Untitled Linux Show 187 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, this week we're back and we're talking about Wine 10 and x86 simulation on ARM. There's a bunch of Linux 6.14 news and there's the NVIDIA RTX 5090 reviews on Linux. It's a lot of fun. You don't want to miss it, so stay tuned. Podcasts you love From people you trust. This is Twit. This is the Untitled linux show, episode 187, recorded saturday, january the 25th. Don't fear the co-pilot. Hey folks, it's saturday. It's time to get geeky over linux and open source all kinds of fun stuff. Uh, I'm your host, jonathan b Bennett, and it is time for the Untitled Linux Show. And it is not just me, of course. We've got a trio of co-hosts. We've got Rob and Ken and Jeff, and they are here ready to go, prepared for action, locked and loaded with some stories, and we are going to let Rob go first.

01:02 - Ken McDonald (Co-host)
Is Rob going to be whining?

01:07 - Jonathan Bennett (Host)
Well, see, I was going to say something about he's gonna get his wine uncle on um, but you've got a story, rob, about wine and no, not fermented grape juice, but the wine is not an emulator project. And, uh, we had a milestone, we had a big milestone we just hit yeah, well, even though it does do some emulation these days also.

01:25 - Rob Campbell (Co-host)
But so this week a new release of Wine has come out and, yes, they have hit a major milestone. And does anyone know what that major milestone is? 10.0? Well, it is released version 10.0, but that isn't a milestone, that is just an arbitrary number. It really doesn't mean anything other than, let's say, it's 10 netto. Now the big milestone is wine has been released with native whalen support, using whalen by default, uh, but it is able to fall back to X11 if required. But who needs X11 these days? That's the big milestone. Wayland Default in Wine.

02:12
Wine 10.0 features an improved Wayland driver with OpenGL support, proper positioning of pop-up windows and auto-repeat key support. Also, on the video side of things, it has improved hdmi support for scaling on modern displays all things that are going to make wine even better for your modern, uh, modern desktop displays. Other features include an initial bluetooth driver, uh, improved arm support implementing x86 emulation interface in its arm builds. So even though Wine is not an emulator, there actually is some emulation going on these days. The devs say it quote takes advantage of the ARM64 EC support to run all the wine codes as native, with only the applications x86-64 code requiring emulation. It also has full Dvorak keyboard support, direct 3D support, enhancements, ff MPEG based back-end alternative to G streamer. It's an opt-in experimental feature back an alternative to G streamer. It's an opt-in experimental feature and for us gamers this, this just means more gaming goodness as these features work their way into proton in the future. So big milestone there, and 10 is just a number.

03:37 - Jonathan Bennett (Host)
And it's just a number. So a couple of questions. First off, you said HDMI support.

03:43 - Ken McDonald (Co-host)
I don't think wine has HDMI support support I said high dpi support uh-huh, I'm sure sounded like it like hdmi, to me too all right, so high dpi.

03:54 - Jonathan Bennett (Host)
Improved high dpi support for scaling on modern displays all right, now that we have that one settled um, the other thing is the arm 64 support. What exactly is the the scenario where someone would take advantage of this?

04:10 - Rob Campbell (Co-host)
I believe this is running Wine on ARM.

04:16 - Jonathan Bennett (Host)
So you've got a Windows binary that's on x86, and you want to be able to run it on an actual ARM64 piece of hardware.

04:24 - Rob Campbell (Co-host)
Yeah, I believe that's what that is.

04:26 - Ken McDonald (Co-host)
Well, looking at the release notes, the ARM64 EC architecture is fully supported with feature parity with the ARM64 support, and the 64-bit x86 emulation interface is implemented. This basically takes advantage of the ARM64 EC support to run all the wine code as native, with only the application's x86-64 code requiring emulation. 64 code requiring emulation.

05:04 - Jonathan Bennett (Host)
So I am wondering who is pushing for this? Who would be interested in running Windows binaries probably games, but just in general Windows binaries on ARM64 targets? Are there some tea leaves to read here?

05:24 - Rob Campbell (Co-host)
And you're already. I know you're thinking like Valve type of things.

05:28 - Ken McDonald (Co-host)
I'm going to say Crawl Shredder.

05:29 - Jonathan Bennett (Host)
So anybody so what this is for is.

05:35
This is where you have a piece of ARM64 hardware, whether that's one of these new ARM laptops or whether this is a handheld gaming rig that runs on ARM instead of x86, and if this would let you much more capably run Windows games on it. So somebody, somebody cares about this, some it's one. It's one of two things either one of the wine developers has one of these ARM64 laptops and they're really interested in, or or maybe it's for the Raspberry Pi, for that matter. I mean, maybe one of the one of the wine developers is a really big retro game enthusiast and wants to be able to play some older games on the Raspberry Pi and you emulating, you know, x 76 windows on the arm Raspberry Pi or on arm laptops or on ARM handhelds doesn't necessarily work very well. So someone has spent, I'm assuming, quite a bit of effort to make this work now, and it sounds like it's actually starting to work fairly well. That's just fascinating to me. It makes me wonder what the exact scenario is that led to this.

06:44 - Ken McDonald (Co-host)
Now would you need 16 gigabytes of memory for doing this.

06:51 - Jonathan Bennett (Host)
It depends on what game you're trying to run, what Windows game.

06:55 - Rob Campbell (Co-host)
Yeah, if you're running a Windows game from 2002, probably not.

06:59 - Jonathan Bennett (Host)
Yeah, probably not. Probably not, Although, ken, are you talking about video card memory?

07:08 - Ken McDonald (Co-host)
Actually no, Just memory in general.

07:10 - Jonathan Bennett (Host)
I thought he was setting Jeff up because Jeff has a story about the.

07:14 - Ken McDonald (Co-host)
Though there could be an option for video card memory with arms.

07:18 - Jeff Massie (Co-host)
Yeah, I thought there was a segue there.

07:21 - Jonathan Bennett (Host)
Yeah, let's kick it over to Jeff and learn about the newest new thing in the insane GPU race that's going on right now.

