Mint, Montana, & Mandriva
Loading summary
Jonathan
This week we celebrate Mint finally getting Wayland ready for Cinnamon Hannah Montana. Linux is back and OpenMandriva has a bit of a community problem. Didn't love to see that one. Oh, and they're changing the way AI attribution works in the Linux kernel. There's a new FFMPEG front end and
Rob Campbell
a whole lot more.
Jonathan
You don't want to miss it, so stay tuned.
Bobby Borsel
This episode is brought to you by Black Hat usa. If you listen to this show, you go deep on the technical detail. Well, so does Black Hat. For nearly three decades it's been where the security industry's most rigorous research gets presented and pressure tested. More than 100 hands on trainings taught by practitioners who've actually deployed in live environments, not lecturers reading from slides. And hundreds of peer reviewed briefings that go well past the overview into the real work across the four areas defining security right now. AI and autonomous threats, cyber conflict, systemic resilience and identity. This year, Black Hat's briefings pass includes all keynotes and main stage access plus business hall entry. You also get breakfast, lunch, Arsenal live tool demos, on demand session access and admission to the midnight in the war room screening. Black Hat takes place from August 1st to the 6th in Las Vegas. If you want the depth this show gets into in person with the people doing the work, this is the room. And we'll be there too. Prices rise on July 17th, so book before then. Use code TWIT for $200 off your briefings pass@blackhat.com US-26 that's B L A C K H A T.com US-26
Jonathan
podcasts
Rob Campbell
you love from people you tr.
Jonathan
This is TWIT. This is the Untitled Linux Show. Episode 262 recorded Saturday, July 11th. Spewing common sense. Hey folks, it is Saturday. It is time for the Untitled Linux Show. That's where some of your favorite geeks get around and geek out over Linux hardware, software, all kinds of fun stuff. It is not just me. We have the, the terrific trio today. We got Rob and we've got Ken, that other guy. Yeah, welcome back Ken. We know we all took off last week from, from 4th of July. So you know we've got quite a bit of news that we can cover here. I am sure we will forget something and if you notice that we do, feel free to remind us you guys that are live, we can add stuff to talk about or you can send us, you can send us wonderful fan mail and remind us that we forgot things. But yeah, it's going to be fun. It's good to be back.
Rob Campbell
I've got.
Jonathan
I'm not sure what's going to happen next week. I think we're going to have a show, but I'm going to be on location in California and so I don't know how that's going to work, but I'm going to be at the Open Sauce Conference.
Ken
Especially if Jeff isn't back.
Jonathan
Especially if Jeff's not back. Yeah, that could be real interesting. Maybe I'll just have to find a quiet spot and do it from there like I did from the airport there a few weeks ago. Anyway, let's get started with the news and the content for today. And I believe we're gonna let Rob go first. And Rob is feeling kind of Minty.
Rob Campbell
Yeah, so after years of telling people not to use or at least implying that people shouldn't use Linux Mint and that it isn't currently suitable for public consumption unless you know how to migrate the security issues, it looks like they are flying finally going to be ready for general use by the end of this year. So there goes my Mint hating Persona. I've spent years criticizing Linux Mint, at least a couple give or take, and Cinnamon for clinging to Xorg X11, whatever one you want to call it. You know, clinging to that long after most of the other Linux desktops had moved away and moved on to Wayland. X11 was designed for a completely different era of computing, and my argument has always been that continuing to build a modern desktop around it creates unnecessary security concerns, additional complexities, and limitations when implementing newer graphical features such as hdr. I have probably been one of Linux Mint's biggest critics over this, and I didn't, you know, I didn't think Cinnamon would be Wayland ready until what I've been telling people is sometime, probably until 2028 at the earliest. So I thought we had a little time here, but looks like I may have been wrong. According to the Linux Mint team, Wayland will no longer be considered experimental in the next version of Cinnamon. Now, due to how they extended out, that's going to be a few more months. But they now say that the Wayland experience feels solid and is almost on par with X11. Both Wayland and X11 will be officially supported when the next major Linux release, Linux Mint release arrives around Christmas of this year. This isn't just a Linux Mint removing the word experimental and hoping nobody notices the bugs. The developers have been fixing many of the problems that kept Cinnamon Cinnamon's Wayland session from being practical. They have improved windows and menu positions, adding proper focus, stealing prevention, fixed numerous crashes throughout Cinnamon and it's a muffin window manager, improved support for multiple monitors and hardware KVM switches, and added better hardware acceleration across the desktop. Cinnamon is also getting full high DPI support, session fixes, improved Nvidia acceleration, and the ability for administrative applications to run as native Wayland clients instead of falling back to X Wayland. This is starting to finally sound like an actual modern desktop, rather than an experimental compatibility layer attached to a desktop environment that would really prefer anyone just stop asking about Wayland. There are still some unanswered questions. Linux Mint has confirmed that Wayland will
Ken
be
Rob Campbell
or has not confirmed that Wayland will be the default session and x11 is is not going anywhere yet. But Mint plans to fully supporting both options for now, so I may still have a little material left. If they ship a fully supported Wayland session and then continue pushing everyone to X11 by default for a while, we'll see what happens. But I also have to give the Linux Mint team credit. They recently changed their development process, their release cycle, so they could release less frequently and spend more time focusing working on major features instead of constantly preparing for another incremental release. At the time, I question whether that would actually help help them catch up or merely just allow them to fall behind more slowly.
Jonathan
I guess.
Rob Campbell
But based on the progress we are seeing with cinnamon 6.8, it appears the extra development time may have worked. That leaves me with one problem. Who am I supposed to criticize once Linux Mint finally modernizes? Maybe I guess there's always canonical. I could fall back to them again, but I don't know. They've been doing some good things too. I'm not giving them complete pass yet, you know. They still have to deliver the release, and almost on par with x11. It's not quite the same thing. As proven, it's ready for every user because you know, people still have their complaints. But but this is is is the most promising update I have seen from Linux Mint in years, and I'm generally gen genuinely happy to see them making the progress. 2027 may finally be the year of the of Linux Mint.
Jonathan
It might be the year that we can convince Rob to try it at least.
Rob Campbell
I've tried it once.
Jonathan
Been there, done that.
Rob Campbell
I used it for a while, but maybe I'll give it another shot a more
Ken
I remember which Ubuntu derivative I'd used when on my Raspberry PI 5 when I was using it as my podcasting desktop
Rob Campbell
ZORN
Ken
no, it wasn't Zorin.
Rob Campbell
I know there's thousands. I don't know which one you're using.
Ken
I'm thinking mx.
Rob Campbell
Yeah, maybe.
Ken
Yeah.
Jonathan
So Rob, if you're looking for somebody to rag on, you could try taking the complete 180 and go rag on Microsoft for a while. Yeah, those guys.
Rob Campbell
Anyway, this is a Lennox show. We try not to talk about them too much.
Jonathan
When did you get here?
Bobby Borsel
Rob,
Jonathan
this statement is not jive with your performance over the past three years.
Rob Campbell
Well, except for when they do subtilinic.
Jonathan
See, they talk about them anyway.
Ken
Isn't it four years now?
Jonathan
I don't know, however long it's been Anyway, so there is another very interesting Linux distro that is back actually and Ken is going to tell. Ken is bringing the memes today. We're going to talk about it after a quick break. Don't go anywhere. This is going to be interesting. We'll be right back. Hey everybody.
