Untitled Linux Show 190 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.

00:00 - Jonathan Bennett (Host)
Hey folks, this week we're talking about the Rust feud as it continues to play out inside the kernel, but there's better news, like NTSync and HDR finally landing. You can't use them yet. We dive into what the timer interrupt is in Linux. Talk about the updates to Handbrake, kde, plasma and more. You don't want to miss it, so stay tuned.

00:21 - Leo (Announcement)
Podcasts you love.

00:22 - Jonathan Bennett (Host)
From people you trust.

00:27 - Leo (Announcement)
This is.

00:28 - Jonathan Bennett (Host)
Twit. This is the Untitled Linux Show, episode 190, recorded Saturday, february the 15th, a Fedora-ish direction. Hey folks, it's Saturday and you know what that means. It's time for the Untitled Linux Show. We're going to get geeky with Linux and open source software all kinds of good stuff. It's not just me and today we've got as co-hosts Mr Rob Campbell and Mr Ken McDonald. Welcome to both of you. Glad you're both here.

00:57 - Ken McDonald (Co-host)
Thank you, I had to tear myself away from Civ 7 today.

01:03 - Jonathan Bennett (Host)
Yes, there is that little game that has released Civilization 7. You running that on your Linux machine?

01:08 - Ken McDonald (Co-host)
Yes, under Steam.

01:10 - Jonathan Bennett (Host)
Yeah, under Steam's Proton.

01:11 - Ken McDonald (Co-host)
Yeah, I couldn't find a native Linux version to buy and install.

01:17 - Jonathan Bennett (Host)
Probably not, not at this point. Some of the earlier Civilization games, I think, did have native Linux ports.

01:23 - Ken McDonald (Co-host)
Well, they said this was coming out with native Linux support.

01:27 - Rob Campbell (Co-host)
Just play Free Civ until the native one comes out. That's a good one.

01:32 - Ken McDonald (Co-host)
Yeah Well it's got to support Steam True.

01:40 - Jonathan Bennett (Host)
Anymore, though, you have to kind of wonder when they say that it's coming with native support. Are they talking about Proton? All right, well, we've got news, and, uh, rob is going to be brave and kick us off on talking about the, uh, the, the big news from this week. In fact, before we got the show started, somebody in our chat room said did you see this news? I'm like, yes, we saw that news. That's probably what we're going to be talking about for most of the show. So, rob, in in the way that only you can take it over and tell us what the latest drama is in the open source world.

02:10 - Rob Campbell (Co-host)
So last week Jeff and I shared some overlapping stories involving some drama, fighting about Rust being in the kernel, and then he evolved into Hector Martin and involved in and saying I'm out of here removing himself as the maintainer for the Apple silicon kernel code and deciding he was no longer going to upstream Asahi code into the kernel. Well, this week the plot thickens and there is a lot to update everyone and I'm going to try to hit the high points in the order that they happen that they came out. So last week it was hinted that Apple Silicon maintainer co-maintainer I mean Sven Peter may find a way to keep upstreaming into the Linux kernel. And good news shortly after the show it was confirmed that Sven will keep patches going upstream, and helping him will be another existing Asahi developer that I haven't really heard of. I think he's only been around for a year. His name is I may say this, wrong. I'm sure I am Jannie Grunau. So Jannie has been working on the downstream Asahi kernel since April, as well as working on a number of drivers. Jannie has the blessing of both Sven and Hector Martin.

03:48
But with this blessing comes some other bad news for Asahi Linux, at least a few days later, where the founder and lead project developer on Asahi Linux, hector Martin, has announced he is resigning as the lead developer, stating it's become less fun over time. A lot of users are expressing frustrations that they don't support the M3, the M4, or certain features aren't working and it's just not being fun anymore. Here's the quote from what he said. Hector Martin says quote I'm resigning as lead of the Asahi project, effective immediately. The project will continue without me. I'm working with the rest of the team to handle transfer of responsibilities and administrative credentials. My personal Patreon will be paused and those who support me personally are encouraged to transfer their support to the Sahi Linux Open Collective. Github sponsors does not allow me to unilaterally pause payments, but my sponsors will be notified of this change so they can manually cancel their sponsorships. I want to thank the entire Sahi Linux team team, without whom I would have never gotten anywhere alone. You all know who you are. I also give my utmost gratitude to all my Patreon and GitHub sponsors who made the project a viable reality to begin with and from me you know know we thank you too, hector, for your hard work, and it sounds like many aspects of the project you know were becoming frustrated. I can understand that, and this fight over the kernel maintenance issues was maybe just kind of a final straw, that that tipped things over.

05:44
So then bring us back to the kernel maintenance issues and the fighting that was going on over rust in the kernel and a maintainer blocking and all that Everything that. If you need a recap, check out last week's when Jeff talked about it. So it appears folks have decided to create a Rust kernel policy to avoid future conflicts. It lays out key topics like how Rust for Linux is not an effort by the Rust project or the Rust foundation. It says how many key kernel maintainers do support Rust in the kernel and how changes are not allowed to be introduced if a C change breaks a Rust-enabled build, and with the exception for Rust subsystems. Also, it states that duplicate C or or rust based drivers are not allowed. However, subsystems may temporarily allow them to make it easier to introduce rust support and get it working smoothly. You know so.

06:57
So this policy, it was uh published by Miguel Ojeda, uh, uh, uh, and he is a Rust for Linux contributor. So you know, personally, I guess I would feel maybe a little confident about this policy if it actually came from Linus Torvalds, greg KH, the Linux Foundation foundation. I'm not sure how individual contributors, especially one aligned with Rust, is doing this policy, and maybe I haven't seen it yet, but I guess at this point it just kind of seems like you know. If an employee you know means like I'm going to write a policy and tell my CEO how he should act, I don't know.

07:46 - Jonathan Bennett (Host)
Miguel is essentially the Rust for Linux guy. He has been the guy that has made all of this happen. So if you were to ask me who is the one person that could write the Rust kernel policy, it would be him.

07:57 - Rob Campbell (Co-host)
But he's also on the Rust side. I guess I would like to see support on the C side of the kernel. Oh, I see what you mean. I see what you mean. You know we're the C guys. You know it's the C guys who are telling the rust guys to get out of here, that your code doesn't belong here. Whatever, you know, I'm maybe exaggerating. So now it's rust guys say oh, here's a policy, so you can accept our stuff. Where I'd really feel more comfortable if is the C guys saying here's the policy, we'll accept your stuff under these conditions, because it's like me saying no, boss, you're going to accept this because I wrote this policy saying you are.

08:37 - Ken McDonald (Co-host)
We've got to remember this actually got started as a hobby for one of the Linux kernel maintainers.

08:44 - Jonathan Bennett (Host)
Yeah, and then everybody realized it might be a good idea to be able to do this. Yeah, so I have thoughts on this. First off, there were bound to be conflicts. We talked about this last week. There were bound to be conflicts between the C coders and the Rust coders. They come at things from a different perspective. It does add to the maintenance burden for the seam maintainer, so, like it's not surprising that there are conflicts and that it's taking time to work through it.