07:30 - Jeff Massie (Co-host)
Yeah, so we're going from handhelds to liftable gaming machines with proper form. And you know, don't use your back, use your legs, be careful, get a good grip. But for those that don't keep up with graphics hardware, nvidia announced the launch of their rtx 5090 graphics card. It was announced at ces and the embargo officially lifted a couple days ago. Right now social media sites are flooded with gaming performance reviews, comparing the 5090 mostly to the previous generation 4090, because it's kind of the direct comparison. Previous generation 4090, because it's kind of the direct comparison. Nvidia themselves even said that the 5090 card is geared more towards the prosumer or people doing computational workloads, and that's the reason they gave it 32 gigabytes of ram and ai workloads I throw in under computational. For gamers they're saying it will be the 5080, which is the dedicated gaming card. But for gaming the 5090 is still going to be faster, it's just more expensive. It's the top of the line flagship product and flagships are not value purchases. Now you know we're bringing this up because there's a Linux spin on this and Michael Larable over at Phronix has a 5090 that was sent to him and he did some testing. Now this is going to be a little different because there actually isn't a released video driver, so there's actually an early release video driver, meaning beta. It's 5070.86.10 driver to be exact. The driver does work with the 5090, where earlier drivers don't. The beta driver has the latest CUDA 12.8 package and NVIDIA says you shouldn't use that driver for graphics and gaming workloads. They recommend waiting until the official driver is released at the end of the month. So today's benchmarking is all going to be CUDA and computational performance, without any graphics or gaming performance. The benchmarks of the older cards were rerun with the latest 565.77 driver and the 5090 used the beta driver. Michael includes cards from the 2000, 3000, and 4000 series and from the low end of each of those series to the high end. The operating system he used was Ubuntu 24.10 with the 6.11 kernel. And keep in mind these results are with a beta driver. But because it's CUDA which has got a small revision, it should be pretty comparable to the release driver, since it's a stable part of the driver stack. You know CUDA has been around for many, many years. I'm sure Michael will check it out just to verify if there are any changes from the beta driver to the release driver. But we're not really expecting anything to shift Now, if we take a look at the results of the over 60 benchmarks which were ran, not only did they consist of CUDA, but also OptiX, opencl and Vulkan compute benchmarks, with the geometric mean of all those showing that the 5090 is around 30% faster than the 4090, which, when looking at the Windows gaming benchmarks, it's about the same percentage.

10:49
I can't say anything about Windows compute benchmarks because I've not found any compute benchmarks run by the normal tech sites. You know Gamers, nexus Hardware, unboxed. You know Level 1 Techs, you know all the big ones. Now, level one techs talked about compute performance, but they're gonna do it on Linux and they're waiting for the official drivers. So nothing was done there. So I only personally know of Michael's compute benchmarks. There could be others, but I haven't seen them.

11:23
Now, 30% sounds like a decent improvement, but the downside is the 5090 costs about 20% more or more for this generation of card. It's $2,000 for the Founders Edition and will be more for the Partners cards. Now, the Partner cards, those will be the Asus, msi, gigabyte, those cards. So based on their custom coolers and all that, they are going to cost more. But I have not seen any pricing yet. They should be out. The pricing list should be out soon and then we'll be able to see how bad it's going to be.

12:03
The other downside is power, and Michael recorded a peak of 584 watts. Now when you look at the performance per watt, it's the high end of the charts for being inefficient. It's not the worst, but it's in the running for the worst. One bright side is, while it takes all that power, the actual core temperature was in the middle of the pack. So the new cooling system that you can see described in a Gamer Nexus video they go into great detail on it it actually does a pretty good job of eliminating that much heat.

12:37
Now I'm going to give my personal opinion based on all the reviews I've seen on Michael's and all the other stuff, because overall Michael doesn't say that much, whether he recommends the card or not.

12:49
He just presents the facts and lets the readers decide.

12:51
For you know he's pretty neutral when he released all the data.

12:58
You know, personally I'm going to follow along with Hardware Unboxed, where they said if the 5090 was released as the 4090 Ti, nobody would bat an eye Because of the performance uplift, because of the price it kind of is what you would expect from a 4090 Ti, but because it's a new generation, people expected more, and the reviewers in general gave a pretty lackluster view of the card.

13:24
Sure, it's faster, but it's taking a lot more power and money, and so the overall cumulative end result is meh. You know, take a look at the article in show notes so you can see all the different benchmarks and you can decide for yourself whether this is an upgrade you want to pursue or you might let this generation pass you by, though I say that with keeping in mind that most people will not get a 5090 and it will be more interested in the mid-range and lower range cards. Over the next few months we should see the benchmarks on those as well, and when they come out, I'll be sure to let everyone know what the results are and keep everybody up to date.

14:01 - Jonathan Bennett (Host)
Yeah, so it is good enough that NVIDIA has yet another money printer, right? For where the market is for the group of people that actually legit need this sort of hardware, it's going to be an instant buy. This is yet another money printer for NVIDIA.

14:23 - Ken McDonald (Co-host)
And maybe even a good space heater.

14:28 - Jonathan Bennett (Host)
And a space heater, sure, and I'm well not thinking, and, jeff, when I say that I'm not thinking about gamers, I'm thinking about people that want compute, and particularly those that are working on, like ai and lm stuff. They're gonna go oh, it's 30 faster. Take my money here.

14:42 - Jeff Massie (Co-host)
I need as many of those as you can give me well, and, and, like I said, that's why it comes with 32 gigs of ram in it. Gddr, I forget. I think it's, I don't remember it's six or seven. Yeah, what do we have to do this day, is it? Yeah, it's, it's a lot faster than uh, the 4090 memory, uh, pathways wise, but yeah, it's, it's going to be for those people, because when you start looking at the professional cards, oh my gosh, the price on those Two grand and you can do your work, versus getting some of the high-end cards. Oh, that's okay, I'll buy 12 and still save money. You know, yes, it's yep, and the power they take. You're not running off of regular like the us 120 volt you're, you're into the 240s or you're switching over to like three phase uh power distribution units, or pdus, to be able to power these uh racks that they put them in yeah, you're hooking it up to a dedicated three-phase power supply coming off with 240 volts in.

15:53 - Jonathan Bennett (Host)
Yeah, so somebody will. Inevitably there will be server cases, full 48U server cases that are just servers with these things racked in them. Somebody's going to build a server farm out of those and it'll power the next big thing in AI, probably. Yeah, this is yet another money printer for NVIDIA. No, question.

16:15 - Jeff Massie (Co-host)
And depending on how many you stick in that server rack, you could be running, say, 480 volts, three phase, 50 amp. Maybe you have multiple of those circuits so it and then cooling is a whole nother. You know how do you keep all that cool and whatnot? Just kind of note. Here is some of the. I think the performance uplift too, being a little lackluster, was also. People have kind of been blaming it on that. The 4090 was built on TSMC's 8 nanometer node and so is the 5090. So it's the same process node as the previous generation.

16:56 - Jonathan Bennett (Host)
That actually makes the uplift fairly impressive.

16:58 - Jeff Massie (Co-host)
then if there's no bump from the node, Well, yeah, but they added a whole bunch more cores and the memory bandwidth went way up, and so I mean it.

