Untitled Linux Show 188 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)
This week we're talking about Jeep Heart's Head hitting 1.7, opensuse's new installer release, what's new with Thunderbird, what's new with the Linux kernel and the big thing that GTK is dropping. You don't want to miss it, so stay tuned.

00:17 - Leo Laporte (Announcement)
Podcasts you love.

00:19 - David Ruggles (Co-host)
From people you trust. This is Twit you trust.

00:27 - Jonathan Bennett (Host)
This is Twit. This is the Untitled Linux Show, episode 188, recorded Saturday, february the 1st. We don't talk about Chrome OS. Hey folks, it's Saturday and you know what that means. It's time to get geeky with Linux and open source all kinds of fun stuff. I'm your host, jonathan Bennett, and today we have David Ruggles and Ken. Welcome to both of you, Good to be here.

00:51 - Ken McDonald (Co-host)
Glad I made it.

00:52 - Jonathan Bennett (Host)
Yeah, I'm glad you made it too. It's been the Jonathan and David show, which I'm sure would be fun, but it's better with three. Oh well, there's a show title for you. So we're going to let Ken kick us off. And Ken, you know I'm a KDE fan. I've always been a KDE fan.

01:13 - Ken McDonald (Co-host)
One of my guilty pleasures is this application that's not so guilty for me, but this week we get three views on the latest update to the application. That is Jonathan's guilty pleasure. We're talking about the open source and free partition manager that lets you resize, copy and move partitions without data loss. We're talking about Gparted this time version 1.7. And the one I use the most is GPAR-TED Live. I still got to download and set up my inventory for version 1.7.

01:52
Now we are hearing from Marius Nestor, michael Larabelle and Bobby Borosov. They all talk about the experimental support for BcacheFS and it's being restricted to single device file systems. Gparted 1.7 also adds support for recognizing network block devices, preventing the Gparted probe from starting LVM volume groups and bumping its libparted dependency requirements. It now requires version 3.2 from exFADS. Excuse me, geoparted 1.7 also improves support from the exFAT partitions by reading file system usage from exFATS progs 1.2.3 or later, fixes a hang when searching partitions of butterfs-progs is not installed, updates the continuous integration jobs for Ubuntu 24.04 LTS and Rocky Linux 8, and fixes serial number for USB keys showing binary data If you use GparDead Live.

03:12
Version 1.7 is based on the Debian SID, which is going to be Debian 13 Trixie but we're talking about the repository as of January 31st 2025, and the Linux 6.12 LTS kernel series. Bobby writes that international users will be pleased to learn that translations have been newly introduced or improved for multiple languages, including Belarusian, czech, danish, german, hebrew, hindi, hungarian, georgian, lithuanian, latvian, polish, portuguese, brazilian Portuguese. I guess there is a slight difference. Oh yes, romanian, russian, slovenian, serbian, swedish, ukrainian and simplified Chinese. Since I've only touched on some of the highlights from each article, I do recommend reading each article that I have linked in the show notes to get each author's view of Gparted 1.7.

04:38 - Jonathan Bennett (Host)
Somewhere along the way Gparted just became the easy button for me for doing things with disks, whether it was deleting or moving or fixing. It's just a nice, cozy interface. It's way easier to use than either Partead or FDisk. And it's just, it works, it's solid. I use it quite a bit.

04:58 - David Ruggles (Co-host)
Highly functional but it's got that nice GUI on top of it, so you're not having to dig around for what code matches what in the partition table.

05:10 - Jonathan Bennett (Host)
It also makes it a little bit harder to accidentally delete partitions and ruin your disks, and all of that.

05:17 - Ken McDonald (Co-host)
Great combination is GPAR, did Venturi and whatever. Fedora, of course, of course, that's Fedora, that's the right answer ken fedora, along with whatever distro you want to play with fedora, that's your daily driver yeah, it's also my go-to for a tool for anything and when I want to go play with something on Linux, guess what?

05:47 - Jonathan Bennett (Host)
I do that in Fedora too, sometimes on my daily driver.

05:52 - Ken McDonald (Co-host)
Or that latest distro that you want to review. That can't be Fedora every time.

06:00 - Jonathan Bennett (Host)
I don't review distros. I'm not the distro hopper of the group. That's not me, that's not my stick. Oh, that's Rob. It's Rob. Get out of here with that.

06:12 - Ken McDonald (Co-host)
So how many distros do you have on your Ventway disc?

06:18 - Jonathan Bennett (Host)
Maybe two, I'm not sure. It's around here somewhere, probably on the desk behind me, like Fedora and Rocky is about it.

06:27 - Ken McDonald (Co-host)
Oh, you do have another, just yeah, but it's based on Fedora. Fedora 40 and 41.

06:36 - David Ruggles (Co-host)
I was expecting.

06:37 - Jonathan Bennett (Host)
Once upon a time that was the case, not much, I think. I also have a one of like the all-in-one distros. It's kind of like just to check your memory, check mem test maybe. For a while I was trying to get the Ultimate Boot CD, but that doesn't play very well with Ventoy. So, yeah, it's mostly just Fedora, I think. For a while I had the Windows install ISO on there, but I think I finally ditched that. Didn't use it enough All right, right.

07:12 - David Ruggles (Co-host)
well, do we want to talk?

07:13 - Jonathan Bennett (Host)
about data centers, sure all right some code take it away.

07:15 - David Ruggles (Co-host)
So I've got a link to an article over at the register and its headline is Tiny Linux Kernel Tweak Could Cut Data Center Power Use by 30%, and not bad for 30 lines of code, which is pretty impressive. School of Computer Science at the University of Waterloo started digging into why they could see user-level networking that was getting significantly more performance compared to kernel-level approaches. This was in actually getting data on and off the wire and into applications, and what it boils down to is how you get data in and out of your buffers. Basically, traditionally, linux networking has been interrupt-driven, which means the computer is just sitting there doing whatever it's doing. It's running a program, it's doing anything else, and then, when data shows up from the network, the CPU goes oh no, something has arrived, I must do something. And so then it pauses its operation. It handles that interrupt which is why it's called interrupt, because it interrupted everything and does whatever it needs to do for that and then goes back to its normal operation, does whatever it needs to do for that and then goes back to its normal operation.

08:45
Now, the reason that this was built was because Linux is derived from Unix, and Unix was and is a multi-user operating system. Linux is too. We just don't normally use it that way. So when you're under a multi-user operating system, you wanted to make sure that all your users had equal access to the hardware of the system, and wanted to make sure that all your users had equal access to the hardware of the system and nobody was getting too much or being starved for resources or anything like that. So it made a lot of sense to do it that way.