09:12
Um, I, I understand that hector is just tired and needs some time away from it, and that's fine. Good for him. Like, good for him for realizing that he needed to step down, taking some time away. Hopefully he'll find you know something else to be able to pay his bills, be able to move on to something that he really enjoys. We, like, we wish him well. I don't think that, especially because we have other people already stepping up to maintain it.

09:39
It sounds like the, the Apple Linux on the Apple Silicon is going to continue on. There's a lot of people that are using it now and excited about it. It seems to me that Rust in the kernel is going to continue on, at least for now, and again, we talked about that last week. So this in and of itself probably should not have been a huge story, right Like, in some ways, of course it is, but on the other hand, it's it's kind of it's normal around the kernel that people burn out, people get tired of doing it and they move on. And you know, dozens of people have gone through the same thing and have not written a huge blog post about it, and we've probably never talked about it.

10:23 - Rob Campbell (Co-host)
Every once in a while we'll talk about one that says I need, I need a break from it every once in a while, and then, sometimes right away afterwards, you have another one that has issues with the kernel so.

10:36 - Jonathan Bennett (Host)
So I've got a story here too and it's uh, it's a follow-on, very much a follow-on. So car Herbst has been a no, no view. No, we always used to say no view and then we got told no, no, it's nouveau, it's French. So the nouveau driver developer, he is a Red Hat employee and has been working on Mesa and some other things. And has been working on Mesa and some other things. He today has announced that he is resigning as the Nouveau driver maintainer and it's interesting, right. So he starts out by saying that he's been thinking about doing this for a long time because he's not really doing the work anymore. Like he's on there by name but he's not reviewing, he's not maintaining, he's not writing any code for the Colonel, he's moving on to other things, um, and he picked this particular moment, when the Colonel was sort of there there was a little bit of a drama, bit of a flame war going on, um to to make his statement that he was stepping out. I consider it unfortunate that he decided to make a sort of a political statement about it on his way out. So one of the other maintainers, in talking about Rust, used the term, use the phrase we are the thin blue line, and this is, this is an american politics thing for so. So for those of you that are not up on american politics, first off, lucky you, uh. Secondly, this is a reference to the police and the idea that the thin blue line is the police doing their job to prevent society from falling into chaos. That is a well okay.

12:31
So in some places, among some people, that is a very controversial thing to say. In other places, among other people, that is not controversial at all. So let me put it this way I live in Oklahoma. Not controversial at all. So let me put it this way I live in Oklahoma. In Oklahoma, we can say this and nobody cares. You even get some nods like yeah, yeah, of course People can imagine what would happen if there weren't a police presence at all in a society. You would have problems. You go to certain other places and you make this statement and you just about get rotten tomatoes thrown at you because, well, it's, in some places it's considered to be a racist statement and I'm sure in some places people do mean it with that sort of a tinge. I will say that in the times that I have heard it, at least in person, I do not think that is the case. So that's the background for those of you that are not involved in US politics. That's the background for those of you that are not involved in US politics.

13:30
Carol Herbst says in response to that this is not okay. And he says this isn't creating an inclusive environment. It isn't okay with the current political situation, especially in the US. And then he says this and this is really interesting A maintainer speaking those words can't be kept. That is an interesting statement.

13:51
I disagree with that statement and I think actually that is a dangerous statement to make, because that is the statement that says someone that is from the other side of things politically needs to be cast out of this project, and that's not what open source is supposed to be about. So I reject this. I think Karel Herbst is wrong, and you can agree with him and disagree with me, and that's fine. That's one of the interesting things here is like I'm okay with you disagreeing with me, but he's not okay with people disagreeing with him. Anyway, I see this honestly as a disgruntled employee that's just trying to make a mess on his way out of the company. I think that is what's going on here and I think we should just treat this as that. This is someone that's mad, that's leaving, is worked up over something or other, but you probably shouldn't take their words too much to heart, too seriously, because that's literally what he is doing he is trying to cause drama on his way out the door.

14:51 - Rob Campbell (Co-host)
So that that is my take on this, and feel free to uh take, have a slightly different take if you guys want to well, you know, maybe in defense of of the maintainer who said it, you know whether or not what his intentions were or beliefs are, or how he meant to use it. The phrase alone is, at least some places used to be, to represent police or being that, that thin line between you know, the public and and everything going bad. So it's it's like some. One defense, at least I've heard, is that you know all he was saying is that, uh, you know, as a maintainer, we're just the people here kind of making sure the bad stuff doesn't come through and and just kind of protecting it. And and you know, saying it that way, I mean maybe if, because the phrase is seen negative in places, maybe just saying you know where the place of the colonel would have been the safer route, but you know, maybe it's from a place like Oklahoma and did not even think about that, the fact that it's, it's.

16:15 - Jonathan Bennett (Host)
Yeah, it's probably worth pointing out that it may not have been the the wisest thing, like it was not the safest way to make his point to do it this way. Um, ken, you have thoughts?

16:26 - Ken McDonald (Co-host)
I do. I'm gonna be putting words into some people's mouths with my thoughts here, so I'm going to apologize up front.

16:35 - Jonathan Bennett (Host)
Do it carefully.

16:37 - Ken McDonald (Co-host)
But from reading everything I get the impression the maintainer was trying to say that as maintainers, we want to try to keep chaos out of the code Mm-hmm. As maintainers, we want to try to keep chaos out of the code Mm-hmm and unfortunately the phrasing he used to say that was a politically charged one.

17:01 - Jonathan Bennett (Host)
Yes.

17:02 - Ken McDonald (Co-host)
I have seen a lot of phrases over the years that I've watched TV, listened to the radio and now even listened to the streaming what streams on the internet. Seen some phrases that as a youth I didn't take any notice of, become politically charged and unfortunately, I think this is one of those situations where a phrase has picked up the charge and shocked somebody else yeah, that may be another thing.

17:47 - Rob Campbell (Co-host)
You know, maybe it's not just the location where he's at or the people he hangs out with, maybe it's. I don don't know how old this maintainer is either. Maybe you know he. He grew up, you know, like you, I guess.

17:58 - Jonathan Bennett (Host)
Yeah. So I would say that we have to, like, make some assumptions. No matter what we go to to understand all the like shades of meaning, you have to make some assumptions and that's that. That's just the fact of it, right? So you have to either assume that he knew that this was going to be sort of a controversial thing to say and said it anyway, or didn't think about it being a controversial thing to say and it's just something that is part of his vocabulary. I, I don't really care. People say controversial things on the LKML all the time. That's not news, that's not new. These things happen. Personally, about all of this story, the one that bothers me the most is the idea that if someone makes this statement, they need to be pushed out of the project, and I am not in agreement with that. I think that is a terrible idea.

