Grafana. Obviously we don't have any data sources or anything and I'm going to walk through adding a first data source as well. And so there's a bunch of different options here. There's Prometheus and Graphite. Prometheus is one of the most common ones but we're actually going to up and add a new connection and you get all of the different plugins that they've got here. The one that I'm interested in is actually MQTT way down here and we can just hit the install button here and it will install it as a plugin and then we can say add new data source and we are now ready. If we had an MQTT server running we could absolutely.
Jonathan Bennett [01:42:39]:
Well, Mosquitto is just an install away. Right? Let's. Let's see if we can do that real quick too. All right. DNF install Mosquitto. Mosquito is an open source MQTT server. Yes please. It's going to go through.
Jonathan Bennett [01:43:01]:
Grab it, do all the things, then system CDL start Mosquito. It's running. That was fast. Again we'll do a status starting. It's running. Config loaded from Mosquito conf. All right. I happen to know that Mosquito is going to open up at TCP and localhost in Port 1883.
Jonathan Bennett [01:43:32]:
So let's go back to Grafana. Wrong button there. We go back to Grafana and we'll
Jeff Massie [01:43:46]:
say
Jonathan Bennett [01:43:49]:
TCP 127.0.0.1 and then Port 1883 and we'll scroll down and we'll say save and test and it doesn't work. Error connecting to MQTT broker Network error. Well, I also have another, I have another pop up here that you guys can't see because I am just capturing the, I'm capturing the tab, but I got an SE Linux pop up and I, I knew this was coming. Is Selinux blocked Grafana from being able to talk to Mosquitto and this is actually fairly easy to work around. So we're going to tail the audit log file and grab the last say thousand lines from it and we're going to pipe it through a command that you get in Fedora and other Red Hat based systems called audit to allow. Yeah, and so it's going to go through and say, hey, here's the Selinux rule that will fix this. Now in this case it's saying just let Grafana talk to TCP sockets. If you were to do this on a real like deployment system, you would want to, you would want to scope that down and say, no, no, we want only Grafana to talk to mqtt, but for here it's fine.
Jonathan Bennett [01:45:20]:
So the full command to run this
Rob Campbell [01:45:25]:
is not that
Jonathan Bennett [01:45:29]:
retail it to auto to allow a M and then the name of whatever it is and I've just called it Grafana tcp. I'm going to run that and it will create a local file and then it gives you the command right here to make this policy active SE module I grafanatcp PP and we run that. It's going to take just a second. Let it do its. Let it cook. Let it cook. Selinux is slow sometimes. All right, it finished it.
Jonathan Bennett [01:46:02]:
Now let's kick back over here to the Grafana interface and see if it works. Now save and test. Yeah, MQTT connected. Then we can start building a dashboard. It's there and you can see under the data sources. You go in and fiddle with it. So I think we're going to stop there for the moment and probably next week we will look at this a bit more and take the next step in doing something actually interesting with grabbing some data from somewhere. And I already have ideas about what we're going to do and putting it on Grafana.
Jonathan Bennett [01:46:51]:
This is something that I've seen people do and I've never done myself. And so this is the way I learn things and I'm documenting it so that, you know, five years from now when I'm going, how did I do that Grafana thing? I come back to my notes from
Jeff Massie [01:47:03]:
this and it is available on Arch as well in the main repositories.
Jonathan Bennett [01:47:08]:
Grafana is going to be available most places. Mosquito is going to be available on most.
Rob Campbell [01:47:12]:
Grafana is pretty big.
Jonathan Bennett [01:47:13]:
It should be.
Ken McDonald [01:47:14]:
So you set up the server on one system and you can access it from other systems.
Jonathan Bennett [01:47:21]:
You could, I would have to go in and do some fiddling Firewall with the configuration of each of these tools and the firewall of each of these tools. It's a whole lot easier to get this working all on one local machine. But yeah, you can, you can do it from other machines too if you really want to. So yeah, we'll build on this in the future. We'll get a pretty cool dashboard. And I probably won't demo all of this at the Ubuntu place. At the Ubuntu conference I will just demo the end result of it. But again, I had to put the
Rob Campbell [01:47:52]:
work in it makes a lot of sense or very use makes it more useful if it's, if it's on a server and actually, you know, run it all the time than on your desktop.