Bobby Borsel
Brief interruption. I want to talk about our sponsor for this episode of entitled Linux Show. It's Bitwarden and we love open source. One of the reasons we love Bitwarden, the trusted leader in passwords, passkeys and secrets management. With more than 15 million users across 180 countries, over 80,000 businesses, it's really a success story. I love it. Bitwarden has built its reputation around trust, transparency, open source security and putting users first. And you know, I talked to Kyle Spiran, the founder and CTO at Bitwarden last month and for those of you wondering, I asked him, he said absolutely. Bitwarden is committed to its free version. The company continues investing in secure, accessible tools that help individuals, families and organizations protect their digital lives without compromising trust and transparency. And they believe that security should be accessible to everyone. That's why they continue to offer a trusted free. Free password manager that and we'll always do that. K said, we'll always do that. That's alongside the way they make their money. More advanced tools for those with families or teams to protect if they're a business. But I love it that they, they care so much about security that they want to make sure everybody has access to a powerful free password manager. And Bitwarden gives you everything you need to stay secure online. I use it. You must know this, generating strong passwords. Yes, of course. Unlimited pass keys, they're great at that. Managing sensitive credentials. I store my SSH keys in there. It will actually generate SSH keys, store them and deliver them. I use it for all my API, tokens all my secrets in AI too. And because Bitwarden syncs securely across devices, I have access to all of that. But only I. Even the AI doesn't.
Rob Campbell
Right.
Bobby Borsel
For businesses and advanced users, Bitwarden also delivers enterprise grade security tools that secrets manager, the vault health reports, secure credential sharing for teams. You don't want your team writing those passwords on post. Its trust me and Bit Warden has just introduced. They did this I think at RSEC earlier this year. I was there for it. Their new agent Access stk. This is a open source developer toolkit. They're offering it to everyone. They want every password manager to adopt it. I did. I'm using it right now. It's designed to help teams securely integrate credential access into applications, automation, workflow and AI agent environments. My AI agents never have access to my secrets. They do it through Bitwarden. So the secrets are stored, they're encrypted, they're trusted. I love it. The SDK enables controlled human approved just in time access to credentials stored in the Bit Warden vault without exposing sensitive information, without granting persistent access. Right. It's just in time. It's designed to support modern development and automation workflows while keeping security and transparency front and center. Never commit that token, that API key that password to GitHub again, never.
Ken
You shouldn't.
Bobby Borsel
And have you ever? I have. That's why I use Bitwarden. And because Bitwarden's open source, its code base is continuously reviewed and audited by both the community and independent third party experts. Bitwarden also complies with major security and privacy standards. Soc 2 type 2, GDPR, HIPAA, CCPA. It's ISO 27001 certified. And bit warden is always evolving to meet modern security needs. That's why they expanded passkey support. They've added secure developer tooling. They have very good flexible self hosting options for users who want additional control over their environments. No, they are happy if you self host. That's okay. In fact it's a great story. The guy who did filevault, which was a Rust. Was it. Is it called File Vault? It was a Rust based Bitwarden self hosted vault. They loved it so much they hired him and they support continued development of that. That's awesome. I think that shows you their commitment to open source. Get started today with a free trial of Bitwarden teams or enterprise plans. Or get started for free. For free forever. All devices, unlimited passwords, passkeys, all of the great features. As an individual user@bitwarden.com Twitter that's bitwarden.com TWIT I know I go overboard with this, but I am such a fan. Bitwarden.com TWIT now back to Untitled Linux
Ken
show and I'm tempted to start off by saying this is a story written by Bobby Borsel, Saurav Rudra and Jack Wallen about Noah Cagle. Yes, that was another meme. Now Noah Cagle is a developer and YouTuber that is reviving the project that was originally built upon Kubuntu 9.04 and that's dating this a bit. And KDE 4.2, another indicator of how old, how far back we're going. And then Noah drenched it in hot pink Disney Channel branding. Yes, Hannah Montana Linux is reborn. Now according to Saurav, Noah built this variant on Debian 13 with non free repositories enabled from the beginning for easy access to proprietary hardware drivers. According to Bobby, the rigly says most of it is a KDE plasma reskin with many parts being modified versions of Breeze. So while it looks like a dedicated Hannah Montana themed operating system, it is actually a Debian Live image with a customized plasma desktop. Now according to Jack Kegel's decision to use KDE plasma makes it a viable Debian based desktop operating system that can do all the modern things you need to do. Now I see this as an example of how with the proper incentive, anyone can customize Linux to their heart's desire. I also agree with Bobby, its main appeal is the absurd nostalgia. I do recommend reading Bobby Saurabh and Jack's articles for their takes and to find out what group Bobby places this distro in.
Rob Campbell
Ken, you gotta dig a little deeper here. On the surface it looks like a Hannah Montana themed distro, but if you really dig deep down underneath, it's really just Miley Cyrus.
Jonathan
Came in like a wrecking ball.
Ken
Actually if you remember the original was based on an Ubuntu so we're moving not down but upstream.
Jonathan
So.
Ken
So you're going to do that? You might as well go with the the Father of Miley Cyrus. I'm trying to Billy Ray Cyrus.
Jonathan
There was some, I don't, I don't know, there was some time in history where and I think it was the Hannah Montana Linux distro was actually notable for more than just being the meme because I'm pretty sure I remember Pharonics articles from the time and you know, Michael, Michael Laravel, you know, you kind of could read between the lines and he's like I'm going to go benchmark this dumb meme Linux because it was. Was it one of the few that actually ran Wayland? Was it like the first Linux distro that shipped with Wayland?
Ken
Something.
Jonathan
It was something weird like that. It's possible.
Ken
I don't know, the articles didn't say anything about how the original came to be, but it's quite possible.
Jonathan
Oh, was it Rebecca Black Linux? Is that the one I'm thinking of?
Rob Campbell
Oh
Jonathan
yes, yes it was. The Rebecca Black OS is the one I'm thinking of was like the first one. Yeah. Here's an article from Michael back in 2012 where he's like, well we're going to try it.
Rob Campbell
Rebecca Black os, you know, if. Yeah, I think hanging in my tana Linux has always just been a meme distro but I think if they really wanted to have some fun here, they should have made a Miley Cyrus distro and then create a Hannah Montana based off of Miley Cyrus. Oh,
Ken
I guess I'm waiting to see if somebody after reading the articles about what and how Noah Keiko did this or watching the YouTube video he put up about doing this, now I'm waiting to see somebody basically following his steps.
Rob Campbell
I'm curious how many people actually use
Ken
it with more science fiction type thing.
Jonathan
I mean it's, I mean sure, you can do that. I, I doubt, I doubt very many people are actually going to run this like on their main machine. This seems like something that you would throw on a laptop just to have
Ken
to show people or just on a vm just so
Rob Campbell
laptop to show people.
Jonathan
Yeah, yeah, the VM makes the most sense but like there's some, there would be some cred you would get from actually installing it on bare metal somewhere. Yeah, Rebecca Black OS was totally. Was the one that I was thinking of. So there you go. That one was actually pretty useful. It was literally one of the first ones that, that shipped with Wayland anywhere. And it let people get in and start playing with it. Yep. All right, let's move on. And this one is another story about a Linux distro, but a little less silly, a little less of a meme and a little less fun actually. And so we're talking now about OpenMandriva and you know, it's not a distro that I run that I've ever run anywhere but you know, it's out there. It does interesting things. Well, there was a bit of a dust up in the community and this is one of those, you know, drama Filled almost soap opera esque stories. We don't have to go into all the details, but it's interesting to look at here what happened. Right, so Angry Penguin, which is kind of the guy that is the lead, I think of openman Driva, he has a blog post. We actually got a link to the Pharonics article and then you can follow the link inside of that to get to the blog post. He talks about the history that OpenMandriva has with David Patrici, who is known for doing Mumble, which I still run mumble in a bunch of places. Apparently David joined the distro at some point in the past and he started hosting some of the infrastructure. He had a mirror from GitHub to a one dev instance, which that's a privately hosted sort of GitHub replacement thing. And there was, there was another contributor that came along into openman Driva and they don't go into any of the details here. I don't, I don't know for sure like who was right or wrong in this particular situation, but someone that was a part of openman Driva behaved in a way that was considered to be problematic and he got himself kicked out of one of the Matrix chats. Essentially you're causing more trouble than you're worth. Get out of here and stop bothering people. Apparently this guy was friend David's and David left the project. Angry Penguin says if he's no longer around, then I'm going to go turn off this mirroring because he's not part of the project anymore. And then what apparently happened is that David, still having some administrative privileges into the GitHub account, went in and again, this is just based on the blog. I don't know if there's, you know, more detail. Like I could imagine that something like this happens accidentally. I can imagine a scenario where this is an accident. Right. So I don't want to completely throw anybody under the bus here, but what's said is that David went in, abused his administrative privileges and sabotaged the distro, essentially doing forced deletes on some of the packages and repositories. So it, like, it's not cool, it's not good. You never like to see something like this happen. You know, whether you use the software or not, whether you're an openman Driva fan or not. And you know, this should go without saying, but no matter how ticked you get at a project, how much you disagree with them, don't, don't go vandalize the project for one thing, that's a crime and I don't know if they're going to pursue it, but like, there could be legal ramifications for doing this. That is a computer crime. Don't do that. Just, you know, be, be a, be a gentleman, be a, you know, take the honorable approach right off into the sunset. If you have a disagreement with someone, don't, don't do this. And you know, hopefully we'll. Hopefully, hopefully like so. And I know this is not super likely to happen, but hopefully there'll be some news report that, oh, it was totally an accident. He was actually just trying to clean up stuff and didn't realize he was. Not very likely. But, you know, I can, I can, I can naively hope. Let me, let me, let me have my naive hope.