18:56 - Rob Campbell (Co-host)
I mean, let's be fair about this. We're talking about the no, how do you say it? Nouveau, nouveau. We're talking about the Nouveau driver, which is like anyone who knows anything. It's like the first thing they try to replace with the actual proprietary driver, unless they are all about freedom and don't want the quality of life that the proprietary one gives you. But I know that's the first thing I did when I had NVIDIA. And well, now the first thing I do is I just don't use NVIDIA, nvidia, and well now, the first thing I do is I just don't use NVIDIA.

19:33 - Ken McDonald (Co-host)
What I want to bring up, though, is you've still got two people that are going to be remaining as Nouveau kernel maintainers, and they're also going to be working on replacing it with Nova.

19:47 - Jonathan Bennett (Host)
Yeah, that's an interesting point.

19:49
Nouveau is destined for legacy support, and they've got some new stuff coming along to replace it. So if we want to be a bit harsh here about this particular maintainer, according to even his own words, nothing of value was lost. He wasn't actually doing the work anymore, and so that's why I say this is sort of like a disgruntled employee trying to make a mess going out the door. He could have just you know nicely said I haven't been doing this work for quite a while, why don't we go ahead and take my name off the list? And he decided to make a political statement, a political statement that will rile up people in exactly the same way as the thin blue line statement did. Right, so you have the people on the other side of this that are going to get mad because he took kind of like me, I get a little frustrated by it that he took this side of an argument and you have people that got mad that he took the other side of the argument, and so like there's just no making everybody happy.

20:54 - Rob Campbell (Co-host)
Yeah, anyway, I don't know, not to bring it up at all and just say you know what I'm, I'm out.

21:01 - Jonathan Bennett (Host)
There is there, there has been. I've seen a couple of projects now that they've talked about this and it's just our rule is we do not talk about, we do not make political statements at all, period. We do not care what side it is the partisan politics thing we just do not do it in any of our official. You cannot come to Discord, into our random off-topic channel and talk about political things. We do not make political statements based from our you know, our ex-account or our Facebook account. We do not ban people from making political statements on their own accounts and I personally I think that is probably the best direction to take with open source projects and so you know in that case. So say that the kernel we're kind of theory crafting here say that the LKML came to that point. Then it would just be a hey, we understand the point you're trying to make, but this is a sort of political statement. Please don't.

21:55 - Rob Campbell (Co-host)
LKML Linux kernel mailing list. Yes, yes.

22:00 - Jonathan Bennett (Host)
All right. Well, let's talk about some better news. Actual progress Code being written.

22:05 - Ken McDonald (Co-host)
I've been waiting to talk about this, Ken tell us about Handbrake.

22:10
I'll be more than happy to, especially since this is coming from Marius Nexter. He wrote about Handbrake 1.9.1 being released this week as the first maintenance update to the latest Handbrake 1.9 series of this free and open source transcoder application for digital video files. It improves support for SRT files with OpenLAD theme subtitles, improves AV1 video decoding by updating to the LibDAV 1D version 1.5.1 library, as well as improving AC3 and EAC3 extra data in MKV files. Handbrake 1.9.1 also fixes an issue that could happen when chapter titles are not UTF-8, titles are not UTF-8,. Fixes FFV1 pixel format selection when using a hardware decoder, fixes GCC14 build failures on ARM64, and updates to the libjpg-turbo version 3.1.0 library for preview image compression. Now I'm going to cut it short here, since the rest of Mario's article covers fixes related to an operating system from Microsoft that I'm not too concerned with myself. If you are, I do recommend reading Mario's article so you can get information about those fixes.

23:54 - Jonathan Bennett (Host)
Yeah, so you know I use Handbrake. It's been a bit but getting like DVDs in particular up into my Kodi instance super useful tool. So good to see them making progress and keeping that project ongoing make in progress and keeping that project ongoing.

24:21 - Ken McDonald (Co-host)
I've used Handbrake with some of the time-shifted programs. I have to transcode them so that my Plex server doesn't have to worry about doing it.

24:26 - Jonathan Bennett (Host)
Yep, yep, I am on Team Cody, not Team Plex, but that's okay, we can agree to disagree.

24:33 - Ken McDonald (Co-host)
Well, actually I've been bouncing between that and just using a network connection and playing with VLC. That works too.

24:42 - Jonathan Bennett (Host)
All right, Rob, you just have all of the controversy today. That's what I do. This one is different. At least we don't have to step into political waters to talk about this one nope just legal waters yes, a little.

24:56 - Rob Campbell (Co-host)
well, potentially potential legal waters at some point here. But so you know, everyone is angry these days and and now, two of Jonathan's favorite companies are feuding Projects, not companies, projects, projects, right.

25:18 - Jonathan Bennett (Host)
Much less. I will just step in here and say much less love lost, lost. Yeah, much less love lost there we go For the company, particularly behind the distro there, than the distro itself. Good point.

25:30 - Rob Campbell (Co-host)
Good point. Good point though. So the two projects that I mean. I know Jonathan is using these right now. In one corner we have Jonathan's favorite Linux distro, fedora, and in the other corner we have a piece of software we all love on the show. We're all using it, we all use it Every panel here, and that's OBS Studio.

25:54
So Fedora, you know they like to package their own flat packs for software and they build it from their RPMs and apparently they don't do a great job maintaining them. So the issue is that the unofficial OBS studio flat pack. It's in poor condition, causing many users to report issues to the OBS studio developers, mistakenly thinking they are using the official build. A few weeks ago, obs requested the removal or yeah, to remove the unofficial OBS flat pack or just present it better. You know, keep it up, make it work, test your work before you publish it.

26:39
And it sounds like folks at Fedora were not taking this issue seriously, with one of them apparently even resorting to name calling by labeling the obs studio devs as being terrible maintainers. So now obs lead joel bethke has stated that if fedora doesn't remove all of obs's branding from the unofficial flat pack by february 21st a very special day for me for other reasons, it's my birthday. Uh, they will pursue legal action against them. And since then, you know, since these threats of legal action, you know this little obs studio project going against, well, someone with the backing of IBM in a way, in a way.

27:36
In a way. I mean, I don't know how that would turn out. Maybe they would just let them fend for themselves, but it looks like we're not going to get to that point. Threats were enough. Fedora got scared or they realized the flaws in their ways.

27:52
So since then, fedora contributor and member of the Fedora Engineering Steering Committee, neil Goppa, has opened an issue to remove the OBS Flatpak from the registry. Seems they already started this and there's a few things wonky and flat pack about it. So you may want to get your OBS studio somewhere else at the moment, but it sounds like it's not going to be in there in the future at all. Or maybe it's not today, I don't know. But yeah, it looks like they're going to back down and stop maintaining that and that's less work for them.

28:32
But you know, this isn't the first story we've had where developers have had issues with unofficial Flatpak releases being poorly maintained and causing issues for them. I think it was just a couple months ago. I had a story on an open SUSuss maintainer and uh bottles having a feud. If you don't recall, I don't know. Look about two months back, open seuss maintainer removed the donation link for bottles because uh bottles about open. Seuss was doing a poor job and I kind of like the the way obs is doing. It sounds like like a more maybe productive. I mean it worked pretty fast, but you know it really brings back. I mentioned this last time If you're going to maintain a package, make sure it works, test it or just don't maintain it. If you don't have the time, you know, focus your attention on what you do have time to do.