09:14
However, in modern throughput-oriented workloads, think about a reverse proxy or a caching process. Reverse proxy is where you've got one connection coming in from the internet that you then need to hand off to a bunch of different servers to share your loads behind the scenes. You can wind up having a single workload that is consuming resources equivalent to multiple traditional systems, and for these kind of applications it can be much more efficient for the application to go check the network buffer than for the buffer to interrupt the application and say, hey, do something with me. And so that is what they did. They actually broke it out so that in times of heavy load, the network stops generating interrupts and it allows the worker process to pull the buffer so it completes its work and then when it's caught up, it goes and says, okay, what else is out there? But then on the flip side, when there's nothing out there, the interrupt process gets turned back on, basically so that now, if nothing's happening on the system and it's not going to overload it, incoming traffic triggers it, and so you still get the speed of response based on an interrupt system with the efficiency, under heavy load, of being able to pull a buffer.

10:50
And while that's really cool on the network side, I think that it's also interesting as a programmer to think about re-architecting your own programs anytime. Where you're doing, where you've got a queue or a buffer system, where you've got processes interacting, whether that's multithreading or between applications. Sometimes it makes sense to do it one way and sometimes it makes sense to do it the other way, but just like they did here, where something was working just fine, but they went back and took a look at it and said, hang on, we can do this a lot better. But they went back and took a look at it and said, hang on, we can do this a lot better.

11:26
It's always good, when you're analyzing your own code and handling your own problems, to always look at it and go hmm, you know, just because I'm doing it the most efficient way in this method, like synchronous versus asynchronous, it might actually be more beneficial to do both in different situations. And the bottom line of this article is that, while it may not be obvious just by looking at it, the significance of not interrupting the processor under heavy load has the potential to reduce the thermal load of that processing to the point of actually saving about 30% of the power use in large-scale network operations. So it's really cool, and at least I find it really cool and therefore I'm sharing it with you.

12:21 - Jonathan Bennett (Host)
No, that's interesting. It kind of suggests that there's probably a bunch of other processes where you could do something similar. I'm thinking of the difference in audio processing. Even this is something that the different audio engines like ALSA, pulse Audio and Pipewire they all take kind of a different approach to this, but it's the same question. Do we go out and pull, do we wait for the interrupts, do we push or do we poll on the data? Basically, and you know what. The right answer there kind of depends upon what you're doing, and it's really fascinating that it dynamically switches between the two. That is really pretty fascinating.

13:01 - Ken McDonald (Co-host)
So we need the code equivalent of a push-me-pull-you.

13:09 - David Ruggles (Co-host)
Yeah, something like that Did somebody's UPS just go off.

13:16 - Jonathan Bennett (Host)
Your hearing aids need their batteries changed again. They're whistling at us. No, it stopped. All right, I've got the next story. We're going to talk about GTK. That's the GTK toolkit, the gnome toolkit, the G toolkit I don't know what it stands for these days. It's one of the libraries that sits behind a lot of programs that run on, well, linux, and on Windows for that matter. It is what makes GNOME work. It is also, I believe, what makes programs like the GIMP work and some other things.

13:51
A lot of things are built on top of GDK, and there is a blog post from today, actually February 1st, the day that we're recording this, and you know there's not a whole lot of interesting things in here, and they're talking about how, you know, they've gotten rid of the GDK display, get default call because of problems with incomplete initialization, and blah, blah, blah, blah, blah. And then they say, oh, spring cleaning. By the way, they removed the old GL renderer. They've removed the Broadway back in, which, if you didn't know, broadway is GTK in the browser over WebSockets. Nobody used it. They're getting rid of that. And, oh yeah, they're getting rid of the x11 back in too, as a clear signal that they're going to remove x11 broadway in gtk5 and oh, and you know they go on to some other stuff, incremental improvements and the people that I know that you know are sort of in this space would wait what. And there was much popcorn memes being spread Because GTK has just said that when GTK 5 comes out now we don't know when that's going to be, but they are deprecating the X11 backend in GTK and when GTK 5 comes out it's not going to be there at all. So you don't get to make X11 programs with GTK5 comes out. It's not going to be there at all. So you don't get to make X11 programs with GTK anymore once GTK5 comes out.

15:11
And the take that I have on this is it's almost as if they're saying that the old X11 is no longer being maintained and therefore they're going to move off of it. And the truth of the matter is that the old X11 is no longer being maintained, essentially, and it is time to move off of it. I found it very, very fascinating that this bit of news came from an unexpected source, but, as Rob likes to say, the Waitland future is here and it's time to get on board. Yeah, super, super interesting stuff, at least to me, part of the deal with this is the timing with Red Hat too. I didn't know this, but a lot of the X11 work was being done by Red Hat engineers, and when they rolled which was it Red Hat Enterprise 6 or 7, rolled from like full support to extended support. Essentially, that means that they moved a lot of people off of taking care of X11. And at that point it basically became unmaintained. So it's time. All right, ken, do you want to talk about OpenSUSE?

16:22 - Ken McDonald (Co-host)
I've been waiting to talk about OpenSUSE. Well, this week Bobby Borisov wrote about OpenSUSE's new, still in development, aguma Installer version 11. It brings several new features and improvements. This release also introduces OpenSUSE SlowRoll as a newly supported installation operating system. As a newly supported installation operating system, aguma 11 significantly reorganizes the web interface, making it easier to set up root authentication before getting the all-important install button. Better yet, the button visually alerts you with an exclamation mark if any configuration issues are detected, and it points you to the precise section that needs fixing.

17:13
Bobby also reports Aguma is set to become the official installer for SUSE Linux Enterprise Server 16 and SUSE Linux Enterprise Server for SAP applications, since these require user registration to access software repositories. Aguma now automatically detects if registration is needed and offers a straightforward interface to walk you through the process. Users aiming to install SUSE Linux Enterprise Server must comply with a license agreement. Because of this requirement, aguma 11 introduces the license acceptance step right after you select an enterprise product. The good news OpenSUSE users can simply skip this new feature and keep moving forward without interruption. Aguma 11 extends the capabilities of its command line interface. You can now run the tool on one machine while installing another, thanks to the new dash dash API parameter. This opens the door to scenarios like remote installations or using scripts that automate tasks across multiple systems. Now I've only touched on some of the features that Bobby covered in his article, so follow the link in the show notes if you want to hear about even more features.