Rob Campbell
Love your optimism there.
Jonathan
Yeah, I'm a hopeless optimist. In fact, the guys I do business with, I've told them this before. I'm like, you guys have to make me meaner. I'm too nice to people to be in business. I'm the hopeless optimist. And sometimes that's not great for the guy at the top. But anyway, I, yeah, I hated to see this. It's never, never, never cool to see this kind of thing happen between open source people and projects.
Rob Campbell
Surprised Mandriva is still around. That's an old project at least. I mean based off of the original mandrake from 1998 and they merged one year and became Mandriva and still around.
Jonathan
Yeah, I don't know. I don't know what their update. So they, they are now on 6.0. I was trying to figure out when that shipped. 20 24, 2025. I'm trying to figure out win 6.0, 2025, April 2025 open Mandriva LX 6.0 rock so it's, it's, you know, there's still some, there's activity there. It's not, it's not dead for sure.
Rob Campbell
I briefly use Mandrake.
Jonathan
I'm trying to remember what the. What Mandrake Linux, like what it was known for.
Rob Campbell
I know it was a Red Hat based.
Ken
I remember, I want to say Mandrake 6 using that. I think that's the one I started with.
Jonathan
So Mandrake from 98, it was one of the first distros to have a graphical installer. You didn't have to dive into the command line to do anything. It was intended to be a Linux distro for newbies.
Rob Campbell
It was popular for a little while there.
Jonathan
Yeah. OpenMandriva one of the things that it is known for according to the Google AI. So your mileage may vary. It tends to build things with Clang. With Clang instead of gcc. And so they're, they're trying to do some interesting compiler stuff which like that's, that's a useful thing to do. You get, you get better code and better projects and you know, GCC becomes better and Clang becomes better. When you have something like actively testing and trying to make compilation work, you find bugs by running things through both compilers. So like that in and of itself is something that's useful. So yeah, interesting. All right, let's, let's move along. Actually, it is time already for another quick break and then Rob's going to talk about AI in the kernel. I'm sure that won't be dramatic or divisive at all. Anyway, quick break and we'll be right back.
Rob Campbell
So Linux kernel developers once again are debating how they should handle code created with the help of artificial intelligence. This time they may eliminate a requirement to identify AI assistance altogether. The current policy says that when an AI tool contributes to a kernel patch, the developer should add an A quote assisted by tag identifying the agent, the model version and any additional analysis tools that were used. But Christian Bruner recently questioned whether the Linux kernels permanent git history should effectively become a free advertising spot for proprietary AI companies. He proposed simplifying the tag to something generic, such as assisted by LLM without naming the company or model. Jeff Layton went even further and proposed eliminating the requirement altogether. His his argument is that the rule is not followed consistently, the resulting information is unreliable, and nobody has clearly explained what useful purpose collecting it actually serves. We don't require developers to list every search engine, documentation, site, compiler, static analyzer, stack overflow. Answer that. That helped them create a patch. AI should not receive its own scarlet letter simply because some some members of the open source community dislike that technology. What what matters is the quality of the code. The human submitting the patch must review it still and understand it, verify that it works, ensure that it compiles with the kernels or complies with the kernel's licensing requirements, and accept all responsibility for it. I cannot add the legally meaningful sign off tag. Sign off by tag. That responsibility remains with the developer. Submitting untested AI generated garbage is irresponsible, but so is submitting untested human generated garbage. The problem is not the tool. The problem is developers submitting code they don't understand. AI can help explain unfamiliar code, identify Bugs suggest fixes, improve documentation and find vulnerabilities that might otherwise remain hidden. Rejecting those benefits because of ideological opposition to AI would only make Linux development slower and potentially leave real security problems unfixed. Judge the patch, test the code, hold the human developer accountable, but stop treating the use of AI as though it is something that must be confessed in every commit. Which is exactly, you know, what I have been saying all along. It's almost like maybe they listen to the show or, or maybe I've just been spewing common sense and, and everyone else is finally catching up. I don't know. But this makes sense to me. What's the point?
Jonathan
Yeah, so I think there's a couple of interesting thoughts here. One is that this was something I had never thought of. I hadn't really thought a whole lot about it, but the result of AI is not copyrightable. What AI writes, you cannot copyright. And I had a conversation with a guy on a flight here recently who was coming back from like a, a Comic Con style event in Oklahoma. I happened to fly out at the same time he was, and he, you know, he wrote for one of the, actually one of the larger comic book companies. And I'm like, hey, this is something that's impacting my industry. I'm curious what you guys are doing about it. What do you use AI anywhere in your, in your tool path? He's like, no, my boss set us all down until we're not allowed to. I'm like, oh, and why was that? Because we can't copyright it. And copyright is a really big deal for these content creators because they make comic books and they don't want people to be able to go copy them. So there and all that to say that there is the potential for some copyright and licensing issues to sort of hinge on whether a particular pull request was written by a machine and therefore sort of in the public domain by default or by a real person and therefore copyright and liable or subject to the GPL and all those things. And so that's, that's one thing that comes to mind. Another is that obviously they were looking for a way to kind of flag these potentially low quality contributions. But I think it's worth pointing out that like, and this is true in a lot of different areas, but like the, the people that would be willing to tag their contributions are the ones that you probably don't have to worry about. That's, that's an interesting observation you can make in a lot of different areas in life. And then the one other thing that comes to mind. This is sort of theoretical at this point, but like you could imagine that at some point in the future you would go like, there would just sort of become an awareness that, oh, that one version of Claude or you know, ChatGPT 4.3, it tended to make this mistake. And so then it might be useful to go back and grep and sort of review all of the patches that let's say ChatGPT 4.3 was involved in to see if it made that mistake in any of them. So like, I can see there's, there's probably a few instances where it could be useful, but on the whole it's not that useful a metadata. So I, I, I could go either way, I'll put it that way.
Rob Campbell
Yeah, it's an interesting thought about the copyright stuff and yeah, I kind of, kind of knew that already. I'm, I'm reading some things about it right now. 1, 1, 1 little snippet I see here is like AI generated code as a standalone work, you can't copyright. However, if the AI generated code is incorporated into a larger project, it may be protected under copyright laws.
Ken
Sure.
Rob Campbell
So there's probably some nuance and a lot of weirdness here that I think is going to need to be refined in the legal system in the future because.