17:10
You know the process node helps, but it's not the only thing. I mean, you're right, you got a lot of design that you can do to make things a lot faster while staying on the node. But it kind of helps to jump both of them up to yep and and we'll see how it's going to work within uh, amd, because the the rumors on the street are that these cards are going to be very limited for several months and right now amd is talking about they're going to release their cards about march because they're waiting until it's ready. They decided they're not going to launch at the same time as NVIDIA because then they have egg on their face because maybe the drivers aren't quite ready or they've got some hiccups and there's talk of they're also building up inventory. So if the NVIDIA cards are all in the hands of scalpers and things are ridiculously expensive, amd comes out with some good cards and plenty of supply. They could really make an impact on the market and get their market share back up at least partially. Yeah, for sure for sure.

18:17 - Jonathan Bennett (Host)
All right. So next we are going to move into story about flop d. I have too much fun saying that.

18:26 - Ken McDonald (Co-host)
But Ken has that one.

18:28
And we'll probably end up using it for updating the firmware on those video cards at some point. But Marius Lester wrote about the fourth maintenance update to the latest release of the open source Linux firmware update utility for Linux-based operating systems. I've got to say it this way too, jonathan flopd2.z.4. This release introduces new features like the ability to record the entire USB descriptor in the emulation data and return defined return code when network metadata refresh fails.

19:17
Slash PID as a full instance ID, a more specific instance ID for QC-S5 Gen2 USB devices, fadvice64 to the systemd syscall allow list and allows disabling of the zero-length pocket for modem manager devices and recovering of the Logitech Bolt receiver in bootloader mode. Now also fixed in this release is a possible critical warning for MediaTek scalar devices, an issue with fire hose padding for some modem manager devices and an issue with UEFI capsule updates when using 4096-byte NVMe block size. Bug fixes include improving FWAPD to correctly parse CSV streams that do not contain trailing nulls, disable reading the option ROM device after dumping and no longer claim the kernel interface to avoid parade downstream port resets. I recommend reading Marius' article, since I've only touched on some of the improvements in Flopd2.z.4, and I feel like Elmer Fudd every time I say Flopd.

20:55 - Jonathan Bennett (Host)
It's interesting that they're playing with the USB descriptor stuff. I've been working with that sort of thing a little bit here recently on another project and it's interesting. You know you've got the VID and the PID. So the VID it's the vendor identifier and you have to pay $6,000 to the USB consortium to get one of those and then you can issue your own PIDs and so you'll have that pair of two-byte. I think it's two-byte identifiers. You know, two-byte for the vid, two byte for the pid. And the fact that uh, you know they're, they're looking at making those. It seems like a little bit more um, fine-grained in doing things. It makes me think that maybe they accidentally pushed firmware. You know one. One company has multiple devices under the same vid but different pid or maybe even a different USB descriptor. That's yet another string that's in your USB stuff. It makes me think that they are accidentally pushing firmware to the wrong device and somebody complained about it. But yeah, interesting updates here.

21:59 - Ken McDonald (Co-host)
Yeah, especially if you've got a vendor that's putting out pretty much the same model. The only difference may be the difference in capacity or something.

22:12 - Jonathan Bennett (Host)
Yeah, you could see somebody reusing a PID and just having a different device descriptor string and that could cause problems.

22:20 - Ken McDonald (Co-host)
It definitely would. If you're using Flopd to update it, yeah yeah, potentially All right.

22:26 - Jonathan Bennett (Host)
Potentially All right, Rob. What is this? What are you doing? What are you doing on my podcast talking about this stuff? Rob has a story about Copilot and, depending upon how this goes, he may or may not be back in the future. Anyway, Rob, take it away and tell us what this is.

22:49 - Rob Campbell (Co-host)
Copilot is here and it is here to stay, and with that many new keyboards are coming out with the new Copilot key. Now I'm sure all the Linux users are concerned that you know now they're going to have another useless key in the future on the keyboard, that you know that doesn't do anything. But you know, just like when that nasty Windows key came out and we had no use for that, we took that key, we made it a super key or a meta key and offered our own functionality to it. So too will we do with the copilot key. As of Linux kernel 6.14, updates to the keyboard driver now map the F23 key to support the default copilot shortcut action. So what will this key do for us? Well, nothing yet. I guess. The support is there in the kernel and now the desktop developers will need to decide what to do with it.

23:58
What are some of the commenters saying? Some have said I would like to use the key for toggling between applications like Alt-Tab or switching to another TTY where HTOP runs. One said it would be bound to his OS update script. And another says or you know, actually this is mine. I say I would like to see it just be programmable by default. Just do whatever you want with it. I know you could essentially program any key, but just make this a programmable by default key. Somebody else in the comments says they're adding a key or they aren't adding a key. They just took away my right control key. Yep, ms took away your right to control and injected their spyware button. So, oh, that's great. It is what it is. Let's make the best of a situation.

25:06 - Jonathan Bennett (Host)
Yeah, so apparently vendors are calling it F23. I didn't know that any keyboards had that many function keys. So as soon as you have the new kernel, then your copilot key will be just another key code and you can do whatever you want to with it. It'll show up in KDE. It'll show up to with it. It'll show up in KDE, It'll show up in gnome, It'll show up in all your applications. You know you're, you're, you're pressing it. Press the key now to bind and it'll just work now with a six 14.

25:34 - Rob Campbell (Co-host)
So don't don't fear that copilot button, don't fear the copilot, assuming you have a meta key.

25:41 - Jeff Massie (Co-host)
I don't IBM type M's, just there's a gap there yep, yep, bring back the gap, yep alright then my muscle memory would be out of place.

25:55 - Ken McDonald (Co-host)
Alright, still in the same location.

25:58 - Jeff Massie (Co-host)
All the keys are still in the same location I'm used to using the meta with the page up for going to full screen oh, I had a page up.

26:13 - Jonathan Bennett (Host)
There are some weird keyboard shortcut combinations out there yeah, that one's available in kde are there any new keyboard shortcut combinations in GNOME, maybe GNOME 48? Jeff.

26:28 - Jeff Massie (Co-host)
There could be. We've been doing a lot of talk lately about KDE and the upcoming 6.3 release, which will be out in a couple of weeks, so I thought we should give a little love to GNOME. So this week we're talking about how GNOME 48 alpha is now out for public testing. If you look at the link in the show notes, there's an article going over what's new in this version. The software package manager has received an experimental plugin which should give improved performance when you're loading updates, when looking at software you might want to load on your system. There are improvements to the upvoting and downvoting reviews so you can better decide what to get and what to pass over and when getting those packages of software, there's now support for including dependencies when estimating download size. There's also improved uninstallation of snaps. Andnome will let you know about microphone permissions when an app has pipe wire access. Gnome 48 has on-screen display notifications for headphone connections and support, if you desire, for screen time and health breaks, and support even for screen time limits. So, coming as well as improved color management, support, detection of preferred primary display devices, frame rate on monitors attached to secondary GPUs in copy mode and better accessibility of the keyboard backlight toggle in quick settings. Which quick settings also has an improved appearance For accessibility.