18:42 - Jonathan Bennett (Host)
Yeah, so I just got dinged on a Mastodon thread and I'm totally distracted. No, the license agreement, that's what it was. That's interesting. You know. There are some other open source projects that have license agreements and one of the most fun things that I've ever seen is somebody got annoyed by that and went it's open source, downloaded the source, removed the license agreement, compiled it and ran that version Like I'm not agreeing to your license agreement. This is dumb. One could make the argument that a license agreement may actually not be compatible with the GPL. I know a lot of people don't agree with it. Apparently Red Hat does not agree with that take, but I could make the argument.

19:28 - David Ruggles (Co-host)
We don't talk about Red Hat and license agreements. That's old wounds right there. Yeah, yelp.

19:39 - Jonathan Bennett (Host)
You're running OpenSUSE, aren't you Ken?

19:41 - Ken McDonald (Co-host)
Tumbleweed. The license agreement is applied to the SUSE Linux enterprise servers.

19:52 - Jonathan Bennett (Host)
SLESS, sless.

19:56 - David Ruggles (Co-host)
I used SLESS many years ago, migrated from Novell Network to SLESS when Novell purchased SUSE.

20:06 - Jonathan Bennett (Host)
It was definitely an upgrade.

20:08 - David Ruggles (Co-host)
Yeah, that was back in the early aughts purchase.

20:10 - Ken McDonald (Co-host)
SUSE. It was definitely an upgrade. Yeah, that was back in the early aughts. I can think back to when Novell Network was an upgrade from Windows networking.

20:17 - Jonathan Bennett (Host)
Yeah, yeah, linux was a very different place in the early aughts. Mm-hmm. Yeah, I did some work. It's not that long ago, but I did some work on a SCO Unix box and I virtualized it actually, which was interesting in its own challenge. But messing with that box really, really made me thankful for all of the creature comforts that we have in Linux, because that was primitive To the point that in Bash the fact that you can hit the up button to get your previous history did not exist in that shell.

20:58
It was sort of painful to use bash is the born again shell.

21:02 - David Ruggles (Co-host)
Um, I think they had the born show on sco, or maybe it was the one before born, just the basic show could be but it was it was the one before Born, just the basic show Could be.

21:13 - Jonathan Bennett (Host)
It was pretty primitive. You had to recompile the kernel, that's what they called it. It was not quite the same thing as a kernel compile on Linux, but you recompiled or relinked the kernel. Maybe technically is what it was whenever you changed hardware. And that was fun. It was a lot of fun and that was it was fun.

21:33 - David Ruggles (Co-host)
It was a lot of fun. Of course, that was back when Sko was an actual software company before they became just an IP.

21:41 - Jonathan Bennett (Host)
I think when I worked on it they had to become their evil IP self. It was an old install when I got to work on it.

21:54 - Ken McDonald (Co-host)
Anyway, that was back when GTK meant GIMP Toolkit.

21:58 - Jonathan Bennett (Host)
Yeah, I looked into that. Apparently it just means GTK. It doesn't stand for anything. Now, All right. Well, what about Ceph? What does Ceph stand for, David Ah?

22:10 - David Ruggles (Co-host)
Ceph Chromium Embedded Framework. So I've got a link over to a Pharonix article about Ceph and the fact that the Chromium Embedded Framework evidently is still working on its Wayland support, so we're dragging some things into the 21st century still. But so embedded frameworks for web browsers are kind of interesting because Chrome itself already supports Wayland, from what I understand. I don't think there's any issues with that at all and I don't think it's relying on X-Wayland.

22:50 - Jonathan Bennett (Host)
So you can run Chrome under Wayland and it has finally gotten to the point to where it mostly works, I think, out of the box. It still uses X11 by default.

23:00 - David Ruggles (Co-host)
Oh, okay, so it's relying on X Wayland. But the Chromium embedded framework is when you want web browser functionality in another application. Or you've got a situation where maybe you're writing a web site or web portal web application that you also want to run as a native application on computers without having to rewrite your entire code base, so you want to just put a wrapper around it. So that's where an embedded framework like this comes in.

23:35
This is the Chromium specific one, which is using the Chromium backend, and they mentioned specifically that Steam, obs, studio, spotify and many other software packages use Ceph for their in-app browser-like experience. So I'm sure you're using Ceph without even necessarily realizing it, because we all use Steam, right? So, in addition to the fact that they're working on Wayland support, I found it very interesting to find out who is working on Wayland support. It is a software developer for Toyota, and the reason is because Toyota uses Ceph for their in-vehicle infotainment plugins, which they now use Wayland behind, entertainment plugins, which they now use wayland behind. So you may be using wayland when you're driving around in your toyota as long as it's new enough.

24:38 - Jonathan Bennett (Host)
Yeah, that's uh, that's always interesting, the whole car thing. I uh I don't know if I've told the story I owned a mazda miata for 24 hours and in those 24 hours I discovered that it was running linux and found that there were ways to jailbreak it. And so I got, I got all excited about you know, running, running my own scripts on the little linux computer in my, in my little miata sports car. And then the bank called me and said you're self-employed and you don't have enough documentation that you make the amount of money you said you make. So we're, you can have the car you need to go take it back. It was for the best. We ended up spending the money on the house instead, so we bought a house. I bought a house instead of a Miata, so it worked out well in the end. But yeah, it was the same thing. It was a little Linux terminal. Their infotainment center was running Linux.

25:26 - David Ruggles (Co-host)
There was much hacking that could be done there so the house is not as portable, unless you have a really bad tornado the house is not as portable, it's true.

25:34 - Jonathan Bennett (Host)
It's true, but I'd rather live in my house than live in my car indeed and also the old house was only like 1400 square foot, and having three kids and that just wasn't going to be any fun. So we, we got it. We got a house upgrade instead of a car upgrade. All right, let's move on then to KDE, because we're just going to talk about all of the desktop things today, of course, and the news here is that we are basically a week and a half away from KDE 3., and it's all pretty quiet at this point. So we've got a link off to Nate Graham's blog and his main point is that it feels like a good one. It's coming. Now. He's got some things teased for 6.4.