Jonathan
Well, Ken, Ken mentioned during the break I had the conversation with, with Tris on Floss weekly a week or two ago and this, he is a lawyer in, in the UK and he does techno. Like his, the law firm that he is with, they specifically do technology and he in particular does technology cases. So he's, he's fairly familiar with the way all of these things work. And we talked about this with him and his, you know, his basic comment was the laws that we're using to try to decide this stuff was all written like in the 1980s, before even the Internet was really a common thing. And so of course we're having problems trying to reconcile it. But you know, that's one of those things that like will you try to update the laws? And you're just immediately, immediately you have this huge political partisan fight on your hand, which, you know, that's no fun. Nobody wants that. And that just sort of makes it difficult to get, to get the common sense things done that need to be done. So anyway, yeah, people are teasing us in the comments that this week in patent law, well, I mean the open source is all about Open source literally is all about licensing. Open source is literally a way to license copyrightable material.
Ken
So open source is A copyright licensed open source.
Jonathan
Depending upon what you mean by free.
Ken
Yeah, or freely.
Jonathan
Indeed. Indeed. All right, let's see what is up next. Ken FFmpeg there's some interesting stuff going on in FFmpeg actually. What are they up to this time?
Ken
Well, this is an FFMPEG because this week Bobby Borisoff wrote about an open source desktop application providing a Graphical interface for FFmpeg called Frame. Frame wraps FFMPEG in a native app and offers users a cleaner way to configure common media conversion tasks. The project describes itself as a native media conversion utility built in Rus using FFmpeg and FF probe underneath for media handling. FFMPEG remains the engine doing the heavy lifting, while Frame is the interface and workflow layer on top. Handling source, probing, FFMPEG argument generation, compatibility checks, task control and progress parsing. You can convert video files, extracting or converting audio, changing image formats, working with subtitles, editing metadata, resizing or cropping media, using presets, and processing multiple files in a queue. Now, according to the project, Frame supports common video formats such as MP4, MOV, MKV, AVI, WebM, and GIF. I had to say that. Now I installed Frame as a flatpak and started it up. Now the initial screen prompts you to add media to start a conversion queue. Excuse me about that. After that you can drag and drop files, click on Open file or Open folder. However, I found that it doesn't recognize or support the video transport stream or TS files that I have saved via PLEX when I record broadcast from a HDO run. Now it would also benefit from documentation explaining what the features it provides are, though I found it has controls for CRF and and bitrate settings, encoding presets, scaling, cropping, rotation, flipping overlays, subtitle burn in metadata, cleanup or replacement, and real time logs. Now Bobby's article provides links to Frame's official website and GitHub page. And of course you'll find the links to Bobby's article in our show notes.
Jonathan
So I was going to suggest that it must be Vibe coded because when I look at their website, I was teasing Rob before the show started about the look that an AI agent gives your website. And when you click on the Frame website it's like, oh this was. This website was written using like Claude or ChatGPT one of them. It just, it's just to my eyes it has that look. I could be wrong, but it's what, this is what I see. And so, oh, I wonder if the whole thing is Vibe coded. Like no, no Hate if that's the case. It could still be a cool project. And then I went and I looked, I'm looking at individual commits and I'm like, nope, this was written by a human. There's not nearly enough comments for an AI to have written it.
Rob Campbell
Well, unless they specifically told it. Be light on the comments. We don't want this to look like an AI.
Jonathan
Yeah, I don't know. No, it looks like it's probably just one guy that's just gone nuts on it. But, you know, who's to say?
Rob Campbell
Make some mistakes in your comments.
Jonathan
Make some mistakes to let people know
Rob Campbell
just once in a while. Not a lot, only like 0.00% or 1%.
Jonathan
Yeah, that's hilarious. No, it looks interesting. Now my question is, how is this different from Handbrake? Why would I. Why would I reach for frame over handbrake?
Ken
With Frame, you could actually go in and edit out worthy starting thing. I've never really used Handbrake for doing that.
Jonathan
That's true. I don't think you can do actual video editing with Handbrake. Okay, that's fair. It sounds like it has its niche that it's really good at. Then next time, next time I go to rip a DVD for archival purposes, of course. I'll maybe take it for a test drive. Do you know can Frame do Blu Ray rips? That would be interesting. Again, for totally legal, totally legal archival purposes.
Ken
Yeah. Moving it from that DVD to that SSD is definitely going to make that video easier to access and last longer than that DVD will, potentially.
Rob Campbell
Yeah.
Jonathan
It also makes it easier to play on a tv, potentially.
Rob Campbell
Your hard drive could die right after you put it on there and guess not.
Jonathan
Surely you've got some RAID 6 backup somewhere where you put all your stuff, right?
Rob Campbell
Maybe you got a backup of that
Ken
hard drive at least.
Rob Campbell
Maybe it hasn't backed up yet.
Jonathan
Yeah. All right, well, let's talk about something completely different. And that is one of those laptop brands that you may or may not have heard of. They actually build a lot of laptops. It's Razer R, A Z E R, maybe better known for their keyboards. But anyway, Razer has a very impressive laptop. It's the Razer Blade 18, the RZ090582. They apparently talked with the USB guys before they named this. And the notable thing about it though is that it is certified for Linux. Yes, Razer is officially getting into the Linux game. They've got a laptop that is certified for Ubuntu Linux and it's got the Intel Core Ultra 9290HX plus platform. Everybody is terrible at naming things and a GeForce RTX 5090 graphics card in it. And apparently based on Michael Larable's benchmarking, this thing is really impressive. He compared it with quite a few different devices, including a framework real similar to the one that I run. And the Razer really, really took the top spot by quite a bit by multiple percentage points. Like an obvious winner here on just about every test, I think. So they did something right both with the processor with the thermals on it. It looks like they won basically every test. Now there's a couple of interesting quirks about it. So one thing that you know Razer for of course is like the RGB lighting in their keyboards. And you may know that getting that supported in Linux, there's a couple of different ways to do it, but all of it is a little bit spotty. There's just, there's quirks with it. And this, it doesn't work. They have it, they have the Razer Blade laptop certified for Linux, but when you're using Open Razer and or Polychromatic from the Ubuntu ppa, so there are some asterisks on this, but still when you're using those two things from the Ubuntu ppa, this laptop is not supported yet. So like you can't control the RGB lighting on it yet, even though it is fully certified by Ubuntu apparently that's not one of the things. That's not one of the boxes that have to be checked to get the certification. That's sort of icing on the cake though. And really what you care about is that the graphics card works, the keyboard works, the sound works, that your processor runs correctly, that it's not going to cook itself. And it seems that all of that stuff is actually working the way that it's supposed to. Now this thing is not cheap. You're going to be talking about like $5,000 for the 32 gig version. If you max it out with 128 gigs of RA, it is a whopping $6,999. So like, that's, that's more than I'm willing to pay. I am, I, I am, I am finally doing well and I still, I don't want to pay that much for a laptop. So any but super cool, super cool to see that Razer is finally jumping on the Linux laptop bandwagon. They've got an Ubuntu laptop now and it's out there and it's available. And apparently it's a real screamer when it comes to performance.
Rob Campbell
Yeah.
Ken
And about 4,000 of that, 6,999 is just for the memory.
Jonathan
Yeah. So when you go from 32 gigs of RAM up to 128 gigs of RAM, it's literally a $1600 price jump.
Rob Campbell
My experience with Razer is with the Razer mouse. That's what I have right now. I like it, except the thing burns through batteries worse than any mouse I've ever had.
Jonathan
Surprise. You gotta pay for performance, man.
Rob Campbell
Yeah, I like my old mouse. I think I changed the regular battery once every year or two. This. A regular battery won't last me a month. So that's why I have these rechargeables now next to my desk.
Jonathan
Nice.
Rob Campbell
I just keep swapping them out. It's more frequent than a. They don't last as long as a regular battery. But who charges when it goes. Yeah, it goes through so many batteries. I don't go.