29:33 - Jonathan Bennett (Host)
Yeah there, there was one thing about this that really surprised me, and that is that, um, I didn't know that fedora maintained uh, these, uh sort of these packages in this way because, like, they have rpms and there are the official packages that are outside of the rpms. It's like why, why does fedora have uh their own like feed of this thing? That's one unofficial, but also not how fedora officially packages things, like I don't see the point of having this extra flat pack. Um, and then apparently also there was. There was something like depending upon how you went to install it, fedora was automatically redirecting you to the fedora version of it, which that does sound a little sketchy sounds a little bit like an ubuntu uh apt redirecting you to snap.

30:25 - Rob Campbell (Co-host)
Yeah, I guess they're not the only ones yeah, apparently so.

30:28 - Jonathan Bennett (Host)
Anyway, all all that to say you know it's a, you know it's a weird story. I'm glad that they finally put their heads together and figured out what they're going to do. You never like to see like two different open source projects threatening legal action. It's just not great.

30:49 - Ken McDonald (Co-host)
And it's not cheap.

30:51 - Jonathan Bennett (Host)
Well, threatening legal action is cheap. Actually following through is not. It costs lots of money. It makes lawyers very rich. Yeah, it would have been nice if the OBS folks had talked to someone, made contact with the right person, rather than having to threaten legal action first. Who is the right person than having to threaten legal action first.

31:11 - Ken McDonald (Co-host)
Who is the right person?

31:13 - Jonathan Bennett (Host)
that sounds like it sounds like neil gompa. He is, uh, he's a. He's a pretty reasonable fellow, um, and I've you know, I've talked to him before, I've talked to him about things and he's part of the fedora release team. He's part of the mainly the kde side of fedora, but he does more than just kde and it sounds like he's the one that finally made stuff happen here. I got to looking and I actually run OBS Studio from inside the Flatpak, and I do that because it automatically comes bundled with things like the Chromium browser extension so that you can show webpages inside of OBS.

31:48 - Ken McDonald (Co-host)
From within side of which Flatpak.

31:52 - Jonathan Bennett (Host)
Well, on this machine it's Pop OS, so it is the FlatHub Flatpak. Yes, the unofficially official Flatpak, because it sounds like. Didn't they say that was going away too?

32:06 - Ken McDonald (Co-host)
No, they were going to be updating the runtime in it. Hopefully, hopefully. So a lot of applications that you can get in Flatpak where they need to update those runtimes.

32:21 - Jonathan Bennett (Host)
So I'm sure what happens is they've got the Flatpak build set up as part of their continuous integration suite on GitHub and it's got the scripts to pull everything down. But to be able to do reproducible, to have a sane build process, they specify the version numbers on things and I'm sure it's just nobody's gone in there and messed with that script and updated that version number of what GNOME libraries are going to be on the back end of that flat pack. So that's fairly typical. That's just busy work that somebody needs to be reminded to go do.

32:56 - Ken McDonald (Co-host)
Set up a reminder. When the next version comes out. We need to update that script.

33:01 - Jonathan Bennett (Host)
Well, you can actually do that, boy. There are some people that are just the GitHub integration wizards and one of the things that they will do is they will set up to automatically run some of those you know go check for updates as a gi run and so like, every time, every time you do a release, you then get ping an email or a new issue. You can also set it up to open an issue. Hey, this library is out of date. Hey, you know this is out of date. You really got to update it and uh, yeah, it's. Uh, it's a. It's a lot like. There's a lot of wizardry in making the, the continuous integration on something like github, really, really sing well and do what you need to do, but when you have it set up and it's actually working well, it's really nice all right, so we can make some money off of teaching that the potentially.

33:50
I'm sure there's classes, I'm sure there's people out there that do it. So I want to talk for a minute about a couple of things that have landed, really cool things that have landed that, for various reasons, are not working or available yet. And the first one is in T sync not in sync but in T sync, not in sync but in T-Sync, and that is a. It is a synchronization primitive from the Windows NT kernel that is now inside of the Linux kernel and it is literally just there to make Wine work better. We've talked about this a couple of times and once just a couple of weeks ago I believe and this it makes a drastic difference on games and some of their fps numbers, um, some of them getting like 200 increases on fps, going from not playable at all to buttery smooth, um, and it's it, just it. It takes this way that the int kernel expects to be able to synchronize things and it moves it into the kernel and once you get everything wired up to use it, it is amazing. Now there are a couple of problems, which is why we're talking about it. One it's a character device, which means it's in slash dev, it's a care device. Slash dev, slash Inti sync was only read-write to users by default, and it is new enough that systemd doesn't know to change the permissions on it, and so that essentially means that it was useless out of the box, because you run all of your Windows games as your user, and that was not the same user as that device. And so you know, know, you either had to go in and change the permissions or set up a new group not something that every linux user knows how to do and so there is a. There's a patch now that's been pushed to the, to the driver. I don't know that it's been accepted, but they have indicated that it's acceptable. Um, that would change the, the defaults on that file to something like a 666 or a 444, make it more accessible. And then there's also work being done to get it inside mainline wine. So the other interesting thing here is you've got wine, you've got wine staging and you've got Proton. I believe there are patches in wine staging and Prot that will make use of NTSync, but it's not in upstream wine yet, and that's because the upstream wine guys are much pickier about the sort of hacks that they're willing to pull into wine. There is a patch there that someone is working on that they're hoping will be acceptable and so maybe it'll show up there soon. But again, I believe this has already landed in Proton and Wine staging. So that is that.

36:42
And then the other thing that has finally landed and I mean finally landed is the HDR protocol for Wayland. They finally merged it and this has been an open merge request for five years now. In fact, this is merge request number 14 on the Wayland GitLab and I am just quickly looking to see they're up to. Okay, they're only up to 381. But still, this is number 14 out of 381, a very early merge request, one of the earliest that have not been either merged or closed. Hundreds and hundreds of comments, people being less than useful in some of those comments, but they finally got, you know, everything straightened out and everybody happy, or at least happy enough. It finally got merged. And so now we finally have hdr and wayland and you say, oh, great, that means that we have it in chrome and firefox and everywhere else too, right? No, nope, nope, nope, nope, not yet. Uh, it is in kde. Um, I believe they did implement it in the still being developed version of gnome, so in like the next release of gnome. It's the still being developed version of Gnome, so in like the next release of Gnome, it's going to have some of these extensions exposed.

37:57
Gamescope has it working and the MPV video player has it working. And in fact I have done a command line tip in the past here where I have an alias that's HDR play. That has the various flags. You need to get the MPV to actually play a video in HDR, because I think at the time it wouldn't, by default. It might now, but at the time it by default did not do so. There is a patch in Firefox being worked on that would enable HDR for video playback.