26:23
But we're at the point to where we're doing little improvements to 6.3 to make sure that things work the way that they're supposed to. For example, in the task manager, the previews, where you have playback controls, it's only going to show it for the window that's playing audio, as opposed to all of them, which that's actually. If that starts working, that would be really nice. That is something I'm looking forward to. If you have multiple Chrome windows, trying to figure out which one is the one making noise, sometimes it's a challenge. There's some other things, like the ability to right click and empty trash. That context menu is going to be a little bit better.

27:02
And then over in Plasma 6.4, which that's kind of the next branch. They've got some other things going on there with the keyboard shortcuts and fixing some things on the system settings, region and language page. But the main thing here is just bug fixes, lots and lots of bug fixes for Plasma 6.3. Getting rid of KWIN crashes, fixing things with color profiles uh, lots and lots of fixes here. And, like I said, it's about a week and a half until we're going to see 6.3 and I am looking forward to it. I will probably try to go and update to it just about as soon as I can, which, yes, may mean running rawhide packages on my main desktop. Again, I like to live dangerous. What can I say? I'll take a walk on the wild side.

27:50
When it comes to Linux, the KDE stuff is also interesting. Another story I've been following I think I mentioned it last week is there's an HDR patch floating around in Firefox of all places, and, yes, it looks like Linux HDR is going to come to Firefox before it comes to Chrome. So long as we don't talk about the Google Chrome OS, we don't talk about Chrome OS. But yeah, that is working its way through and we're waiting for another revision of that patch to land from the guy that's running it, who is, interestingly enough, a Red Hat employee. It's a Red Hat employee that's working on the Wayland stuff for Firefox. I was kind of surprised by that, but like that's his job, apparently that's one of his main jobs there at Red Hat is working on Wayland.

28:40 - Ken McDonald (Co-host)
Firefox, because they need HDR support so much.

28:44 - Jonathan Bennett (Host)
I don't know. I mean, it's kind of the natural next thing for Firefox when it comes to Wayland support.

28:53 - Ken McDonald (Co-host)
Got those enterprise users with HDR-capable monitors, I'm sure a lot of them do.

28:59 - Jonathan Bennett (Host)
Yeah, yeah, I don't know for sure what all the use case is there, but that's one of the things he is working on and that's coming along too. He is working on and that's coming along too, but the cool thing is that that landed the part of Wayland that that talks to actually landed in KDE 6.2. So it's already there and so, like in Fedora and all of those places, just as soon as Firefox lands that patch, it's going to fire up and start working in KDE at least.

29:27 - Ken McDonald (Co-host)
Here's going to be a scary thought for you, Jonathan. What's that? When is wayland going to give you a good support for hdr capable webcams?

29:42 - Jonathan Bennett (Host)
uh, that's not really a wayland thing. That's more of a video for linux, either a v4 l2 or that might come first via lib camera. Um, that's kind of the replacement for v4 l2 lib camera with pipewire. I don't know if pipewire supports hdr at all or not. Um, that isn't it. Just that is actually a really interesting thought. Um, there's quite a bit of that stack that I don't know if it's ready. So, like I know, I know there's a way to do HDR over HDMI, but you would have to have, I think, a capture card that was capable of doing that. And then, yeah, I don't know how much of that is wired up Probably quite a bit of work left to do there to be able to do live capture of HDR, I would assume.

30:24 - Ken McDonald (Co-host)
I can tell you for a fact you don't want to see me in HDR.

30:34 - Jonathan Bennett (Host)
It's bad enough that we've got you in a full, definite, full full hd. All right, let's see what is up next. Ah, yes, thunderbird. Thunderbird, which I will get you a quick plug. I don't know who's going to co-host for it, but we've got one of the guys from thunderbird is going to be on floss weekly. I think he's going to be onhost for it, but we've got one of the guys from Thunderbird. He's going to be on Floss Weekly. I think he's going to be on Floss Weekly this Tuesday, but we're not going to release the episode until next Wednesday. We're actually a week ahead there, but Ken wet our appetite for that. What is up with Thunderbird?

31:03 - Ken McDonald (Co-host)
Well, this appetite wetter is coming from Bobby Borisoff and Marius Nestor is coming from. Bobby Borisoff and Marius Nestor. They both wrote about Mozilla's release of its latest development of Thunderbird 134. Now they both start off by warning you Mozilla Thunderbird 134 is a monthly development release that only should be used for testing purposes. Marius states this release brings one new feature a notification system for real-time desktop alerts. Bobby says with this new feature, thunderbird can deliver immediate alerts on your desktop whenever new mail arrives, ensuring that important messages never go unnoticed. Mozilla Thunderbird 134 also contains numerous bug fixes that should make Thunderbird more stable and reliable. It fixes a startup crash when using POP3, fixes a hang when downloading message headers from a news group and fixes APOP authentication for POP3 to prompt the corrected password. Performance improvements include smoother message list updates as well as reduced delays in large virtual folders. According to Marius, mozilla Thunderbird 134 makes it possible to drag embedded images from a message to a composed window. As always, I recommend reading the articles linked in the show notes for more details.

32:42 - Jonathan Bennett (Host)
Yeah, you know, I am actually a Thunderbird user. I'm a bit of a Firefox user these days too, and I was pretty much chrome everywhere. And then we interviewed the firefox guys and they were telling me oh yeah, we extensions work on android now. Like what, something I was waiting for for the longest time. So they kind of converted me and I've got, I think, more tabs open in firefox on my phone right now than I do on chrome. Um, but uh, yeah, I've been using thunderbird for a long time because it's just, it's just the, the. I mean, it's really the webmail or the excuse me, the desktop web app. If you don't want to use your webmail. Whatever your webmail is, thunderbird is pretty much the only other game in town, that's.

33:22 - Ken McDonald (Co-host)
That's really particularly good yeah, do you use it uh to doubt uh, as a pop for pop three, or do you configure everything for imAP? So it's still in the cloud.

33:35 - Jonathan Bennett (Host)
I think IMAP. I would have to check. I think I primarily IMAP, though, but there are reasons that you might want to POP it. You may want to POP and IMAP for that matter.

33:47 - Ken McDonald (Co-host)
If you've got it set up on a server and have that as the pop and then IMAP for all the other devices that you use to check your email with.

33:55 - Jonathan Bennett (Host)
Yep, Yep. Or if you want to have a, if you want to have a device somewhere that is your, your kind of your archive, you could just make sure you've got Thunderbird running there, running as pop and it'll. It'll suck down all of that email and keep it there forever.

34:09 - Ken McDonald (Co-host)
And that that particular one has a filter or filter set up so that it moves it into archive folders.