Jonathan
I'll tell you a secret. I hate wireless mice and keyboard. I will absolutely put up with a cable on them to not have to mess with batteries. Of course, I also use trackballs everywhere, so, you know, who cares whether that's what. In fact, one of the weirdest things in the world is a wireless trackball. It's like, why.
Rob Campbell
How.
Jonathan
Why would you even want to. It's such a terrible idea. But, yeah, wired mice. Wired keyboard.
Rob Campbell
For me, my. I mean, my last mouse, the battery probably lasted two years, so it wasn't a problem. Now this one, when. When I'm using a rechargeable, I'm playing a game, and it just stops, like, right in the middle sometimes. I've never, like, been right in the middle of, like, some crazy battle. It always seems to pick an opportune time. Fortunately, that's handy.
Jonathan
One thing I do like about Razer is at least maybe not all their mice, but some of their mice, the ones that I've seen, they actually have a USB port on them. And so, like, if you want to run them wired, you can plug them in. Yeah, look at the. Look at the front of the mouse where it would be.
Ken
Do you notice any difference in latency between plugged in or wireless?
Rob Campbell
I've been using wireless for years.
Jonathan
Yeah, it probably uses. It probably does not really even packetize the data, so it probably is basically zero latency.
Rob Campbell
And supposedly this razor is supposed to have way better latency than, like, my old logitech I had.
Ken
But I'M like you, Jonathan. I prefer the ported. In other words, I like tails on my mice.
Jonathan
Yes. I don't like that we like tails on our mice.
Rob Campbell
I don't like that in the way.
Jonathan
But yeah, I don't know, it's just, it's more important to me that it always works than anything else. I can put up with a whole lot of stuff in the way as long as it works all right.
Rob Campbell
And the battery is really easy to change too. Like the battery cover is magnetic, so I just grab it. Oh, it slaps right in real quick, right on top. So at least they knew you're going to be changing the battery a lot.
Ken
I don't know about you, Rob, but I'm ready for another break.
Jonathan
Do you. Do you shout out reloading every time you go to change your batteries out now?
Rob Campbell
I try to pretend nothing happened and keep on going because it's a little embarrassing that, you know, I'm playing a game with a bunch of people and so, and people say I, I don't use a wireless. I don't, I don't want that to happen. I just.
Ken
You mean you don't swap the batteries out before you start the game?
Rob Campbell
Sometimes, but usually no, I don't take a moment.
Jonathan
All right.
Ken
And that's when they'll die.
Jonathan
That's when it dies. Of course, I don't imagine that Rob plays like real competitive, like ranked games though. Somehow that doesn't seem to be his. His cup of tea.
Rob Campbell
I'll do some shooters though.
Jonathan
Do some shooters.
Rob Campbell
And I'm. I'm usually the one getting shot, even with the better mouse. But hey, yeah.
Jonathan
All right, let's take a quick break and when we get back we're going to talk about Fedora and AI. So don't go anywhere. We'll be right back.
Rob Campbell
Not long ago, we talked about two major Linux distributions. Preparing for an AI focused feature. Ubuntu announced plans to make its operating system more useful for AI development while also exploring local inference, accessibility features, agentic workflows, and eventually a more context aware desktop. Then right afterwards, Fedora proposed, I think it was a week later, proposed a different approach, an AI developer desktop. Rather than integrating AI features throughout standard Fedora experience, Fedora planned to create an optional atomic desktop image designed specifically for developers working with artificial intelligence and machine learning. It would provide easier access to accelerated workloads, AI development tools, optimized containers, and better support across hardware from amd, Intel, ARM and Nvidia. At the time, I wasn't sure which approach I preferred but the Fedora want to sound pretty good to me. Ubuntu appeared to have the more ambitious version, but I like Fedora's idea of providing a dedicated and completely optional AI environment. Developers who wanted it could install it, while everyone else could continue using regular Fedora as, as normal and just not even think about the AI stuff. I was excited to see both distributions experimenting and I was waiting to see how their ideas developed before deciding which one I might use. Unfortunately, Fedora's proposal became tangled in arguments over Nvidia, proprietary CUDA components, the ethics and energy consumption of artificial intelligence, and whether AI development belongs within Fedora's open source mission. Now the Fedora Council has closed the official discussion. Boom. Nothing else. Not talking about no more. It's also concluded that Fedora's broader community initiative process was not working effectively and paused that entire process while it develops a replacement. The Council says supporters can continue developing the AI desktop independently without us. You know, if it eventually matures, it could return as a Fedora remix or follow the process of becoming an official Fedora offering. But not today. So, but, but the idea, it's not necessarily dead forever. It's just no longer an officially supported Fedora initiative right now, and the planned Fedora AI developer desktop is just not proceeding. Meanwhile, Ubatu is still moving ahead. Canonical says plans to spend 2026 improving local inference, integrating optional agentic workflows, developing AI powered accessibility features, and preparing Ubuntu for increasingly capable consumer AI hardware. For me, it's a disappointing situation because I genuinely liked Fedora's approach. A whole system just designed just for this, you know, this was an opportunity to provide developers with a ready to use AI workstation, you know, without having to force AI on anyone else who didn't want to use it. So, you know, in the world of AI desktops, where there was once two and now there's one.
Jonathan
Yeah. Personally I think this is more about the way Fedora works and how they get on the same page and how they like, steer the way the desktop is going to go. I think it's more about that than it was about the, the AI question. And I think you will, you will definitely see the AI, the AI desktop come back in some form or another.
Rob Campbell
Somebody still pushes it forward. The idea, yeah, it doesn't have to be an official one if it's still a thing, but.
Jonathan
Right, and so like there's nothing that would prevent them from making a spin, you know, adding, adding packages and then pushing a spin that use those packages by default. There's just, there's just now not a sort of top down agreement that hey, this is an important thing that we want to do in Fedora because the Fedora guys sort of understood now that their method for coming up with one of those was sort of fundamentally broken. Now I don't know all the details as to what all happened to bring them to that decision. I'm sure that when someone said let's make an AI desktop for Fedora, you had a certain group of users that just went nuts and keep your AI out of my Fedora. Which I suppose that's understandable, but at the same time nobody was going to force them to run it.
Rob Campbell
Some version like you said, likely to go forward, but it would still be nice to have the official backing of, I mean basically IBM Red Hat. I mean IBM is known for AI and I'd imagine a lot of, you know, since Fedora is an upstream to them, I'd imagine, you know, Enterprises would like to get their hands on some, some version tool of some sorts like this. But yeah, maybe it's just not yet.
Jonathan
Yeah.
Ken
And I think part of the problem is, at least from Bobby Barsoff's viewpoint in the article I posted in the Discord, is that this actually exposed a deeper issue with the community initiative itself.
Rob Campbell
Yeah, they said they shut that down. Mention that.
Jonathan
Yeah, yeah, they, they killed that as a. So it's, it's no longer there. And so if you think about this like over in Ubuntu land, they've got much more of a top down structure.
Rob Campbell
Right.
Jonathan
So like what's his name?
Rob Campbell
Mark Shuttleworth.
Jonathan
Shuttleworth, yes. Shuttleworth says, hey, this is what we need to do. And his engineers go, we will make it happen, sir. And sometimes they're terrible ideas like mer, but sometimes they are pretty useful ideas. And some of the things that they demoed at the launch Unity, like Snaps. I mean Snaps is honestly pretty interesting. What they were showing off with doing AI instead of Snaps. It was fairly compelling. So you know, it's, it's a, it's a give and take.
Rob Campbell
Right.
Jonathan
There's two different ways to do business. Neither are necessarily wrong, just different ways to go about it. All right, let's talk Pipewire then. Ken, what is new in 168? Anything we all need to rush out and update for?