38:30
Um, and I I don't know that it's um it is not made a whole lot of progress here. Recently I I was told that it actually worked for a small time, like a real short amount of time. It actually worked, uh, but it it now no longer works and I've tried a couple of times to um to compile it and get it working, and so you know it's. We're waiting. It's last updated two days ago. That one's been open for five years as well. It's Mozilla bug 1642854. And unfortunately the guy that is doing the work on that has not been around for a few days. So we are still sort of waiting to see what happens with that. So we've got two really cool things that have finally landed in Linux and you can't very easily use either of them, but it's closer we're getting closer, soon the hard stuff is done.

39:33
Yeah, that's actually true, that's actually true, I've really been thinking about going to Chromium and opening a bug and just say, hey, it's here, it's here, let's do the thing. Come on guys, I haven't found one. I haven't found an open bug in the Chromium source code. There's like a really old bug where someone's like work at x11?

40:00 - Ken McDonald (Co-host)
I don't think so, so I might just go reopen that one and update it. Ah wayland, so with this, does this mean we'll also start seeing blue filter like options?

40:06 - Jonathan Bennett (Host)
oh, those are already out there. Um, you can already run. Uh, like nightlight, I believe is. Is that what kitty calls their solution?

40:13 - Ken McDonald (Co-host)
yes, it is yeah, so that'll.

40:15 - Jonathan Bennett (Host)
That'll do red tent and blue tent, depending upon the time of day which I ran that for a while. I don't know if it actually helped me sleep, but it was cool.

40:22 - Rob Campbell (Co-host)
But you could now with this. It means I need a new monitor.

40:26 - Jonathan Bennett (Host)
It means you need a new monitor? Yes, go go, invest in like an oled.

40:32 - Rob Campbell (Co-host)
Uh, some big 4k hdr oled monitor uh, I have been thinking I really need to get a nice new gaming monitor. You know some of the games I've been playing with people and they're describing what they have like. Oh, that seems like that could be kind of nice here make sure you get one that has a display port.

40:53 - Jonathan Bennett (Host)
Oh yeah, because hdmi is sort of crippled, then you need to get kde, kde plasma 6.3 for managing that new monitor there are some interesting things in 6.3. Uh, ken, you've got that story, don't you?

41:07 - Ken McDonald (Co-host)
yes, I do well, let's do it and this story this week is coming from sarav Rudra and Marius Nestor. They both wrote about the latest KDE Plasma 6.3 desktop environment release. Now, according to Sarav, it aims to be the ultimate desktop for digital artists. Plasma 6.3 ships with a revamped drawing tablet page in the system settings app. That allows artists to easily map the entire screen area of a drawing tablet.

41:41
Surface shows information related to tilt and pressure. This is during your stylus testing Allows tweaking the pressure curve and range of a stylus and remapping or swapping of stylus buttons. Now KWIN will now snap elements to the screen's pixel grid, eliminating blurriness and gaps while producing a sharper, more crisp image on that new monitor. Rob, the overhauled fractional scaling makes the experience of using a high-resolution display much better than before. According to Marius, the ability to clone a panel, support for remembering the active virtual desktop per activity, an option to prefer screen color accuracy in KWin and support for viewing battery cycle count in InfoCenter are some other highlights of KDE Plasma 6.3.

42:42
The KWIN window and composite manager received improvements in choosing a default scale factor for devices with small screens and support for the automatic scale factor chooser to select a scale factor that's rounded to the nearest 5% rather than the nearest 25%. Also, it's now possible to temporarily disable KWIN window rules instead of deleting them. Also new is the ability to show low battery notifications for wireless headphones that properly expose battery information, the ability to configure the touchpad to be automatically disabled when plugging in a mouse, and the ability to set keyboard shortcuts to move windows between custom tile zones based on directionality. Now I've only touched on some of the highlights from Marius and Sarad's articles. If you do want more details, then follow the show note links to the respective articles.

43:54 - Jonathan Bennett (Host)
Yeah, I am not running 6.3 yet I have.

43:58 - Ken McDonald (Co-host)
I'm tempted to, I'm tempted to to again run the bleeding edge packages and I'm waiting to see, uh, which one I'm going to have when I reboot into tumbleweed tomorrow you'll probably have whatever you had when you left it, and you'll have to do updates to get something new I can't remember if I did an update right before shutting down or not yeah I usually do yeah, when was the actual 6.3 release date? Oh, let's see.

44:27 - Jonathan Bennett (Host)
I said I want to say it was back around the last couple of days looks like the 11th so four days ago is at least, when this says so cool, yeah, I will, so I will get already updated to it.

44:46 - Ken McDonald (Co-host)
I just need to pull up the uh about system information to check and see which one I've got.

44:53 - Jonathan Bennett (Host)
Yeah, yeah, all right. So, rob, we've talked about politics, we've talked about legal issues. What more could Linux geeks get offended and upset about? What else do you have in your toolbox?

45:09 - Rob Campbell (Co-host)
All right, here's one You're really going to get angry about here. System D is all that's left right System D. System D is all that's left right System D. So this isn't going to be a very long story and very few of us are likely to be using this feature, but I just thought it was really cool so I wanted to bring it up and share it with everyone and to all you System d haters, too bad.

45:39
So you all know, uh, well, maybe you don't all know, but many of you probably know what pixie boot is. Uh, pixie spelled pxe. So that allows you to boot from an, an iso on the network, you know. Kind of like, instead of booting from a USB or an optical drive or your internal hard drive, you can boot from a Pixie server set up, you know, set up for this. Well, system D.

46:10
I just wanted to bring the drama today, so I'm I hadn't got the love, but system D is adding a feature similar to Pixie Boot, except it is allowing us to boot from HTTP, as in boot from the internet, the ability to let SystemD boot directly into a disk image downloaded via HTTP. They keep saying HTTP everything I write. Hopefully they have HTTPS at least someday I don't know, but anyway downloaded via HTTP within the initial RAM disk in it RD during the Linux boot process. Eventually, this should allow you to point a UEFI to a single URL where it can then load the unified kernel image from and then in turn picking up the root file system and booting it up.

47:14
The guy who's been working on doing all this work. He's already looking to extend this into supporting NVMe over TCP and other features in the future, which is cool. So you know, maybe that's something everyone's going to use. That's a freaking awesome feature in system B, I think, and you know I like to see you do that with your system by five. I was combining V and five system V system five in its system.

47:46 - Jonathan Bennett (Host)
Yeah, this, this is. This is actually kind of fun. I'm reading through the Macedon thread here about what he is hoping to accomplish with this, and for him it's just to be able to spin up images and test them out faster, which you know that's cool in and of itself.

48:04 - Rob Campbell (Co-host)
I mean for that purpose if you're, I don't know, it's not too fast, it's not too slow to do it, just to Pixie boot.

48:11 - Jonathan Bennett (Host)
but I suppose if he doesn't want to download it, yes, yes, he also talks about being able to boot directly from a tarball. You can boot into a tarball now with just a uki, so like there's some, there's some interesting stuff there. Um, I would not terribly surprise me to see, um like netboot images, start to use this, yeah oh yeah, that'd be a pretty, a pretty slick.