34:18 - Jonathan Bennett (Host)
You could do that. Yep, it is neat to see Thunderbird continuing to make progress. You know, for a while there Thunderbird was the forgotten stepchild of Mozilla and they have really here in the last I think about the last year they've really stepped forward and and started working on it again. Um, I understand that apparently they're getting some like direct donations from the community directly to the thunderbird project. Um, it actually makes me wonder if there's a there's a potential future where thunderbird spins out from mozilla. Um, we'll see. I don't know. I don't know if there's a potential future where Thunderbird spins out from Mozilla. We'll see. I don't know if that's in the cards or not. It's something I can ask about.

34:56 - Ken McDonald (Co-host)
Does anybody remember the SeaMonkey project?

35:01 - Jonathan Bennett (Host)
I know the term. I tried to look into that a few times and never made much progress with it. What exactly was c monkey? It was a community effort to develop uh, an internet application suite so it was like your browser and your uh, your email and maybe something to edit the web, all at the same time.

35:28 - David Ruggles (Co-host)
Yep.

35:31 - Jonathan Bennett (Host)
Yeah, I never actually used SeaMonkey. I always thought it sounded interesting, but I think by the time that I finally became aware of it it was already sort of dead. So all right, let's see what is up next. Aho, david, you going there? Huh.

35:50 - David Ruggles (Co-host)
I'm not going very far.

35:53 - Jonathan Bennett (Host)
Tell us about the open AI. That isn't.

35:56 - David Ruggles (Co-host)
Yes, I mean we love to not talk about AI, and so I've linked to a TechCrunch article and the headline is Sam Altman open AI has been on the wrong side of history concerning open source, and this is actually a summary of a Reddit AMA with the OpenAI team. So one of the interesting it's the internet, so you can get some very off the wall questions and some crazy out there stuff. But when you do have especially high profile people doing amas, you got people that aren't afraid to ask the hard questions and it's interesting to see what the responses are. So the article goes into a lot of different stuff out of the AMA, and I am neither here nor there on OpenAI, so I don't really want to dive into all of that, but I did think that the section about the open source strategy was interesting.

37:09
Altman, specifically, is quoted as saying not everyone at OpenAI shares his view that they they continue to produce better models. They're going to look at open sourcing some of the previous models as they're no longer making as much money off of them. That was my editorialness, kevin. Well, the OpenAI's chief product officer also replied to that question thread and said that they're considering open sourcing older models that aren't state-of-the-art anymore and he said they're definitely going to think about doing more of that. But neither Sam Altman nor Kevin Weill went into greater detail on that. So I mean, it's a company that's got open in the name, but that seems to be about as close to open source as they're willing to get oh, get um oh. But it is interesting to see what uh other activity, what you know, the market pressures and things like um deep seek and china's ai development and stuff. Just there there's a lot of uh tension and it's interesting to see how, how openness either does or does not play into all of these relationships out there.

38:52 - Jonathan Bennett (Host)
Yeah, so the folks at OpenAI are special people. You know that's a nonprofit company right In by their government filing. At least that's a nonprofit.

39:08 - David Ruggles (Co-host)
Well, part of it is it is a nonprofit by their government, fil, at least that's a non-profit, well, part of it is, it is a non-profit.

39:12 - Jonathan Bennett (Host)
by their government filings, they are attempting to take it from a non-profit to a for-profit business, which is insane to me, because that is a complete no-no. Um, I, I just I cannot comprehend how any lawyer signed off on them ever attempting to do that. That is not a thing that you do. That's a thing when people actually do that. That's the thing you go to jail for, right, like that's a anyway.

39:41 - Ken McDonald (Co-host)
Treasury Department standing right there waiting. Yeah, I guess.

39:45 - Jonathan Bennett (Host)
Well, it's not necessarily the Treasury Department, it's going to be like the IRS and the Federal Trade Commission, places like that, because that's just not the way business law and nonprofit law works in the United States had back in I think it was September. They made the news because several people left on their leadership board and come to find out it was like the, the I hate to say the good guys, because it's not. It's not a black and white issue, but it was the people that were saying, hey, let's be a little bit more careful about this thing.

40:24 - David Ruggles (Co-host)
Those are the ones that left um, or got pushed out, or got pushed out right, right, right, which?

40:30 - Jonathan Bennett (Host)
which version of the story you read? Yeah, uh, they've got open right in the name of open ai and they are not at all one of the actual open source ai providers. Um, you know, they're talking about spending hundreds of billions of dollars and the chinese deep seat comes out and like, oh yeah, we did this on a shoestring five million dollars and we got this done. Now, to be fair in the comparison, the thing that they built is equivalent to the thing that OpenAI built for $10 to $20 million. Right, so it's not quite as out of balance as it seems, but it is still hilarious to me. It is still hilarious to me that DeepSeek actually uses the MIT license and is closer to an open source AI than what the nonprofit called OpenAI is willing to push out Now, according to OSI.

41:24
I asked Phipps about this because OSI has their definition now of what an open source AI actually is, because it's kind of complicated and the thing that is missing from what DeepSeek pushed out is that they do not have enough data, they don't have enough information out there about how they trained the thing, and so that's actually a part of what OSI considers to be an open source AI is, you have to also either give people the exact training data or give people the details on what that training data was, because in some cases of AI, your training data is going to be things that you just like, maybe even by law, you can't share publicly, right? So like what if your training data has medical information in it? You train your AI because you want to do a medical AI. You train your AI on medical information. Well, that stuff is protected to the moon and back. You cannot share that.

42:16 - Ken McDonald (Co-host)
Even if you anonymize it.

42:18 - Jonathan Bennett (Host)
Well, I mean, you could make the argument that anonymizing it is training the AI, right, but the actual data itself, there's just no way to share it legally. So there's kind of an allowance for that in OSI's definition of an open AI. That is, if you don't give the world the actual data, essentially you have to give them enough of the recipe that you could replicate that data. You could do the same survey over medical data and put the same equivalent stuff into it. Um, and yeah, except for that uh, deep seek is is open. Right, it's almost what osi considered an open source ai and it's literally out under the mit license. So it's. It's just hilarious to me.

43:08 - David Ruggles (Co-host)
It's quite fun oh, we do live in interesting times.