Ken
I wouldn't rush out and update. I just wait for you to distro to come out with the updates when it does. But, but you can thank Bobby Barself and Marcus Nestor because they also wrote about my Favorite Linux based open source software for handling multimedia pipelines via a server and user base API. And yes, as Jonathan said, we're talking about PipeWire version 1.6.8. Now highlighted in the release notes of this version is fixing a data race in jack that could cause lost MIDI events in Ardor as well as some unbounded memory allocations. Now, according to Bobby, this update avoids unnecessary graph recalculations by fixing a bug that could occur when suspending an active node. The modules area also gets a fix for a potential memory leak in the error path of client node. Now, According to Marcus, PipeWire 1.6.8 addresses a crash in GStreamer support caused by the metadata listener being registered twice. PipeWire 1.6.8 also fixes a potential leak when Bluetooth transport fails to start and fixes issues with filter graph volumes that were observed when the filter is loaded inside a node with hardware volume. As always, you can read Bobby and Marcus's articles for more details and of course I have links to them in our show notes.
Jonathan
Yup. Lots of. Lots of solid fixes, it sounds like.
Rob Campbell
Yep.
Ken
I don't like that hearing about memory issues.
Rob Campbell
Memory.
Jonathan
I mean, pipewire talks directly to hardware in some cases. And so like you. Well, yes, but also like you, you would say, oh, let's rewrite it in Rust, we'll fix all this and we'll know. To talk directly to hardware, you've got to do Rust unsafe. And so, you know, there's no getting away from the possibility of memory errors. Yeah, you have to talk, you have to talk to the raw memory to talk to the hardware.
Ken
So especially if the memory error is caused by unplugging that device and then plugging it back in,
Jonathan
that could cause problems. All right. Oh yes, we've got a. This is kind of different.
Ken
This is.
Jonathan
For me, this is the story that I found. This is. This is not exactly a Linux story, but man, it really, it really hit me as being super duper interesting. And there's one, there's one moment in particular that I just had to. I just sort of of stopped and marveled and I want to bring that moment to you. So this is. This is a hackaday article about a. And it's, you know, it is vibe coded, right? Like with this right from the top. This was written by an LLM. Okay. So Fabian, Fabian Kubler said, hey, and he got, he got access to Fable, right? So the new, the new sort of alternative to Claude from. What's the name of the company that makes Claude? I don't even know. Anyway, the company that makes Claude, they now have Fable. It's their new Anthropic. Thank you, guys. I too, have memory issues. So Anthropic's new Fable, of course. Did we talk about this? They released it and then the U.S. government said, oh, no, no, no, no, no, this can be backdoor. And they're like, okay, fine, nobody can use it now. It's finally available again. Anyway, so he said, I want to know how good this new Fable thing is. And so he came up with this reasonably crazy project. And basically he said, hey, Fable, let's write some Super Nintendo code. Let's do some coding on the Super Nintendo. And he got sort of a hello World going. And then he built a. Trying to remember what was next. He has a compiler, he has a hello world, and an emulator harness as well.
Bobby Borsel
That's it.
Jonathan
To be able to run all of these things in emulators and speed up the process so you don't have to wait on the real Super Nintendo hardware. But then he said, hey, build me micropython. He said, do a micropython port to the Super Nintendo. And what's really interesting is that he got partway through it and then Fable disappeared. And so he had to go back to the older Claude, like Opus Model and Opus 4.8, and it just could not handle it. The older model just couldn't do it at all. And then he got back to Fable 5 again and started making progress. So, you know, it's interesting to see the differences in these models and how. How much more capable some of them are at certain tasks. But he said in just a few hours he was able to solve those problems. I'm trying to remember how long this took in total of Fable time within several days, though, from what I can tell. So anyway, you found several real bugs in Micropython itself and a handful of bugs in the compiler that he was using, one of which being the negative Y pointer bug. And then he had another bug he found that was an alignment problem. And you come to find out that if you dive deep enough in the code, you find out that when you use the aligned tag, it's a no op. It just doesn't do anything. It doesn't do what it said it does. So a couple of these different things. So finally he got it done, Fable got it done for him, and he says it passes 91.9% of MicroPython's core test basics suite. So 468 tests that they'll run on Micropython, his port to the Super Nintendo passes 430 of them, which is pretty good. So he's got a game, it's called Stage. Apparently this is an existing game. It's made for these little DIY handhelds. It's a known Python game. And he's got it running on the Super Nintendo. And if you click all the way through, and this is the moment that I decided I really needed to bring this here because this is what got me. If you click on through, you go Fabian Cooler's website. The actual post about it, you scroll down, it looks like a picture here where it says press start, right? And if you click on actually starts doing things, and then it goes to a blue screen, it says Micropython on SNES booting takes a moment. And I went looking at this and I'm like, oh, it's a video he captured off of Super Nintendo, right? And then, you know, I mouse over it and I start clicking around. No, no, no, no. This is not a video. This is an emulator. This is an emulator running, like, in either JavaScript or WebAssembly, emulating Super Nintendo, running Micropython, running this game now at like 1fps, right? But still, it just. It was this moment of, wait a second, wait a second. We're emulating a Super Nintendo in the browser, and then that Super Nintendo image is sort of kind of emulating Python, which we're two layers deep now. I don't know. I just, I. I find this to be very, very impressive, but also very fun. And it's also like, it's interesting. We were talking before the show started. Rob and I were just about the point that we've come to with LLMs writing code, and they've really gotten impressive, like, really, really good at it. And I was reminding Rob that I used to be something of a skeptic. Like, no, no, no, it's never going to be as good as a human. I don't know about that anymore. I'm not convinced that's still the case. I don't know, maybe we humans need to get. Get good at digging ditches again because,
Rob Campbell
I don't know, five years, we'll all be retired in one form or other, I guess.
Bobby Borsel
I don't know.
Jonathan
I don't know what that's going to look like. But anyway, that was. It was fun, though. I think it's. It's pretty cool that he made this work and that you can run it in your browser. That's just neat to me.
Ken
So Anyway, run it on Atari Jaguar.
Jonathan
Not yet, but I bet you it would be portable. Does the Jaguar use the same CPU
Ken
chip as the SNES M68000?
Jonathan
No, I don't think that is. The snes used the Ricoh 5A22, which was based on the 65C816. So, no, not the same thing.
Ken
Jaguar had Motorola 68,000 CPU, two megabytes of RAM, cartridge ROM at OX 8,000 or 80,000, and two custom chips called Tom. Guess what the other one was called Jerry. Yep. Tom handles graphics while Jerry takes care of DSP tasks.
Jonathan
Is the Super Nintendo based on the 6502? Yeah, it is. It is a 6502. So many things were based on the 6502 back in the day that it
Ken
was a great ship.
Jonathan
Yeah, yeah, that thing was super interesting
Ken
because the original Apple, Atari, Commodore 64, I can't think of all the games consoles that were based on it.
Jonathan
Get a lot of Acorn had a couple. The Apple One, the Apple II, the Apple IIe, the Atari 2600, all the way up to the Atari 7800 and 800. And the Atari Lynx was based on it. BBC Micro, the Commodore PET, Commodore VIC 20 and 64 and 128. Of course, the Nintendo, the Super Nintendo was roughly based on the TurboGrafx 16. A lot of different things were 6502 based. That's crazy. So much. So much of the heritage that we enjoy today came from that particular chip. Yes, it's great.
Ken
And then we moved from that 6502 to the X86, one of the next big ones. Yeah.
Jonathan
All right, that's pretty much it for the news for this week. We got some tips to get into. We're gonna take a quick break, one final break for before we end the show, and then we'll get back into some command line tips. So don't go anywhere. We'll be right back. All right, Rob, what do you have for us? What is your command line tip for the day?