48:36 - Rob Campbell (Co-host)
I mean, yeah, they already have in-the-net boots a very minimal download ISO and just gets it all from the internet. I mean you could just put it at the point where you just point to it and boom install, Yep.

48:49 - Jonathan Bennett (Host)
That's cool, very cool. I've got a bit of a geeky story here about the kernel and that is the timer frequency, and so there is a proposal to change the kernel's default timer frequency. I figure we could talk for just a minute about what exactly this means. The default frequency right now is 250 hertz and that essentially means that 250 times a second the timer interrupt runs. That means that 250 times a second the timer interrupt runs. That means that program execution goes away from user space, enters kernel space and the scheduler runs. It looks at what's running, what needs to run, pushes out the next slice of time in the scheduler. It runs at 250 hertz and the proposal is to push that to 1,000 hertz, essentially to schedule the system in one one-thousandth of a second slices, and that's fairly interesting.

49:50
And so Michael Larable looked at that and said well, I wonder what that's actually going to do on benchmarks, as he is wont to do, and he ran several different things. Some of them did much better under the new proposed 1000 hertz. Some of them did better under the old 250 hertz. Lama, for example, the AI benchmark, did much better. Nginx did better under the 1,000 hertz, but then you get to things like Darktable did much better under the 250 hertz, and it kind of extrapolates out to programs that are just going to. They get a task and they churn on that task for a significant amount of time, like what Darktable does. And so by significant I mean in like in kernel time it's going to churn on that for a while. Then you will probably do better with fewer interrupts, with fewer of these timer interrupts. But when you're looking at something that is sort of very, very quickly moving from thing to thing and sometimes Nginx is going to be one of those you might do better with the faster interrupt timing.

51:08
And so it's interesting just to look at the different use cases. Some games do better with one, some games do better with the other, and for some things it seems like it's basically a dead heat. Uh, it has been suggested I've seen it elsewhere that this might improve system responsiveness to user input just a little bit. Um, I don't know if that's enough that the majority of us would really be able to tell, but theoretically it could. Uh, and then the. The one other thing is that a lot of distribution kernels so ubuntu's kernel already compiles in and says let's use the 1000 hertz. Uh, so it's just it's an interesting little change that may or may not land, but I figured it would be good to talk about what exactly that means, why we care about I think believe this is also called the tick rate why we care about that.

51:59 - Rob Campbell (Co-host)
So is that something that would be user configurable, or would you have to compile your kernel for that? Do you know?

52:05 - Jonathan Bennett (Host)
I believe it's a, I know it is. It is a compile time option. I'm not sure if it is possible to change that on the fly or not.

52:15 - Ken McDonald (Co-host)
Or you have two versions of the kernel uh compiled and set up your boot menu, whichever one you're using, so you can switch between one to the other, depending on what you have to do that would take a reboot, though, so you know that's quite a bit longer than just being able to cat a value to something in slash, slash, slash, sys, slash proc yeah, I'd hate to reboot every uh time.

52:39 - Rob Campbell (Co-host)
I want to do a different uh, you know, run something else, but I guess it's better than recompiling the kernel every time.

52:46 - Ken McDonald (Co-host)
Absolutely, I'm just wondering, you know like unless there's some way to switch kernels without rebooting uh only sort of only sort of basically run up a vm well, no, there's, there's more than that.

53:03 - Jonathan Bennett (Host)
So you can you can jump execution. In fact, this, this used to be a thing that was done more commonly you can jump execution into a new kernel, but it's actually very similar to a system reboot. It's basically doing a reboot without turning the power off.

53:20 - Rob Campbell (Co-host)
Yeah, the old system V. Couldn't you just do a new init? And I don't remember.

53:30 - Jonathan Bennett (Host)
Yeah, there was a way to do it. I think there's even a CPU instruction to do it. I don't know, that is not something that I ever had to do it. I don't know, that is not something that I ever had to do and it's not something that hardly anybody does anymore. So it's sort of black magic to me. I'd be.

53:48 - Ken McDonald (Co-host)
I'm just curious if I could improve my web server's performance with the It'd almost be like having to reboot and bring back up just to run a one application, like switching from dark table to engine x yeah, it's, it's, but you're replacing the running kernel out from underneath itself.

54:12 - Jonathan Bennett (Host)
I will see if I can find this, this bit of internet lore that I have obviously forgotten. It's been enough time.

54:18 - Rob Campbell (Co-host)
Use that as your tip next week.

54:23 - Ken McDonald (Co-host)
How to change your kernel while running.

54:25 - Jonathan Bennett (Host)
See, that's going to make it sound like I endorse this idea. I do not Do a real reboot, it's much safer. Endorse it All. Right, ken, tell us what's new with OpenSUSE Tumbleweed.

54:40 - Ken McDonald (Co-host)
And this is coming from Bobby Barsoff. According to Bobby, big changes are landing in OpenSUSE Tumbleweed, starting with Snapshot 250211, so it's actually already out. Selinux will become the default mandatory access control system for new Tumbleweed installations. Now SELinux, or to say it the full name, security Enhanced Linux, is a security feature built directly into the Linux kernel that controls what applications and users can access on a system. It sets strict rules about what programs and users can do, which helps prevent unauthorized access or damage if a system is compromised. Now we've heard Jonathan talk about that sometimes. Now, according to Bobby, the maintainers have confirmed that existing tumbleweed installations will remain untouched. Thankfully, if any user prefers AppArmor during a fresh installation, the installer will provide a simple option to switch to AppArmor. It is worth noting that this shift does not apply to OpenSUSE's Leap 15.x series, which will remain on its existing security model For Tumbleweed. The first boot after installing SELinux can take a little extra time to complete system labeling so don't be alarmed if things seem to run a tad slower immediately after setup.

56:22
Bobby also includes a link to OpenSUSE's announcement in his article. As always, you will find a link to Bobby's article in our show notes.

56:33 - Jonathan Bennett (Host)
Yeah, so kegsec is the command I was thinking of of and that will be very interesting to look into. Um, but the, the idea of a uh sc linux system relabel or, in this case, a, an initial labeling of the system, that takes some time. Um, yeah, I have, I've had to sit and wait for machines to finish that before they will continue booting up. Uh, the, the relabel.

57:00
It can take some serious time enough time to get a cup of coffee or to eat a full meal it depends upon how big your hard drive is and how fast it runs it depends on how big your cup of coffee is too that too, so like if you have a big raid array and you have to relabel everything on it and it's multiple terabytes maybe just go to go for a vacation I I've more than once I've started a process and then just gone to bed and hope it'll be done in the morning wake up and it's only halfway that's usually not a good sign.

57:33 - Rob Campbell (Co-host)
Uh, it's been years since I've done that, but I've done it. Yeah, no, I mean doing something like a dd rescue on a disc okay, I guess, yeah, I've done things like that, but it's still been quite a while since I've had to do that. I've got it. I've got a disc it.