43:12 - Jonathan Bennett (Host)
Oh, yes, very much so. I guess I'm glad to see Sam Altman come out and say that he realizes that they're on the wrong side of that, but I'm not exactly holding my breath that they're going to actually make meaningful changes. Yeah, they've not necessarily proved themselves to be the best players in that particular space. All right, I've got a story about the colonel's problem. That may or may not be a problem, and this is something that colonel people have been talking about for a while, and that is maintainers and programmers too, but maintainers dang it. Maintainers and programmers too, but maintainers dang it. Um, they.

43:55
We have had a couple of colonel maintainers step down recently, and we have also, for geopolitical reasons, had colonel maintainers to get dismissed as being maintainers. Um, and we've talked, we've talked about that where the, the, the people that the people that were their email addresses were at Russian corporations that were on the United States Treasury's list of, do not work with these companies, and so that was fairly straightforward in actuality. That one was unfortunate that it was not spelled out why they were being removed as maintainers, but what was actually done was pretty straightforward, and so we have an article here that a DRM developer driver developer had to orphan his drivers due to health reasons. Atheros engineer for the past decade and was the apparently the sole maintainer for the Linux wireless uh drivers, is now also stepping down as maintainer. Um, he is, uh, let's see. Jeff, another kernel developer, is going to continue to maintain a handful of drivers from Atheros, but the driver slash net slash wireless tree, the whole thing, does not have a maintainer and, as Callie says, there is no replacement at the moment. If anyone is interested, please let us know.

45:27
But this in and of itself is not great. I'm sure someone I hope someone will step forward to do this, because having working Wi-Fi drivers is sort of important and there's a lot of companies out there that are pretty invested in having working Wi-Fi drivers drivers. So I imagine there are people at Google that are on the on the staff, right Paid programmers at Google that would do kernel work that could step in and do this. Um, some of the other places that that actually you know, do a good job paying their programmers to have people work on the kernel. So it's not this in particular that I'm concerned about that. Well, that that the um, the Linux leadership team, is concerned about.

46:06
But I know that they do have this concern that the maintainership pool, the people doing maintainer work in Linux, are getting older and grayer and maybe crotchetier. But that's not exactly what we're talking about today. They're going to age out eventually. Just the reality of the situation. People will get old and have to stop doing it because of health reasons and other things.

46:33
And the question is are there enough young people that are coming up and hacking on the kernel and doing it in such a way that they might become maintainers to continue the work? And this is something that the people at the kernel are particularly interested in pushing for. And you know they've got some official projects where they're trying to get more people into the kernel and you know sort of grooming them to be maintainers. And then you know, unofficially there's things like it only takes you three or four patches to the kernel to land a job offer. So if that's the sort of thing that you're interested in, start sending patches in and someone will probably send you an email with hey, would you like to do this for us?

47:18
So there are definitely on-ramps, there are definitely paths to get in there, but there's just this question like the kernel is so, it's so big, but it's also so important Do we have? Do we have that group of people going forwards? And I think so. I think it's going to be a problem that takes care of itself, but I don't know that for sure. And so it's going to be interesting to watch over the next decade or two probably is when we'll watch this play out what sort of the average age of kernel maintainers tends to be. If that's going to continue going up, we might have a problem.

47:53 - Ken McDonald (Co-host)
So when's the mentor program going to start?

47:55 - Jonathan Bennett (Host)
It's already started. There are several of those mentor programs out there.

47:59 - David Ruggles (Co-host)
Yeah, I think the challenge is less from like mentor programs and stuff and more from interest.

48:08 - Jonathan Bennett (Host)
That's just not as much interest as would be ideal yeah, you don't get recognized so much for maintaining as you do for developing new uh code for uh or expanding new uh horizons well, mean so the maintainer started out pretty much all of them started out by writing code for the kernel and you, eventually you write enough code for a particular part of the kernel that you sort of you know. You get the email like, hey, why don't you, why don't we make you officially responsible for this, this code that you've written? But yeah, it's just a question of are there enough people coming on board? But yeah, it's just a question of are there enough people coming on board, new people coming into the project to continue to have a new life and sort of reseed that maintainership pool, that leadership pool. And there are people at the Linux Foundation, at the Linux Kernel, that are thinking about this a lot and working on how to take care of it and make sure that it continues.

49:18 - David Ruggles (Co-host)
It's kind of related to what you started with there, but I actually just dropped a link in the rabbit hole about a story that I had come across, not really looked into a whole lot, but of course the whole sanction challenge and who's allowed to contribute to who and stuff. The Linux Foundation has actually released a comprehensive guide, evidently within the last week or so, to help navigate all of that mess, but that's something you're having to keep up with as well.

49:41 - Jonathan Bennett (Host)
Yeah, fine, all right, let's move into some command line tips then. Is that the gong at the end of the news segment unintentionally unintentional gong?

49:57 - Ken McDonald (Co-host)
all right. Maybe we need to make it intentional uh, maybe I don't know.

50:03 - Jonathan Bennett (Host)
I don't know that we need any more bits on this show. Let's talk about Pipewire.

50:11 - Ken McDonald (Co-host)
Yes, let's do, I'm going to be going into. The command that I'm going to be covering this week is PW-V4L2. It may sound familiar because last week I'd covered V4L2-CTL, but for those of y'all listening, I've just brought up a screenshot that I took of running PW-V4L2 space dash H. Basically it runs a command using a compatibility layer that maps pipe wire video devices to be visible. Applications using video for Linux.

50:58 - Jonathan Bennett (Host)
You'll find a link to the screenshots I took in the show notes. You're beeping again, man. Oh, there he goes. We need to get you a buy me a coffee fund so that we can replace your battery backup. It's stopped. Now You're good.

51:15 - Ken McDonald (Co-host)
Okay, the first screenshot I have here shows what the options are for PW-V4L2. It's only 3-H for getting this brief help screen. Dash r if you need to do a, get into a remote, you put the name. Then dash v for verbose debug information. Now the second screenshot is displaying the command that I typed in to run. This particular one is where I use a PWV4L2 with the dash V so I can debug the output before running V4l2-ctl. And then the rest of it gives the device that I want to adjust the brightness on. I'm setting it down to zero.

52:17
And then the third screenshot is what happens after it did. It gave a lot of output that I personally didn't really understand all of it. I could recognize some of what was going on, but it basically did what I asked it to do. It adjusted the brightness down to zero and you can see the result of that, because the screenshots got the terminal at the bottom and where I was using video ninja for capturing the output of my video camera. And then with the, the fourth one, the fourth and fifth screenshots. The fourth one showing the command that are typed in for adjusting the brightness to 255. The fifth screenshot that I've got displayed right now is showing that the brightness has been adjusted to 255, again with the debug information, so you can look and see what's going on.