Rob Campbell
Now, I need to remind you, this is the tips section, not command line tips. So they're not always command line for this week. Step a company you've already talked about today. Anthropic, has finally released an official beta of Claude Desktop for Linux. So maybe something you want to try out. Still in beta, but personally, I am not much of a Claude user, mainly because it costs, but. So I haven't tried it out and I don't have anything to show, but I wanted to share this with with those that may be interested or like AI or Claude fans. So I'm just going to talk about this one unlike almost all the time when I have something to display for the viewers. So this tool. This brings the same basic experience already available on Windows and Mac os, including regular Claude chats, Claude code and cowork. It can run multiple sessions in parallel, visually review code changes, and provide an integrated terminal editor and live application preview. There are still a few limitations. Computer use, which would allow Claude to interact directly with the desktop applications, is not yet available on Linux, although Anthropic says it is coming. Voice dictation is also missing, and some desktop features such as notifications and quick or quick entry may not work reliably during the beta. This is also a client for Anthropic cloud service, so not a locally running AI model. So you do need an Internet connection and Anthropic account. And for installation instructions, check out the Show Notes, system requirements and check the link in our Show Notes for all this. And you know, one thing I was thinking about when I was doing this, you know, talk about all this Vibe coding, I kind of wonder, I'm kind of curious if Anthropic eats their own dog food and Vibe coded this whole thing. Sure, I, you know, I kind of hope so. But then if so, I wonder what's taken them so long to add all the missing features. It should be something pretty easy to do in an afternoon from my experience. It's funny, I also know on my experience lots of times AI will push you to just do like one step, add one feature at a time instead of like 50 of them that I tell it to do. So I don't know, maybe. Maybe that's just kind of their model and they follow that is do one thing at a time. And then they test that. Even though when I do my own Vibe code bank, no, I want to just test this all at once, do these 50 things and then I can. Then I then I don't have to like spend.
Jonathan
It's quicker.
Rob Campbell
It's quicker.
Jonathan
Yeah, you get a little more control over it. Doing it one thing at a time, I find.
Rob Campbell
You know the, the one I mean, I suppose if you do the right prompts. But I know one thing I'm concerned with if I do it one thing at a time is that like five things in it'll change something from that first thing, even though I didn't tell it to anyway. So I was like, let's just do it all at once.
Jonathan
And then why this is why? You set your vibe code project up as a git repo and you commit things in between so you. You can review it now. Well, you shouldn't. Then you go back and you smack your LLM. So you weren't supposed to touch that. Go put it back.
Rob Campbell
Oh, I have. I. I talked to my LL like like it's a person. I'm like, you messed it all up. Why did you do that?
Jonathan
You should feel bad about yourself.
Rob Campbell
I try to make it feel bad about it.
Jonathan
Code is terrible and so are you,
Rob Campbell
but I know it doesn't.
Jonathan
It doesn't care. Unfortunately, the. The cloud Linux beta is Debian and Ubuntu only at this point. No, no RPMs for me to go
Rob Campbell
play with, so they are targeted. But you could maybe try to use a tool like Alien to make it work.
Jonathan
Yeah, there's ways to go about it anyway. Actually, I tell you where I use Claude, let's use it right inside of Vs code works fairly well. It works well enough for me for what I'm doing.
Rob Campbell
Do you tie the API right into it? Is it API, right?
Jonathan
Yeah. So you grab their extension and you log in and it'll open a browser window. You log in and then it gives you a token, a key to use, and then it talks. Works pretty well. All right, Ken, what about doing things locally on our own hard drives? What's some tooling that we can use there?
Ken
Well, this week I'm going to discuss a command for managing, monitoring, configuring and troubleshooting NVME storage drives. Directory from the Linux terminal. Now, let me go ahead and bring up my screen here so I can share with y'.
Rob Campbell
All.
Ken
Is that big enough for everybody to see?
Jonathan
Yeah, not too bad.
Ken
Okay, but the command itself, it can be either referred to as NVMe or nvme-cli, depending on the distribution you're running it on. But you'll find that what you actually type is NVME followed by a sub command. In this case, I put the sub command help so that can print usage information to display to everybody. Quite a bit of usage information there. I'm just going to touch on a couple of those. The first one, of course, is going to be list with list. What you can do is list all the NVME devices you have installed. And as you can see, for those of y' all listening, I just did NVME space list and it came back with the list of my NVME devices. I only have one, and it's showing the NV flash dev slash NVMe 0N1 as the device and then, then it goes on and gives the serial number for it and the model. It's a Kingston SNV2S 1000G. So you can guess what size it is from that.
Jonathan
Probably a terabyte.
Ken
Especially since that's what it says as you go on down.
Jonathan
I promise I didn't cheat.
Ken
And it's showing that the usage of mine is 1 TB out of 1 TB.
Jonathan
Is your hard drive.
Ken
No, that's just showing what you have there.
Jonathan
All of the, the. The. All the data is. Or all of the disk space is spoken for by partitions, right? Yeah.
Ken
Now another command that you sub command you can use is smart log and this is going to give you a lot of information again. Now this one you do need to run as root. So I've got sudo space nvme space smart dash log and this is going to sprint your smart the log for a device NVMe Zero and it's prompting me for my password. And yes, I've got to go in and set up a alias for pseudo so that it doesn't display how many characters I've typed.
Jonathan
Yeah, apparently smart log is an invalid argument.
Ken
Oh, I forgot to copy the. Well, that's because I didn't tell it what device to look at. Ah, I need to follow that with the/dev/nvme suo. There you go. And there we go. And that's the joy of doing it live. Now, this gives you a lot of information. It tells you the device that it's looking at its namespace id, which in this case right now is. I don't want to count the S. So if y' all are listening, I'm going to leave you in suspense.
Jonathan
A number of Fs.
Ken
I think I'd get us kicked off YouTube if I tried. But one thing that some interesting things that you want to look for is the temperature, the percentage use. And here's another one that you may find interesting. Unsafe shutdowns. That's where for whatever reason I had to just shut off the power to the computer.
Jonathan
Yeah. So this is, this is an NVMe, right?
Ken
Yep.
Jonathan
There's actually one number on there that is probably the most useful of any of the temperature is super useful. But yeah, right up there towards the top, the available spare. That's probably the most interesting number on there because that really is telling you how much life you have left to in that NVMe drive. So it's got, it's got A certain number of sectors that are considered spare and it will use those as sectors die. And so right now it's got 100% of those left. And when that gets down and so there the threshold is 10%. But when that gets down to zero percent, that's basically when your hard drive will, will start to die and, and lose data. No longer work the way you expect it to.
Rob Campbell
All those apps, isn't that the color white or is there too many?
Jonathan
I don't know, I'm not going to count them.
Ken
But it also indicates that this particular hard drive has three temperature sensors.
Jonathan
Yeah. So some chip in there is running at 60 degrees Celsius and some chip is running at 50 and the overall temperature is 38.
Ken
That's Celsius. And when you look at the. It does the conversion for you to Kelvin And Fahrenheit at 60 degrees Celsius is quite hot. Yeah. Another one that you'll Find useful is NVMe space ID CTRL. And of course you're going to use sudo with that again and the device is going to be the NVMe 0 again. And it gives you a lot of data. Yeah. So there's a lot of data that you can look in there, get a
Jonathan
serial number out of it. That's useful.
Ken
Now what's nice is you can also use sudo mvme git dash feature and let me go ahead and type the whole thing into the landline. But after get that feature, I have the device dev slash NVMe zero. Now I'm following that with a dash capital H or dash dash human, dash F is dash dash feature dash ID. And here I'm having that equal ox zero D quickly. Can you convert that to decimal 13? Yeah. So I'm looking to get feature ID number 13, which happens to be the home host memory buffer. And as you can see, I've got the enabled Host memory is enabled.
Jonathan
It gives you some information about it.
Ken
Yeah. Yep. To include that size, that's 16 gigabytes.
Jonathan
Ah, cool. All right. Very cool. Some good, good info for getting off your NVME and watching it for its health.