57:49 - Jonathan Bennett (Host)
I just gave up on the dd rescue because on a 500 gigabyte disc I was down to five megs that it couldn't pick out in various places along the disk and so just today I started the copy back over to the new SSD and then we'll pop that into the computer and see if it's got enough of it to be able to boot or not. We'll see, hopefully. Anyway, I see Linux. I'm a fan of SC Linux. I think it is worth it. Um, it's very fascinating to see open susa going in sort of a even more fedora ish direction. Does for use app image oh no, fedora uses, uh, sc linux, oh oh, they're going. You know, never mind.

58:32 - Rob Campbell (Co-host)
Yeah yeah, yeah, yeah, so um yeah.

58:33 - Jonathan Bennett (Host)
Yeah, yeah.

58:35 - Ken McDonald (Co-host)
So Ubuntu uses app armor Correct.

58:41 - Jonathan Bennett (Host)
I'm more familiar with app armor but yeah, I think they do basically the same thing, so much so that I don't think you can effectively run them at the same time. I think I step on it, I'm not sure you'd want to run them at the same time. It's like having two antiviruses installed at the same time on a Windows machine.

58:57 - Ken McDonald (Co-host)
Like I said, I don't think you'd want to run them at the same time.

59:01 - Rob Campbell (Co-host)
When things don't work, I don't run either. I just turn the whole thing off.

59:05 - Jonathan Bennett (Host)
Yeah, when things don't work, you just walk away. There's no running required. If you have to run away, things really went badly, all right, oh, all right, rob, let's talk about etsy keeper etsy keeper, uh, etc.

59:25 - Rob Campbell (Co-host)
I'm gonna pronounce it, etc. Keeper, um, I'm not gonna bring up anything, uh to display. So this is gonna be an audio only you get to look at. You get to look at me if you're watching the video, which is pretty awesome too. But ETC Keeper, I linked some documentation, the show notes, to the Ubuntu documentation on it. But you can use. You know it integrates with like app TNF, yum. I know there's a Pac-Man integration, even though I didn't necessarily see that was official pacman being arch. But anyway, what etc. Keeper is because that's what I've been calling it it's a version control for your system configs or your etc. At c folder. So when you install this and initialize it by default, it commits daily. But then it also commits before and after package installation.

01:00:25
If you got the hooks into your package manager, which most will be by default, you can also manually make a commit to ECT. Maybe you were in there and made a manual change to an ETC file, an Etsy file, and you want to save that, make a commit. You just do. The command is sudo etsykeeper, commit quotation and you put like a comment reason in there for your commit. It's very much like git, but for etsy.

01:01:04
Um, you can do other things such as uh, etsy keeper space, vcs, uh space log, and then do a specific file like slash, etsy, slash password, and then you can see the, the commit log for that specific file. Or you can do them all. Would be crazy. You can do the same thing etsy keeper vcs status, which will just show the status, like it'll show that maybe you're on the master branch and if there are any, any changes pending to be committed if you change a file or something.

01:01:41
And then you could also do you know lots of other things. You could do like a diff to see what's the difference, see what's changed since you've done a commit. So if you want to keep versioning of your Etsy just in Etsy folder Etsy directory Folder is a Windows term, don't know why I let that slip out your Etsy directory. You know this is an excellent tool for you because it's pretty much automatic, has hooks, your stuff and it'll just do its thing. And then you can always look back at old commits and see what changed. If something's not working right for you and you don't remember what you changed, you can look back at that too, just by uh, you know, doing a log and compare it with this commit to that commit and make your change.

01:02:35 - Jonathan Bennett (Host)
For saying folders instead of directories. For absolution you must say the GNU, linux, copy pasta three times I'll see myself out, bye-bye. Nah, that's fun. All right, ken, you've got yet another Pipewire command.

01:03:00 - Ken McDonald (Co-host)
Yes, I do, and this week is PWConfigure. This command can be used to debug Pipewire server and client configuration files. Now, as with all the Pipewire commands and let me go ahead and bring up my screenshots here you can use the dash H or dash dash help to show a basic help screen and of course, you'll see, for those of y'all listening and you'll see here for those of y'all listening, I've brought up a terminal showing pw-config-space-h and one of the commands that it mentions that you can use in this dash dash version, which I also ran, which gives you the compiled and linked version that the command is working with. Now the default action when running PWConfig, without any options or flags, is to list the default pipewire config paths. And for those of y'all listening, I just switched to another screenshot of my terminal where I showed running PWConfig and the default config path in this case is slash user, slash ser, slash pipewire, slash pipewireconf, slash pipewireconf. So that's what my pipewire system's default configuration is running off of, and you can also use a dash end or dash dash name and give the name of the configuration file that you're wanting to look for.

01:05:17
In this case I did pw-config slash pipewire, slash pipewire, dash pulse dot config and it also showed that I've got two other configuration files set up to override that. One is home slash dad slash dot config. Slash pipewire, slash pipewire dash pulse dot conf. Dot itself is RT dot priya dash 83 dot comp. This is to change the real-time or configuration options for the pipe wire pulse. Next thing that you can use a PWConfig for is to list out the configuration file, and it will start off by giving you the config name, or configname, and the name of the file, its path, and also list out everything in it like context properties, contextsba, libraries, contextmodules. This actually takes several pages or several screenshots here as you. I'm going through For those of y'all listening. There's a lot of text there.

01:07:04 - Jonathan Bennett (Host)
A lot of it's commented out too. Yes.

01:07:08 - Ken McDonald (Co-host)
If you wanted to change it. You can take and uncomment, like one of them is for Pulse Rules. You can uncomment the client name equals Firefox and then uncomment applicationprocessbinary equals teams and then uncomment. Applicationname equals speech dispatcher and it uses the tilde as a wild card. In front of that, can you guess what that might do, jonathan?

01:07:44 - Jonathan Bennett (Host)
Probably a match. All it's going to that exact change. No, I'm not sure.

01:07:54 - Ken McDonald (Co-host)
I haven't tried it myself since I don't have Teams installed, but I suspect it would allow you to set it up so that it would use Firefox with the Teams binary for passing speech.

01:08:10 - Jonathan Bennett (Host)
Could be, could be.

01:08:12 - Ken McDonald (Co-host)
I haven't played with that part yet. I haven't played with that part yet. And then you can also use the command by putting dash in the name of the configuration file list and follow it with contextproperties, so it will list the properties that you have. In this particular example, it shows that it's going to configure properties in the system where you can set memwarn, memmlock, which is a Boolean. You can set a true or false. It also shows what you can set the defaultclockquantum limit. You can say have a rule set up so that you can change it to 8192. It gives you quite a bit of capabilities.

01:09:13 - Jonathan Bennett (Host)
Did you come across the dash R flag in that?

01:09:18 - Ken McDonald (Co-host)
Yes, I did and I played around with it, but because I didn't have anything worthwhile, You'd use that when you're doing the merge to reformat, basically.