53:48
And then we've got the sixth screenshot, which is where I tried running OBS from within PW-V4L2. And here's what came up afterwards and it launched OBS fine. It gave me some screen sharing options where I was able to select whether which screen to share with a console, which screen to share with a console. And then when I tried to start the virtual camera in OBS, I got an error message saying starting the output failed. Please check the log for details. Note if you are using the NVENC or AMD encoders, make sure your video drivers are up to date. And your final screenshot is showing that window with that message so you can read it easier. But PWMv4L2 does give you the option of controlling your cameras and any other Pipewire video devices from within Pipewire.

55:18 - Jonathan Bennett (Host)
Now I am trying to run that command on my laptop here and I am in the weirdest little issue. I am getting a command pw-v4l2 not found but can be installed with the pipewire-bin package. Well, I have the pipewire-bin package installed In fact I did a reinstall of that package and it still can't find that application. So I will have to play around with that a little bit. Something about the Pipewire bin package in Pop OS here apparently doesn't include it.

55:52 - David Ruggles (Co-host)
Pop OS.

55:54 - Ken McDonald (Co-host)
Well, there's your problem.

55:56 - Jonathan Bennett (Host)
I'm sure it's on the Fedora desktop behind me.

56:00 - Ken McDonald (Co-host)
This is Ubuntu 24.10.

56:08 - Jonathan Bennett (Host)
Yeah, it works just fine on the desktop behind me.

56:16 - Ken McDonald (Co-host)
I did find that I could get the virtual camera to launch in OBS by making one change. By making one change pw-v4l2 space, sudo space, obs, aha. But before you do that, you want to put do a sudo dash v and enter your password so that you and then run that command because it doesn't give you. Let you actually, because it's in the middle of the.

56:51 - Jonathan Bennett (Host)
Yeah, because it's in the middle of the command stream Makes sense. All right, David, you've got a command here that is near and dear to my heart as a fellow web dev guy. Occasionally I'm not really a web dev guy, but I occasionally wear that hat.

57:07 - David Ruggles (Co-host)
Yes, so we haven't talked about it before, but there is a command line program called CertBot and this is the website for CertBot. It's not really something that I can demonstrate very easily, but it's something I've been working with quite extensively the last couple of weeks, so that's why it was top of mind. But what it allows you to do is take advantage of what's known as the ACME protocol and get automated certificates to secure web communication, and why I'm saying web communication? If you look at the website here and most of the time when people talk about this you're talking about websites. But the certificates are for encrypted communication across the web TLS and can be used for anything that needs a TLS certificate. So you can use them for mail servers if you're trying to encrypt your mail traffic, and the big advantage of this is using it in concert with a service provided by a company called let's Encrypt, which provides free TLS certificates. They're relatively short-lived I think they're 90 days Actually. The interesting thing is Steve Gibson's been talking about that because there was a discussion about them going to six days. So if you don't watch security now, he's been talking about that quite a bit over the last couple of weeks.

58:45
But it's a command line utility and you can come in here, select what you're using. I always use Nginx, what you're running on, and you've got a choice of Snap or Pip for Linux and I do Pip because that is the Python installer and then you get the option of which one you're doing and you just simply walk through the steps and tells you what to install, how to configure it and you run it and as long as you've got a functioning web server already, you can have a cert working pretty painlessly and then you can set up automated renewals of that and everything, and it's just a process to move forward. Acme is not specific to let's Encrypt. Acme is a standard for automated certificate requests and renewals and there are other providers out there that support acme, but let's encrypt is probably the most well-known one of the options out there. So anyway, if it's something that you're doing on the server side, just remember that certbot is available to encrypt all your web traffic. Do you remember how?

01:00:04 - Jonathan Bennett (Host)
hard. It was or hard or expensive to get a certificate HTTPS certificate before let's Encrypt, oh.

01:00:13 - David Ruggles (Co-host)
There's a slide buzzing around me. It wasn't super hard, but it was pretty darn expensive.

01:00:23 - Jonathan Bennett (Host)
There was like one other provider that would give you free certificates. It was a little bit sketchy and their website looked like it was out of the 90s. I don't remember the name of it I don't either.

01:00:33 - David Ruggles (Co-host)
I never used them Before let's Encrypt. I was doing paid ones through the various providers. There were a couple of different ones I used. Network Solutions was the original and they're like the gold standard in the fact that they want your gold to give you a certificate. So, I never used them, but it was always paid certificates.

01:00:58 - Jonathan Bennett (Host)
Yeah, and then for a while you know, the extended validation certificates were a thing. It's like, yeah, you can pay us $5,000 and we'll make the browser bar extra pretty in the web browser.

01:01:09 - David Ruggles (Co-host)
Yep, and the interesting thing about that is you can still get those extended validation certificates, but when the web browser quit making a big deal out of it, everybody was like well, it doesn't look any different, so why pay the money? Yep, Yep All right.

01:01:23 - Jonathan Bennett (Host)
So I've the money. Yup, yup, all right. So I've got a command line tip, and this is because I have been uh, I've been working with getting rock M working on Fedora. You should install Fedora, um, and so my command line tip is rock M info. It's installable on Fedora, it's installable on pop OS and if you run it it'll give you a bunch of information. But really, what you're looking for is it'll have. It has agents. So it'll start with agent one and that's going to be your processor. And so this is actually a useful command just to quickly list what the processor and GPU is on a system, because it'll tell you exactly what your processor is.

01:02:09
So in this case, my laptop here is a Ryzen 7 5800 or no. That must be on. That's on the desktop behind me. That's on the desktop behind me. That's why that doesn't make any sense. I'll get back to that in a second. You may have to run it with sudo or you may have to add yourself to the render group on the computer that you're working on. So, on the laptop, this one makes more sense.

01:02:36
Yes, so on this laptop, it is a Ryzen 7 Pro 5850U with Radeon graphics. That's agent one. And then I also have an agent two, which is Radeon graphics. That's agent one. And then I also have an agent two, which is Radeon graphics, a GFX 90C, and that is the little embedded GPU that's in the laptop CPU and then on the desktop behind me through the power of SSH I can run it there, and I see that that has a Ryzen 7 5800X eight-core processor as agent one, and agent two is the Radeon RX 6700 XT, which shows up as a GFX 1031. Now what that means is I have enough of the Rock M.