28:11
Gnome has Orca, which is a screen reader, and it has improvements as well. They're replacing the handling of when a program admits too many events, from handling to protection, so the program stays responsive. When a program floods Orca with too many items, so it goes off the rails and suddenly floods it. It got better handling and not just overwhelmed. There's also better logic so it can tell if an object is only layout and not actually conveying other information, and it's now better support for Orca to speak the digits for a telephone number, independent of the user settings.

28:51
Many apps inside GNOME also got some love, such as a calculator which receives some pressure units and a keyboard shortcut for clearing up history. The Epiphany web browser now has a warning when you want to disable a website's data storage and it has better history function. The control center now has support for battery charge limitation in the power panel and other apps like the front viewer, clock, snapshot, sudoku and many other items are receiving feature improvements and polishing. If you take a look at the article linked in the show notes, you can see all the details and, if you so desire, there's a link to the actual discourse or gnome announcement page where you can look at the changes per library. So see, and it just goes for you, rob, I can give gnome a little love I'm happy to see that, yeah uh, it sounds like it's not a world rocking release pretty much.

29:45 - Rob Campbell (Co-host)
You missed the killer feature in there what's the killer feature?

29:48 - Jeff Massie (Co-host)
which one's that?

29:50 - Rob Campbell (Co-host)
uh updates to sudoku yeah, I mentioned, sudoku had updates no, you mentioned it, but uh, but jonathan's saying it's not a big killer, uh, update he apparently didn't hear you say that yeah, I, I just I guess I don't.

30:08 - Jonathan Bennett (Host)
I don't hold sudoku and quite as high of esteem as I should you're right, it's.

30:15 - Jeff Massie (Co-host)
It's kind of 48 is going to be a lot of polishing and kind of refinement and not any major or shaking. But you know that's probably good, because a lot of times that's when they break the api and make everybody mad and I was.

30:30 - Jonathan Bennett (Host)
I was just thinking about that. It's like what gnome decided they weren't going to make everybody mad this time. Yeah, I guess that's in every other release, maybe. Maybe gnome has a an odd. No, is it? Is it? They're odd releases that tend to take everybody off.

30:43 - Jeff Massie (Co-host)
I don't remember well, maybe they're just saving it for 50 I guess, that's what I was thinking.

30:48 - Jonathan Bennett (Host)
50 50 will be the big one.

30:52 - Ken McDonald (Co-host)
Yeah, yep, yep and then you'll see another fork, inevitably I mean a project as big as gnome forks from time to time are inevitable.

30:56 - Jonathan Bennett (Host)
It's just and then you'll see another fork Inevitably. I mean a project as big as Gnome. Forks from time to time are inevitable. It's just a question of how popular they get versus the original sort of mother project.

31:08 - Rob Campbell (Co-host)
The next will be turmeric.

31:10 - Jeff Massie (Co-host)
Turmeric. Huh, because they have cinnamon, he's making a spicy joke. It's spicy he's making a spicy joke. It's spicy, well, and the thing is, you know, the hard part is when they branch off, a lot of times it can cause problems later on, like going to wayland. That was a major issue for some of those pre-wayland forks and yeah, I believe they've gotten there, but it's taken a it's experimental but yeah, but it's taken a major effort because they kind of wound up coding themselves into a corner, so to speak.

31:48 - Jonathan Bennett (Host)
Yeah, and so there's the question of do you do all the work yourself or do you try to rebase on top of the upstream uh gnome? And if you do the rebase, are you going to accidentally not accidentally, but are you going to end up pulling in all the code that you wanted to avoid to start with? And yeah, it's a challenge. Um, it's kind of nice, at least for forks, that we are now on the other side of the whaling tissue, most for most, most places, and that, uh, not going to be quite as big of a deal going forwards.

32:17 - Jeff Massie (Co-host)
But yeah, so we should. We shouldn't have any major groundbreaking until we replace wayland, which hopefully it'll last, you know, at least 20 years even even open.

32:28 - Jonathan Bennett (Host)
Seuss got alex cute on uh on wayland recently, so yeah, you know it is a little different though, because with wayland the desktop environments are doing to some extent they have to do more of the work themselves, whereas with X11, a lot of that work, a lot of code, lived in the X11 code base. In Wayland there really isn't any code that lives in the Wayland code base. The Wayland code base is really just the specification. You sort of get to do all that work yourself inside of your, uh, inside your desktop environment. So it is just a little different. Well, right, right, right, it is the compositor, specifically that that interfaces with you know, your, your graphics driver and all that, and there's not as far as actual running code. There's not actually a lot of wayland that exists up there. It's just sort of the specification.

33:17 - Jeff Massie (Co-host)
That's why they have that reference compositor, just to.

33:20 - Ken McDonald (Co-host)
WL Roots.

33:21 - Jeff Massie (Co-host)
Yeah. So they can just say well, here's an example, you know, you can.

33:26 - Jonathan Bennett (Host)
I think there's actually some desktop environments that just use WL Roots. Isn't there? I can't remember off the top of my head which one it is, but there's not any still using it.

33:34 - Rob Campbell (Co-host)
There probably are.

33:36 - Jonathan Bennett (Host)
I'm sure there are.

33:37 - Rob Campbell (Co-host)
Yeah, even I don't remember what it was. But last time I was on the show, the desktop environment I talked about still used WL roots, but they planned to create their own compositor. I remember that much of it, but I can't remember what desktop environment I was talking about.

33:54 - Ken McDonald (Co-host)
Yeah.

33:56 - Rob Campbell (Co-host)
That was Budgie, I think.

34:01 - Ken McDonald (Co-host)
I think Budgie still is is using it, but they plan to make their own yep, and then I talked about a lxqt last week, about how they're setting it up, so you can just combine it with any compositor you want is sway a desktop environment.

34:17 - Jonathan Bennett (Host)
I think it I think it's the compositor. Yes, so a building block of a desktop environment? Um, yeah, I was looking. I was looking quickly to see if I could tell which ones were out there that still use wl roots and or and or sway. Um, not immediately seeing the answer, but I'm sure there's at least one of them.

34:36 - Ken McDonald (Co-host)
LXQTs are going to be able to use it.

34:40 - Jeff Massie (Co-host)
There, you go All right WL roots, it'll at least get you off the ground, you know, because you still have to support probably a lot of other code, and that at least the code surrounding the compositor you can get ready, and then why you then write your own compositor so you can kind of attack it from two different angles. It's sort of a bootstrap process with all of this uh multiple compositors and everything.

35:07 - Ken McDonald (Co-host)
Are you going to need a way to boot between multiple live uh cds?

35:14 - Jonathan Bennett (Host)
well, that's an interestingly great segue to Ken's next story. This is something that I actually run as well. I don't know if it was Ken or Rob or Jeff. One of you guys really got me started on this and it's become an indispensable part of my toolkit, and that is Ventoy Ken. What is new with Ventoy?