Jonathan Bennett [01:48:02]:
Well, I mean it depends on what you want to do with it, right? Like if you're just wanting to be able to look at your local machine and see what your RAM is doing. Well then maybe on your, on your local machine. But yeah, if you want to, if you want to watch more than one system or you want to be able to get to it from anywhere, then yeah, put it on a server somewhere. But all of this stuff is going to be applicable. It's basically the same process, just a few more steps to get it accessible more globally. So yeah, there we go. That is that, that's command line tips. We'll let each of the guys get the last word if they want to or have plug something, what have you.
Jonathan Bennett [01:48:36]:
Rob, Rob is in the first place slots and so we will let him, we do it like we're going around the clock and so we'll let Rob go first.
Rob Campbell [01:48:47]:
Well, I mean for my screen we're going the other way. I was like, well from me it's this way better.
Ken McDonald [01:48:52]:
Remember mirroring?
Jonathan Bennett [01:48:53]:
Yeah, yeah, yeah, Yes. I, I, I've had.
Rob Campbell [01:48:57]:
Anyway, all right, first I want to start off. I feel like I missed thanking somebody who donated a coffee to me or something. I, I don't remember thanking them. I was just looking at this during the show here to make sure I didn't miss anybody. And I already thanked Mike last time when he donated to, to me, George, Paul and Ringo here. And I don't know if it's the same mic down below who's also donated in the past, but there's one right in there that says love the untaled Linux show from Donner that I don't think I thanked him. I don't even think I noticed. Somehow, somehow I missed that they donated.
Rob Campbell [01:49:37]:
I opted dig in some time and see how longer they donated. But thank you for the coffee I will appreciate it. And if I already thanked you once, well, I guess it's a bonus thank you. And for those who want to get thanked live here on the show by me, all you have to do is go to my website, robertp Campbell.com on that website, you can find links to My Little buns, to my LinkedIn, my Twitter, my Blue Sky Mastodon, and that. I know I can't think of a good word, but the amazing, wonderful place where you could donate a coffee to me, and I will thank you live on the show, whatever you put on there. You can be anonymous or you could put your first name. You can put your whole name, whatever you want. And I will thank you, I guess, as long as it's appropriate.
Rob Campbell [01:50:26]:
Don't. If you put an inappropriate name on there, if the system even lets you, I don't know, I'll just thank somebody.
Jonathan Bennett [01:50:35]:
Yeah.
Rob Campbell [01:50:35]:
So anyway, come connect with me and donate a coffee or, you know, whatever.
Jonathan Bennett [01:50:42]:
Pretty cool. All right, Jeff.
Jeff Massie [01:50:45]:
Not much. Not much to say. Those of you that have do want to connect, you can go through LinkedIn, but you have to be a friend with Rob first or else you won't get through because I get too many of them. Other than that. And I thought about reading this in the theatrical voice. It's just another haiku, but I'll give it to you standard. Chaos reigns within. Reflect, repent, and reboot.
Jeff Massie [01:51:12]:
Order shall return. Have a great week, everybody.
Rob Campbell [01:51:16]:
Chaos reigns within.
Jeff Massie [01:51:19]:
I thought about that.
Ken McDonald [01:51:22]:
Actually. That worked better for me the way you did.
Rob Campbell [01:51:26]:
Didn't work for me, Jeff.
Jonathan Bennett [01:51:30]:
Fun. All right. And then, Kim, just want to remind
Ken McDonald [01:51:33]:
everybody, especially for those of y' all who are in the US where tax season is coming to an end in what, three weeks? Make sure to back up all those records.
Jonathan Bennett [01:51:51]:
Yeah, that sounds about right. Print them out. I actually printed mine out.
Rob Campbell [01:51:55]:
Kill a tree.
Jonathan Bennett [01:51:57]:
All right.
Jeff Massie [01:51:59]:
Yeah.
Jonathan Bennett [01:51:59]:
A whole tree for 15 sheets of paper. Right? Anyway, thank you guys for being here. It is. Yeah. Real small tree. It's been. It's been a blast. And thank you everybody for watching.
Jonathan Bennett [01:52:10]:
We appreciate it. And we will be back. We. We appreciate those. Whether you watch or listen live or on the download. And we'll be back next week on the Untouched, A Living Ship.