01:03:21
Rock M is AMD's fancy. Let's do compute on your GPU and your CPU, for that matter, but mainly GPU. Rockm is their kind of answer to CUDA. It is actually open source, which is nice. It's part of HIP or HIP is part of RockM, I forget which way that goes. But anyway, if you get both agents on RockM info, that means you've got enough of the RockM system installed on your machine that it is aware of your GPU, and so this is a good first step to being able to actually do compute on your GPU, if that's a thing that you want to do and of course, there's been some very interesting things in the news recently about why you might be able to and might be interested in doing compute on your GPU.

01:04:10
We talked about DeepSeek. I've not gotten DeepSeek to work yet. I've gotten some other things to work, but when you go to run DeepSeek it's really interesting. You will get different answers doing it locally versus doing it through the DeepSe service, because deep seek is a company out of china and so the chinese communist party. There are certain things that they are told you get to censor this you don't get to tell people about. You know tiananmen square, why? It's why winnie the pooh is hilarious in regards to china, like it is a huge no-no. They absolutely censor that stuff. Um, and you know some other things like that. And so there is there is a legitimate reason why you might want to try to run one of these llms locally, because you can get the uncensored answers out of it which you know in some cases, something you might want to do. So anyway, uh, we are going to uh.

01:05:02
Over the next couple of weeks I think we're going to go down that the Rock M rabbit hole just a little bit. And it actually works. In Fedora it took a bit of fiddling and playing with Conda and Pip 3 and juggling Python prerequisites and all of that, but I finally have it working and doing some interesting things. So that will be the command line tips that I cover for a few weeks here. So it's a show, anything you guys want to plug. Uh, want to get the last word in before we let folks go. David, did you have something?

01:05:39 - David Ruggles (Co-host)
Oh, uh, oh no, I didn't know about Winnie the Pooh in China until you mentioned it. So that's the rabbit hole I'm going down oh yes, that was fun.

01:05:50 - Jonathan Bennett (Host)
Just you know, you can google it. I'm sorry. I totally got this show censored in china.

01:05:55 - David Ruggles (Co-host)
If anybody listens there, I'm so sorry yeah, yeah, I googled it, so anyway, that's neither here nor there. I was just like what when you mentioned that.

01:06:06 - Jonathan Bennett (Host)
Yes, all right, ken, you've got something in your ending notes. What is your link about? Ken is muted. Tsk, tsk.

01:06:23 - Ken McDonald (Co-host)
But, that also means we're not hearing the UPS. It's true, are you hearing the UPS? No good, because I can't see the light from here to see if it's uh gone red or not the package already arrived but the uh.

01:06:40
I've got a link in the show notes to article michael larlle wrote about the cloud provider Equinox ending support of xorg foundation or freedesktoporg's cloud hosting needs, and this will be effective April 30th 2025. It's a good read and if anybody's got any suggestions on what they can do to help with finding a new cloud host or raising money for a new cloud horse, maybe put in some suggestions yeah, I, I have thoughts banning the lead developer of one of the more popular alternate desktop environments.

01:07:25 - Jonathan Bennett (Host)
maybe it wasn't a great move. Free Desktop is the home of Wayland and all of that, so it is fairly important. At the same time, if Free Desktop itself goes down, I'm sure that Valve will be happy to step in and maintain Wayland going forwards Portions of it. Yeah. So yeah, good to know.

01:07:52 - Ken McDonald (Co-host)
How many projects does the XOrderFreeDesktoporg Foundation sponsor or maintain?

01:08:01 - Jonathan Bennett (Host)
I'm not sure. I know there's several. There's two primary ones right now and that's Wayland and Ex-Wayland. Those are the two big ones, but I think there are a few others. They do some of the other video stack stuff some of the graphics drivers.

01:08:16
I don't think LibCamera is part of there. It might be. I could probably find out LibCameraorg. No, no, not that I see. I mean it's possible, but it doesn't look like it, not with a, not with a quick google. Looks like it is sort of its own thing. Um, yeah, looks like it's, uh, it's its own thing. Not part of uh, not part of x. Uh, all right, well, we will, uh, we'll follow that story and if they find somebody else to sort of be their ci sponsor going forwards, we'll definitely let folks know about it. Uh, appreciate both david and ken being here was a lot of fun. I'm sure we will have some of the regular crew back next week. Although you guys are definitely welcome, I want to can his can. His regular crew. David is welcome anytime he wants to jump in, of course, uh, but I appreciate you guys being here good to be here yeah, all right.

01:09:15
Uh, I do want to quickly plug over at hackaday. We've got floss weekly still and that is, uh, we had it's been kind of an interesting week because we have a couple of interesting weeks coming up. I did an interview with Key Jeffries from Session this past Thursday. I'm hanging on to that until next week, and then next week we're doing an interview with one of the guys from Thunderbird and I'm hanging on to that until the week after next, because in two or three weeks I'm actually going to the Zero Trust Conference in Orlando, florida, and that's going to take most of that week there around the 21st of February and I'm actually going as a Twit host. So that'll be fun getting kind of the VIP treatment as Zero Trust World is or as Zero Trust is one of the sponsors of Twit these days. So very much looking forward to that.

01:10:10
The only other thing I want to plug is, of course, I've got the security column that goes live also on Hackaday every Friday morning and we have a lot of fun with that. If you want to keep track, stay on top of what's going on in the security world, you can do that. You do that there, follow me there, appreciate everybody watching and listening. Those will get us live and on the download and we will see you next week on the Untitled Linux Show.

01:10:34 - Leo Laporte (Announcement)
Hey buddy are you a geek? Are you a tech enthusiast? Then I would love to invite you to join a tech community like no other. You can gain exclusive access to our incomparable quality tech content for just $7 a month with Club Twit. As a member, you'll enjoy all Twit TV shows ad-free plus access, private video feeds for insider shows like iOS Today, home Theater Geeks and so much more. Dive into the members-only Twit Plus bonus feed for behind-the-scenes content, club discussions and special events. But here's the best perk Join our incredible Discord community to watch live show productions, chat with hosts and participate in exclusive members-only activities. It's your backstage pass to the world of Twit. Whether you're a tech enthusiast or a lifelong learner, club Twit elevates your knowledge while entertaining your interests. Get two weeks free when you sign up now for $7 a month and unlock unparalleled access at twittv slash club twit. That's twittv slash club twit and, from the bottom of my heart, thank you and welcome to the club.