35:32 - Ken McDonald (Co-host)
Well, actually, bobby Borsroff wrote about what is new with Venturi. The popular multi-boot utility for creating bootable USB drives for ISO files, released a brand new version 1.1. The new version updates the latest shim, effectively fixing the verifying shim sbat data failed error. The update also resolves any associated secure boot issues, making the entire boot process more secure and straightforward. On the bug fix side, the developers tackled the boot issue affecting System Rescue 11.02+, so you can rest assured that your rescue and recovery environment will be ready whenever needed. Some minor bug fixes have also been polished, rounding out the overall user experience. On the OS side, ventoy 1.1 expanded its support to the I hope I'm saying this one right UOS ISO.

36:47
It's a muscle-based, lightweight, general-purpose rolling release distro. Now I looked this up just so I could understand what muscle-based meant. Now I looked this up just so I could understand what muscle-based meant. It's basically an easier way to say. Based on the C standard library developed by Rich Felker, the muscle provides consistent quality and implementation behavior from tiny embedded systems to full-fledged servers. Also, I wanted to give you a tip when using Venturi servers. Also, I wanted to give you a tip when using Venturi. F2 will allow you to directly browse and boot files on your local disk. As always. Check out Bobby's article for anything I didn't cover.

37:30 - Jonathan Bennett (Host)
Yeah, muscle. I'm trying to remember exactly what that's a reference to. It's like the Mew C library, something like that, but you see it on. I'm trying what that's a reference. It's like the, the mu C library, something like that. Um, and but you, you see it on, trying to remember what, uh where exactly muscle gets used. Uh, open WRT might use muscle, I don't remember. I think, uh, alpine Linux uses muscle, if I remember correctly. Yes, there's a couple of them out there that use it.

37:54 - Jeff Massie (Co-host)
Um, out there that use it um.

38:01 - Jonathan Bennett (Host)
It's an interesting little alternative. Uh, c standard lib.

38:03 - Jeff Massie (Co-host)
Um, that sometimes the record ain't to work with, I'll say, and for the record, I was the one that uh introduced you to ventoy. It was in episode 12, august 7th 2021, yep that long ago.

38:14 - Rob Campbell (Co-host)
I am the one who introduced you to I ventoy well.

38:17 - Jonathan Bennett (Host)
Well, that's true. That's two different projects. What's the difference between those two, Rob?

38:21 - Rob Campbell (Co-host)
iVentoy, that's the Pixie, boot.

38:23 - Ken McDonald (Co-host)
Right, right, right, right, right, right yeah yeah, and you pay to be able to use it for ARM devices.

38:30 - Jonathan Bennett (Host)
Yeah, thwandal in the chat points out that it's Alpine Linux that makes use of muscle. I in the chat points out that it's Alpine Linux that makes use of muscle. I was thinking that I may or may not have said it, but if I didn't say it, it is Alpine Linux. There we go. My brain is already on the next thing, trying to keep it going. Yeah, I'll have to go grab the Ventoy 1.1 update. I will say this If you go to update your ISOs on Ventoy, test them before you need them, because sometimes, like particularly with a bleeding edge distro like fedora, you will occasionally run into problems if you have not updated your ventoy often enough. And then you end up in desperate times like having to boot back into windows. It's terrible, terrible terrible, yeah, that's why you?

39:19
have two, two venturi, uh drives that would be the way to do it. Yep, yep, all right. Well, rob um, I've told people for a long time just don't use suspend. And for the longest time I know suspend was terribly broken on linux, on almost all hardware, and not as bad as hibernate. That's true.

39:40 - Ken McDonald (Co-host)
Suspend is not as bad as hibernate um as well, just shut down if you're going to use them yeah, well, that's kind of what I figure, yeah I mean I kind of thought upgrade an ssd you know made startup and shut down so fast that who cares right?

39:55 - Jonathan Bennett (Host)
that's not necessarily the case, though, is it?

39:56 - Rob Campbell (Co-host)
you've got a story about this oh, you know, I right, I thought it made startup and shut down fast enough. I thought an ssd, you know, would make suspend and resumes fast enough. Uh, you know, faster than I could ever need. But uh, some, for some you, you can never be fast enough and I guess that is why updates.

40:19
Another update to the 6.14 kernel included ACPI updates switching from the msleep function to the usleep underscore range function within the acpi underscore os underscore sleep function call. The ACPI underscore OS underscore sleep function call In the kernel reducing sleep time due to timer inaccuracies. The Linux ACPI slash PM maintainer, rafael Laisaki of Intel, who authored this change, noted that it could spectacularly reduce the duration of system suspend and resume transitions on some systems. This addresses bug reports back from 2022 on a Dell XPS laptop taking eight seconds to suspend and eight seconds to resume. Another report suggests this ACPI change noted another Dell XPS laptop giving a kernel resume time of 1.9 seconds down to 1.1 seconds.

41:32
Raphael said, quote the extra delay added by mSleep to the sleep time value passed to it can be significantly or can be significant roughly between 1.5 nanoseconds on systems with hz equals 1000 and as much as 15 milliseconds. Yeah, yes, 15 milliseconds on systems with HZ equals 100, which is hardly acceptable, at least for small sleep time values. And another bullet point he says, quoting him he says mSleep on the default HZ equals 250 in Ubuntu on the modern PC takes about 12 milliseconds. This results in over 800 milliseconds of spurious system resume delay on systems such as Dell XPS 139300, which use ASL sleep parentheses five milliseconds in a tight loop. So keep looking for more efficiencies, even if sleep resume hibernate is a waste of time, like some of us think. If you like using it, it's going to go faster for you now.

42:53 - Jonathan Bennett (Host)
Yeah, that's interesting. I wonder if this change is just. I think it's just related to sleep and resume. It's not necessarily a. Yeah, they switched from mSleep to uSleep range. It was inaccurate, right? So they would say in the code it might say, well, do an msleep for five milliseconds, and it would actually do an msleep for 12 milliseconds, which is hardly anything, just those two. You would not ever notice those right. But in the midst of like a, not a shutdown script, but like a hibernate script or a suspend script, you might have 75 of those calls and well, you have enough time. Add up through all 75 of those calls that suddenly you have, you know, an extra second or multiple seconds of time. So yeah, it's just one of those things where someone sort of did the math right. This is what this sounds like. He did the math and went wait a second, this doesn't add, this doesn't add up. Dug into it, go. That's why it doesn't add up. It's because m sleep is not accurate for small sleep times.

44:06 - Rob Campbell (Co-host)
So very cool so does it even really work on windows because I know one time even windows had problems with all the sleep and I think pretty much all os's sleep and hibernate work a lot better than they did, say, 10, 15 years ago yeah, I'm trying to remember.