Ken
Yep. And you could even set up a script that would allow you to have that monitored every five minutes and warn
Jonathan
you send you an email when your hard drive is starting to die.
Ken
Email or pop up a notification depending on where that hard drive is connected.
Jonathan
Yeah.
Ken
While I was looking into this, this is also used with fiber. I'm trying to think of the phrase again. Fiber channels.
Jonathan
Yeah, that's a, that's a That's a fancy like big data data storage thing, like sans, like storage area networks and all that, these fiber channels.
Ken
But if you follow the link I've got in the show notes for Me and NVMe CLI, yeah, it'll take you to the GitHub page where they have a link to nvmexpress.org where it talks about it and gives you the specifications for NVME and drivers. That way you can look up some of the other features that you might want to pull off your NVMeS.
Jonathan
Nice.
Ken
Something to explore. Sure.
Jonathan
All right, well, there's something that I have not explored a whole lot on Linux. I know it exists, I've used it a little bit and that is local MDNs. What are these locals on the network? Well, MDNS is a link local dynamic name system that doesn't use hierarchies. It's just all a device can advertise that it is on the network with a certain name. And it's interesting and it gets used for a lot of things like printer auto discovery and various things. But you may say to yourself, I wish I could use that better. I wish I could do more things with that. Wish I could set DNS names with locals. And turns out that you can on your Linux desktop you can run the AVAHI publish address and you can just say, hey, this thing local is at this IP address and it will start advertising. Avahi is the name of the daemon that does it and you can control it and set up your own domain names that way without having to log into your router to set them. This is a Hackaday article from Al Williams and he goes into a lot of detail about how to make this work on your local machine. And it's, it's pretty cool. Like go set up your own dot local if you want to. And you know, if, if, if you have a, I don't know what, a Raspberry PI somewhere and you want it to be on like proper DNS. You want it to be, you know, Raspberry PI or PI or you know, Pi5, whatever its name officially is, but you want to give it a second name like Gizmo Local. Well, you can absolutely do it. And MDNS is one of the ways to, to do so. Where it's really useful, I think is you've got a device that has like, some printers are like this where it uses say its serial number for its domain name that it sends up through dhcp. Or maybe that ability to send the name up through DHCP is just Broken. I've seen some devices like that. You can use this to give it a human readable name on your network. So, yeah, I thought it was pretty cool. Al pitches it as sort of the replacement for slash EDC hosts on modern networks, and I thought that was pretty cool.
Ken
And you could probably set that up if you've got a local server for it to do that for you.
Rob Campbell
You could.
Jonathan
But the main advantage it, like, the really sort of killer feature about it is that you can run it from any machine on the network and it will. It'll advertise and everything else can see it.
Rob Campbell
Yeah, if you have a DNS server already, you don't really need that, but it's kind of. If you don't want to go through and set that kind of stuff up.
Jonathan
Because some people, you know, their, Their, their ISP may control their router and this so they can't get into it and make changes. All kinds of.
Ken
All kinds of, like AT&T controlling their fiber router.
Jonathan
I haven't. I wouldn't get a name. Any names just say ISPs. But, yeah, they're one of the ones that does that sort of thing. All right, well, that is it. We have reached the end of the show. I'm going to give the guys a chance, if they want to, to plug something. We're gonna let Rob go first. Rob's gonna unmute himself before he starts talking. Yeah, there you go. What do you got for us, Rob?
Rob Campbell
That was. All righty. So I just got the usual. For those who want more of me, Want to. Want to be my friend. Wanna see what I got going on outside of the show here, you can come find me at robert p.campbell.com. it's at the top right now. All throughout the show. It's always at the bottom. And on that, you'll get to my webpage. And on that page you can find links to my LinkedIn, my Twitter, my Blue sky mastodon, a place to donate coffee. And if you want to learn more about me, you can even scroll down on the page and find other things like, oh, you know, I got an about me section and magication, my work history and a resume. If, if you, if you, you know, got lots of money to throw at me and want to offer me a job or something. I don't know, whatever. Anyway, that's my page. That's where I'm at. Come connect. I'll see you on the socials.
Jonathan
There you go. And Ken.
Rob Campbell
Ah, perfect.
Jonathan
Now Ken is like.
Rob Campbell
I like it
Ken
now if, like Me you have been running an Ubuntu 2510 flavor. Then I recommend reading Marcus Nestor article reminding us that 2510 reached the end of its supported life last Thursday. That's July 9, 2026 for the. Yes, those of y' all who download this later and what he suggests that you do now.
Rob Campbell
Yeah, or just upgrade.
Jonathan
Cut right to the chase there, Rob. All right, thank you, guys.
Ken
Are you sure that's what he to do?
Rob Campbell
I don't know what he says, but that's my recommendation.
Jonathan
Unplug your computer and throw it in the river.
Rob Campbell
Yeah, go buy a new one. It's outdated.
Ken
He gives you step by step instructions on how to.
Jonathan
Oh, that's so boring though. It'd be much more exciting to throw
Rob Campbell
it in the river if you're, if you're running, if you're running an interim update like that, there's chances, good chances, unless you're new, that you've done it before.
Ken
True. All right.
Jonathan
Appreciate the guys being here. Appreciate everybody that's out there. If you want more, you can, of course, check out Hackaday. That's where Floss Weekly lives. We have a lot of fun there as well. Other than that, though, I just want to say thank you. Oh, I'll be at Open Sauce if you're in the Bay Area. I will be at Open Sauce this next weekend. You can come by the meshtastic booth and say hi. Should have some very fun stuff to play with there. Anyway, other than that, we just want to say thank you to everyone that watches that listens, those that get us live and on the download. And hey, we'll be back next week on the Untitled Linux Show.
Wayfair Ad Voice
Hey there, it's Wayfair here, where delivery and setup are as easy as a few taps on your phone. You're relaxing in an old hammock, scrolling Wayfair's app when you spot it. A brand new patio set. Next thing you know, Wayfair delivers it right to your patio and sets it up. Oh, you need a new grill too. All right, Wayfair's got you covered. With Wayfair's room of choice delivery and fast expert setup on qualifying orders, life gets a little easier. Visit Wayfair.com or the Wayfair app.
Ken
Wayfair. Every style, every home.
Rob Campbell
Hey, it's Ryan Reynolds here for Mint Mobile. Now, I was looking for fun ways to tell you that Mint's offer of unlimited Premium Wireless for $15 a month is back. So I thought it would be fun fun if we made $15 bills, but it turns out that's very illegal. So there goes my big idea for the commercial. Give it a try@mintmobile.com Switch upfront payment
Wayfair Ad Voice
of $45 for 3 months, $90 for 6 months or $180 for 12 month Plan required $15 per month equivalent to taxes and fees. Extra initial plan term only greater than 50 gigabytes. Me slow when network is busy. See terms.
TWiT.tv | Aired: July 12, 2026 | Host: Jonathan | Panelists: Rob Campbell, Ken, Bobby Borsel
This week’s episode rounds up the latest in the Linux world, focusing on significant desktop updates, fascinating community drama, new tools, and the evolving impact of AI on open source processes. The panel digs into real progress for Linux Mint and Cinnamon with Wayland, the return of the infamous Hannah Montana Linux, a controversy in the OpenMandriva community, ongoing debates around AI code attribution in the kernel, Fedora’s AI desktop initiative stalling out, new FFmpeg front ends—and a look at impressive Linux hardware. Light-hearted moments, thoughtful takes, and deep technical dives make this episode a must for open source and Linux enthusiasts.
alien package conversion).nvme or nvme-cli for monitoring/configuring NVMe drives.
nvme list—see available drivessudo nvme smart-log /dev/nvme0n1—view S.M.A.R.T. data, temperature, available spare, unsafe shutdowns.id-ctrl, get-feature) for deeper info, including host buffer size.avahi-publish-address to announce/customize local network hostnames (e.g., raspberrypi.local) even without router access.