01:09:32 - Jonathan Bennett (Host)
Yeah, the other thing that I found useful just now looking at it "-r will go in and strip out all those comments. So if you don't want to see all the lines commented out, dash R will just give you the meat of your config file. Yep, all right, I've got a pretty neat command line tip. At least I think it is. I'm going to see if I can share. There it is. You'll be the judge of that. You'll be the judge of that. Well, that's fine.

01:09:58
So, um, I am looking at the pw dump command output here and it is a very, very long lot. This is all from a single run of pw dump, lots and lots of output, because it's dumping the entirety of your, your pipe wire status. Well, it's in, it's in JSON and there is actually a JSON parser and that is JQ. And so I I spent a little bit of time fiddling around with this and it took me a little while to to get exactly what I wanted this to do, but I finally found it, and so you. So you could pipe the output from pwdump into jq and then you can do something like this and it will give you, essentially you're writing a query. It's sort of similar to an SQL query in some ways. But we can break this down one thing at a time. So if we just pipe PW, dump straight into JQ, it basically just passes it through unchanged. We can then, and you generally, put your JQ command inside of single quotes. So we can do like just, I think a dot is going to pass it through unchanged. But if we were to do dot and then single square brackets, it's going to essentially strip the top level of brackets off. So it's, and in JSON terms what it's doing is it's going into that first array and it's just giving us the output of the first array. Well, so then, once you do that, you use the literal pipe symbol to pipe from your individual bits inside of JQ. So our dot square brackets dives into the first array because the PWDump output puts everything in an array.

01:11:51
And then we can say we want to select where the ID. So I'll pull this back up and you can see it here the ID field is set to 19. And so this is another object inside of this JSON field. We want to select the object where the ID key is set to 19. And so it's just a select and id equals equals 19. You know, we run that and it gives us just that little piece of the entire PW dump.

01:12:21
And so you could you know you could be a little bit more creative here if you wanted to. I believe you could say something like type, and obviously there aren't any type 19s, but we could say I want to see the types where it's a pipewire interface module and that's going to give us all of those, and so that's an easy way to kind of navigate and pick out just the bits of information you want. And JQ is super interesting for a few other things. For example, I use it in a GitHub continuous integration pipeline. We use JQ to pull things from the GitHub API and it gives us to it in JSON and then we run that through JQ to pick out just the one little bit of data that we need to set into a variable to be able to do the next thing. But it is a really useful piece of kit, piece of tool to have in your s of sysadmin and command line cowboy tool bag.

01:13:23 - Ken McDonald (Co-host)
I think I can find some uses for that just with exploring my pipewire configuration.

01:13:32 - Jonathan Bennett (Host)
Yeah, no, that's what. Last week when you saw it, when you did that and I saw that it put out so much JSON output, like, oh, there was a there was command that did this and I went and looked for it. It's like, yeah, jq, that's what. Last week when you saw it, when you did that and I saw that it put out so much JSON output, like, oh, there was a there was command that did this and I went and looked for it. It's like, yeah, jq, that's right, had to do a little bit of reading on exactly how it's how you do the parsing and how you do a select and how you do piping and all of that, because you know, the first time I went to do it I didn't include the single quotes and so I tried to pipe it and bash is all like we have no idea what the select command is.

01:14:01 - Ken McDonald (Co-host)
Keep fine. Yeah, had fun with those types of responses.

01:14:08 - Rob Campbell (Co-host)
Yes, yes, dnf install select.

01:14:13 - Jonathan Bennett (Host)
Yeah, and I lag we. I don't know if we've done JQ. We may have done it, but we didn't dive into the syntax and it was just a great fit for what Ken came up with last week.

01:14:25 - Rob Campbell (Co-host)
So it's a follow-on.

01:14:28 - Jonathan Bennett (Host)
I may have mentioned it as a way to parse JSON, but now we actually dove into.

01:14:35 - Ken McDonald (Co-host)
Here's how you write JSON code with it, because I need to use it to dig through to find out where I got a loopback device being generated in my configuration files yeah, you might be able to do exactly that.

01:14:48 - Jonathan Bennett (Host)
Uh, it calls it the. Uh, it's like said for json. That's what they say on the official site jq is like said for json.

01:14:55 - Ken McDonald (Co-host)
So they said jq's like said for JSON. They said JQ is like said for JSON.

01:14:59 - Jonathan Bennett (Host)
Yes, they said it first. They said it first. All right, I'm going to let each of the guys give their plugs whatever they want to Get the last word. We'll let Rob go first, all right.

01:15:10 - Rob Campbell (Co-host)
Whatever we want, this one's going to be exciting. Guys, you ready, here we go. Come connect with me To find me. You go to robertpcampbellcom and there you'll find near the top and a nice little semi-transparent gray rounded corner rectangle when it says I'm Rob Campbell. At the top, at the bottom of that, you'll see some icons. 'll see a linkedin icon. That's way to the right. That takes you to my linkedin profile where you can connect with me there. Or, if you prefer, twitter I'm not all that active there, but I have an account there if you just want to uh connect with me and uh show off to your friends that you're connected with me. Um, but not a lot of activity.

01:15:58
The place I have been the most active lately is Blue Sky, and that's just because that's somehow where I suddenly got the most followers. I'm over 500. So let's have a goal and let's get me to 1,000. So Blue Sky. Or I am still active on Mastodon. Lots of times I double post these. So if you like the freedom of Mastodon, come and connect with me there. And if it doesn't matter to you or you want to do them all or you want to do none of them it doesn't matter. This last one. Here is the place where you can donate a cup of coffee to me, and I haven't had one for a good it feels like a month or two since around Christmas time, so I really could use a coffee.

01:16:45 - Jonathan Bennett (Host)
Could, it be because nobody has the money to.

01:16:48 - Ken McDonald (Co-host)
They're still trying to pay off that debt from Christmas it might be.

01:16:52 - Jonathan Bennett (Host)
All right, Ken, anything you want to plug?

01:16:55 - Ken McDonald (Co-host)
Yes, I do have an article by Liam Proveman where he talks about an old sweet learning new tricks. It's in the show notes, so follow it and read what the Liberty Office announced at Fostin.

01:17:13 - Jonathan Bennett (Host)
An old sweet learning new tricks. I will have to take a look at that. All right, thank you guys for being here. I appreciate it All right.

01:17:23
So if you want to follow me, you can find my stuff over on Hackaday. It's where Floss Weekly is at. That's where my security column goes live every Friday morning. We may skip it for a week. Uncertain yet the the future is very uncertain for me here for a few days, um. But if I do end up going off to a security conference, which is the current plan, then probably won't have the security column this week and we'll double up on it next week, um, but anyway you can. You can follow me there at the hackadays, um. Other than that, we just appreciate everybody being here, appreciate those that support twit. If you're not a part of Team Twit, you should really come join Club Twit. About the price of a cup of coffee per month, you can support the shows and the network that you love and we would love to see you there. But regardless, we appreciate everybody that watches and listens, both live and on the download, and we will see you next week on the Untitled Linux Show.