44:26 - Jonathan Bennett (Host)
I've seen a couple of videos from people that actually knew what they were talking about discuss how, how much of a colluge, just a terrible, terrible hack, all of the like, sleep states and hibernate, all of that is. It's just all sort of flaky and held together with duct tape and bubble gum, so it's not all that surprising that it goes wrong from time to time. I have fixed multiple people's computers for them just by finally getting the thing to come out of hibernate, right Like you got to unplug it. Multiple times I've seen this where, like on a laptop, you got to take the battery out, the wear out the power button for a minute. Then you can put the battery back in. It'll do like a fresh boot rather than trying to come back from hibernate or come back from suspend bonus windows tip power cfg space forward, slash hibernate space off or even just slash h off space off disable hibernate?

45:20
yeah, I do that quite a bit. It saves you. Saves you some gigabytes on your hard drive space too, doesn't it?

45:26 - Rob Campbell (Co-host)
that's usually why I do it when somebody's running out, I'm like you don't need, you don't need uh to hibernate anyway yeah maybe a 20 gigs yeah, or save more space and don't have windows that was an option.

45:39 - Jonathan Bennett (Host)
Windows for the longest time had this terrible bug that it would just fill your hard drive up with hundreds of gigabytes of log files from I think. It was trying to compress the log and the compression would fail and it just would never clean up the old files. I've ran it. This has been a couple years ago now, I guess, but I ran it.

45:58 - Rob Campbell (Co-host)
I've run into that problem on several computers I've seen 50 to 100 gigs of log files and that's why you'd reinstall windows every six months that's one of the reasons.

46:09 - Jonathan Bennett (Host)
Uh, all, right now let's talk about something that's not windows but you can have a lot of fun with, and that is the steam client. Jeff, what is new in the Steam client?

46:20 - Jeff Massie (Co-host)
Well, we've talked about a new series of GPUs, we've talked about a new desktop. Now it's time to talk about gaming. Steam had a January client update and I think we should take a little time to go over what they've done, because after all this hard work, we need to relax a little. So the update was released on january 21st. If you don't have it yet, you don't need to worry about actively downloading it, as your steam client will automatically update itself the next time you start it up. Valve added a global setting for improved management of game updates, along with the panel to manage per game exceptions to the global settings. If you have a game that's actively downloading, you can click on the game image on the downloads page and it will now take you to this, the game's library page. There's also an updated layout for the actively downloading game, so the client now has better caching of soundtrack album covers in the library and improved library asset loading times for large libraries. If you use streaming, things should be better now, because they fixed a case where Steam would prefer to stream from another PC on your local network, even if the game was installed on the PC that you're sitting at and you could play it without streaming. So even if you were sitting at your local machine, you were like, oh, I'm going to play this game, it would try to stream it from another machine in your house that had that same game. So now it shouldn't do that.

47:56
Now, for Linux-specific issues, they fixed a case where Steam could permanently stop the screensaver if it crashed while holding the screensaver lock. There was a potential crash when exiting a game when game recording was enabled on the system. With recent versions of Pipewire, if you had a high single-core CPU usage with a large library during Steam startup, this should no longer be a problem. And when using Steam player-to-player voice chat in games, there should no longer be a crash. They also fixed a crash in the game recording viewer caused by unstable V-A-A-P-I drivers.

48:38
And finally, for Linux, they fixed a rare crash when hot-plugging a mouse and keyboard into your system. There were some other little bugs in Steam input that could sometimes cause crashes, and those should be gone now as well. Now there are many, many other fixes that have been added to the client, so I suggest you take a look at the article in the show notes for full details, and there's also a link in the article to the actual Steam client update announcement from valve, where every single update is listed and organized in an easily readable format. So happy gaming fun stuff I gotta.

49:17 - Jonathan Bennett (Host)
I have to look into the uh, the, the fix where it, uh it blocked your screensaver. I always had had trouble with my computer failing to sleep the monitors overnight and particularly worrying about burn-in with that. So I'm now down a little bit of a rabbit hole trying to figure out what's the command to tell your screens to go to sleep. It's like I'm sure there is one.

49:47 - Jeff Massie (Co-host)
Well, based on the bug, I believe you need to go in and delete a lock file. Yeah, that makes sense too, Because it would basically have that lock file crash and then that lock file just stayed there.

50:02 - Rob Campbell (Co-host)
Mm-hmm, mm-hmm, that lock file just stayed there, mm-hmm.

50:05 - Jonathan Bennett (Host)
Mm-hmm. Yeah, I have found somebody's project to be able to put their machine to sleep and it's a bash script. That's essentially one line in a bashsh, and people on Reddit are, of course, flaming this guy for why did you make this a bash script when it's a single command? It's hilarious, so look forward to that in a future command line tip. Yeah, all right, let's see. Ken, you've got the last one from you guys. I've got one story to follow up with after this, but Ken is talking about Rhino Linux. What's new there?

50:47 - Ken McDonald (Co-host)
Yes, this, but Ken is talking about Rhino Linux. What's new there? Yes, and we can thank Marius Nester for writing about the latest release of Rhino Linux 2025.1. Rhino Linux is an Ubuntu-based distribution offering a rolling release model on top of a stable desktop environment. So there is a version of Ubuntu that you can use that's rolling release.

51:08
Rhino Linux 2025.1 includes support for dynamic workspaces in Rhino's XFCE-based Unicorn desktop to automatically create new workspaces when opening applications, a new custom grub bootloader theme for more modern fill and a new testing meta package called Rhino Stampede. In addition, it introduces a brand new companion application called Hello Rhino, written in Rust and IcedDTK that makes it easier for users to access Rhino Linux's homepage, blog, discord, community and, most importantly, documentation. Under the hood you find Rhino Linux 25.1, powered by Linux kernel 6.12 and using Paxtel 6.1.z as an AUR-inspired package manager. It also updates the kernel on the Pine64 images to Linux 6.9 and the kernel on the Raspberry Pi images to Linux 6.11. Rhino Linux announced they will be at FOSDEM 2025. They will be giving a talk on the distro's track about the creation, maintenance and future of Rhino Linux and PaxDoll on this coming Sunday. Rhino Linux and PaxDoll on this coming Sunday, february 2nd, at 2.30 pm.

52:52 - Jonathan Bennett (Host)
Universal Time Interesting Is.

52:58 - Ken McDonald (Co-host)
Rhino Linux, one that you would recommend for some of our listeners to give a try.

53:04 - Jeff Massie (Co-host)
If you want a rolling release, yeah, and who doesn't? I? I actually watched the uh release distro tube on youtube. Dt he had uh, he went over it and it looks pretty nice the one. The one thing is I wish it had kde but, otherwise I thought it looked. It looked really nice.

53:24 - Jonathan Bennett (Host)
Yeah, rhino is not an official Ubuntu or canonical thing, right? It just happens to use their. Yeah, yeah, all right, interesting, very cool, very cool, all right. I do have a story and it is a follow-up to something we've talked about in the past, and that is the BcacheFS merge that got blocked for the Linux 6.13 kernel. I did not make secret that at the time and I'm still of this opinion that that is a terrible policy to keep particularly bug fixes out of the Linux kernel as a punishment for anyone. I just think it's a fundamentally flawed policy for multiple reasons, which I'm going to do then. I will not go over again, but it's a bad idea. Please stop it.

54:16
But I'm happy to say that the 6.14 merge window of the kernel is now open and the bcachefs changes changes, which was quite a large set of changes, was pulled in with absolutely no, uh, no problem. It was not held up, there was no, no fuss made it landed and you're going to get all of your new fixes and bcache effects in 6.14. So they they did indeed follow through the linux of Conduct Committee in Torvalds. They followed through that. It was only to be a one version ban and things are once again working the way that they are supposed to, so for that we are thankful. I am glad that this is the story we are covering and not something else about BcashFS and all of this mess. So it's good to be back. It's good to have them back. No kerfuffle, thank goodness. No more kerfuffle, yes, absolutely.

55:17 - Ken McDonald (Co-host)
Basically, we, the users, got punished for that.

55:22 - Jeff Massie (Co-host)
Yes, that is a part of why I think it is such a bad idea only the bcash users, which is really small well, you know, if I read that and I looked at that story and it seemed like, yeah, it kind of it definitely punished the users a little bit, but it seemed like there was some extra coding that went on. So this code that got merged in is a little more polished now and but it. But it is a different format. So you've got to uh, have a conversion because the base structure, whatever changed and they've got revision numbers on it and but there's, I guess, a tool that should convert it for you.

56:08 - Jonathan Bennett (Host)
Yeah, so that was one of the other things that had been complained about with BcacheFS, completely separate from the code of conduct stuff. This is something Torvalds complained about. They tended to sort of play fast and loose with pushing code in that maybe wasn't quite as polished as it should have been. So there is a sort of a devil's advocate point that we should not let them push code. You know, we should always force them to be a month behind like this and maybe we would get better code out of the bcash fs project, and that that's sort of a separate issue. Um, but I I'm just, I'm just glad that they're back and that we can push things forwards and hopefully we won't ever have to have a story like that one again.

56:51 - Jeff Massie (Co-host)
uh, maybe, maybe that's a little over optimistic for me, but I'm hopeful oh, and I and I looked it up the the disc format change go. The version goes from 1.13 to 1.20 and they're saying that should be the last format change they're planning the last one, for sure.

57:12 - Jonathan Bennett (Host)
Really, really this time guys, yeah, r3, yes no, this is an rc, sorry, yeah, yeah yeah, all right, let's get into some command line tips then, and we're going to let Rob kick us off first with Ignition.

57:32 - Rob Campbell (Co-host)
So, as I start to use Linux Mint, one thing I notice is it does better than, say, ubuntu is how a user manages startup applications, how a user manages startup applications. Ubuntu startup applications makes you navigate using a file manager, pop up to the exact desktop file or runtime needed, which is you know. It's okay for dabs and images, less obvious for apps in other formats. Linux Mint is more user-friendly and informative with simple toggle, startup delays and simple ways to add new startups. I'm not going to tell you how to run Mint startup application, ubuntu and other desktops, but instead offer an alternative called Ignition. Ignition comes to us from the developer of Warehouse, which is a flat pack manager we've spoke of in the past. Ignition does everything the Ubuntu startup app does, but easier and better. It's an opinion, but Ignition provides a simple UI to add or remove, modify startup entries on your computer. Ignition can add apps, scripts and arbitrary commands to the login. You can try it yourself in Flatpak.

58:56
For those watching, I have a quick little side-by-side here. On the left here is Ubuntu startup applications and on the right is Ignition. It looks about the left here is as ubitu startup applications and on the right is ignition. It looks about the same here. Now if I go to add a startup application in ubin to put a name, I could browse and it doesn't look all that friendly. That looks not great. But and you put a comment and not a whole lot else, that's great.

59:28
Let's go over here to Mission. If I hit New, I can add an app or a command or script. Let's start with the app. Look at this. It shows all my installed apps and I could just pick one. That seems quite a bit easier.

59:45
Let's just run software on startup and you can have it enabled. You can do a name, you can do a comment, the command, the commander script there you could show it. A terminal wouldn't make sense for this. And then you can create. Also, if you wanted to do a commander script, you have the add a command or script and with that you can basically just put in here any command you want. And when I was playing around with it I put in this echo hello command which, and then show the terminal. So when I start up I get a nice little ho which, yeah, not very useful, but uh, I was just testing it to see how it plays around. So, um, it still could have add some new features like delays, but if you're comparing that against the very basic ubuntu startup one, it's it well, I think it's a lot better all right, good stuff.

01:00:50 - Jonathan Bennett (Host)
Now the question that I have is can you run ignition on other distros, or is it meant only?

01:00:58 - Rob Campbell (Co-host)
uh, other distros. It's not even meant at all. This is uber to. I was running it on here. All right, it's a flat pack so you could run it anywhere. Flat pack could run.

01:01:07
One caveat uh, for example, if you already have startup things um added via, like, say, the uber 2 startup um, startup preferences, when you go over, when you first start up ignition, they're not going to show up in there because of the sandboxing. Now anything you add into Ignition will show up in your other startups. A quick way I found to get around that is like for those. When you're looking here you see, like this SSH key agent. That was a default one that was already in Ubuntu right at the beginning and it did not show up in ignition at first. I just disabled it in here and as soon as I disabled it it showed up here and I was able to enable it again and manage it within ignition. So that's a little caveat to be aware of if you already have things added through the normal interface and then you start using ignition because of the sandboxing yeah, all right, very cool.

01:02:10 - Jonathan Bennett (Host)
All right, jeff, I see you have one here that is very much in your, in your wheelhouse yeah, yeah, uh, the command line tip I ran across.

01:02:21 - Jeff Massie (Co-host)
well, I ran. I ran across this command line tip when I was trying to help a friend pull data from a hard drive. It was a Western Digital NAS that died on them and they couldn't access the disk. But they'd read it was a Linux file format. Well, they don't have Linux and they're not very computer savvy, so I was asked to help.

01:02:39
Now there were some issues and I needed to see where the drive was located in the system. And what I mean located, I mean is in the slash dev mount point, and it wasn't readily visible. A little searching had me find the ls scuzzy command. So it's ls scs I, which is a command that lists information about scuzzy devices in Linux. Now, that might not sound super useful to most people, because most of us don't run a SCSI device, but it also picks up ATA, pcie, sas, usb and many other drive interface formats. So it's more, and the simplest way to use it is simply type in your shell lsscsi. Now it'll give you a readout of the drives on your system, tell you a lot of information about what your disk is, the name of the drive where it's mounted in the slash dev directory there's a dash h so you can get the size of the drive and human readable format. There's a verbose option which gives a lot more information and you should really read the man page for what every item is. And of course there's a dash dash help.

01:03:57
But you know I'm not going to go into the deeper commands and interface specific options. You know, because there's specific. You know SAS commands, sata commands, all that stuff. So without making this, you know, an hour long, even longer podcast I'll just have, I'll just say I'll leave some experiment or investigation to the listener and viewer. And if you need to have more information about a drive, take a look at the program. And there's a link in the show notes to a full and detailed description beyond the man page. It's a special page set up, I think, from the people that program it. So it's very helpful, it's organized really well. So each device interface has got its own little subsection. So you know, give it a whirl. I found it very helpful, so I highly suggest it.

01:04:50 - Jonathan Bennett (Host)
Yeah, interesting. I learned something just now. I think I may have noticed this in the past. So most computers will have like a slash dev, slash SDA, that's if you have a SATA device. The S in SDA does not stand for SATA, it stands for SCSI. Sda is SCSI disk. A Linux sees almost all disks as SCSI disks. That's fun.

01:05:13 - Jeff Massie (Co-host)
Yeah, and I think isn't the SATA standard. I think it's kind of a SCSI standard, I thought.

01:05:20 - Jonathan Bennett (Host)
It may be somewhat derived from SCSI. A quick Google shows that they are not the same. People you know are, I think, sas derived from SCSI. A quick Google shows that they are not the same and people you know are there it's. I think SAS is actually SCSI. Oh, maybe that's not.

01:05:32
Yep, all right, I've got a command line tip and I am going to I'm going to attempt to do a screen share with this. Yeah, there we go, that should work. So my command line tip is VA info. Which VA info is the VA API info, and when you run it you should hopefully see this. And this is essentially giving you information about how your video card is capable and your system as a whole is capable of using your video card to decode video codecs and so on. This laptop, obviously it is a Radeon card, it's running the Radeon SI graphics driver and then these are all the profiles that it supports MPEG-2, simple, mpeg-2 Main, all the H.264s, several of the HEVCs and then the VP9 profiles.

01:06:30
And the reason that I discovered this particular command is because I was working on trying to get Firefox to do something with live video or with decoding video, and I was having trouble with it, and one of the things that came to mind is I wonder if I have VA API set up to be able to actually kick it out to the video card. Found this command, ran this command, to discover that, no, I did not have any of the VA API libraries installed, and so I went on a bit of a hunt to get the desktop behind me to show a list very similar to this, the desktop behind me to show a list very similar to this, that, yes, indeed, I can decode video right on the video card and my my hunt for actually what it is it's for, hdr, is not quite done yet because the patch is not quite there in Firefox, but it's close, it's coming soon, and I will definitely let you know about it when all that lands. Nice, yeah, all right, ken Ken has something that I think dovetails rather nicely onto that.

01:07:33 - Ken McDonald (Co-host)
I do, but I just wanted to say that I found out earlier when I was playing with VA info that it will determine whether or not you have a Wayland or X11 using It'll try both Wayland first and then, if it doesn't find Wayland display, it'll switch to X11.

01:08:00 - Jonathan Bennett (Host)
Oh, interesting.

01:08:02 - Ken McDonald (Co-host)
Interesting, but what I'm going to talk about today it has to do with controlling your video devices or even just identifying them. It's a V4L2-CTL. It's part of the V4L utilities. I didn't provide a link to that. I'm going to put that in the show notes after we finish here, but I'm starting off by showing the first part of the output. If you do a dash h or dash dash help, it gives you all the general and common options that you can run. I tried the dash dash all Boy. There's a lot of information there, but I'm not going to display that today. What I am going to show you, though, is the version. In my case, I'm running V4L2-CTL1.28.1. I believe that's actually the latest one, and it also shows you the remainder of those options from the help. The next thing I'm going to show you in my third screenshot. By the way, for those listening, I do have a link to a document with all the screenshots so you can pull it up and look at them yourself. But with the third one, I went and choose the dash dash list dash devices, and, for those listening, it lists all the devices it found at the time that I had running. One, of course, was my webcam, and it actually sees that as three different devices Device video one and dev video zero and Dev Media 0. And then it also finds my OBS virtual camera as Dev Video 2. And it says it's a V4L2 Loopback device.

01:10:21
Going to the next screenshot, I'm showing how you can use V4L2 to list all the controls that you have. It breaks it into user controls and camera controls. If you've ever used OBS Studio, you may be familiar with some of these, because they do appear in there. Appear in there, but the nice thing about this is you can see the minimum maximum for your controls, what the default value is and what the current value may be. As you're looking, there's also another command it's our option dash, dash, dash controls, dash menus, which pretty much shows the same thing, though it does give you some other information that the other one didn't, depending on the device you identify for it. Speaking of which, you can also put in there a dash D and specify the device before doing the dash dash list, dash controls, dash menus. That way it gives you the specific information for the device by default. It does go with the dash slash, dev slash, video zero device. If you don't do that, because you'll see it's repeating all that. Now you can try to.

01:11:50
I even tried using a device that was not there, and you'll see that when I did one that wasn't in use, it didn't give me anything for it, and then I started using it to play with some of the controls. Here you're seeing where I used it to get the brightness that I had, even though I already had it shown up there. But it'll come back and give you that the brightness is currently set to 130. And give you that the brightness is currently set to 130. And then on the next screenshot it shows where I actually set that brightness down to 65 and you'll see it doesn't make a bit of a difference in how I look. And then I also set it up to 195. So you can see how it does impact your video camera's operation. So this is one way that you could, from the command line, adjust your video camera.

01:12:58
Yeah, very cool and if you got one that supports pan and tilt, you can actually adjust those Ooh.

01:13:08 - Rob Campbell (Co-host)
What about focus? What about focus? I have?

01:13:10 - Ken McDonald (Co-host)
a problem with focus can you do that? Yes, you can you know, and uh, there's also a command out there that you can use, that is a qv4lt, that gives you, uh, like a video test for adjusting it. That basically, so you can test your setups and just different things interactively. So if you've got your camera being captured, like into OBS studio, you can then adjust it and see the change immediately in your display in OBS Studio or in Video Ninja, which is what I was using to show that screen capture there.

01:13:59 - Jonathan Bennett (Host)
Those of us watching Ken's feed probably are asking ourselves whether V4L2 lets you control your camera's frame rates. It depends on the camera. Yeah, all right. Um, well, that is it, that's our show, that's our command line tips. Um, we appreciate everybody being here. I'm going to give each of the guys the last word if they want to plug anything, and they can certainly do that, and we're going to go to Rob first. What does Rob?

01:14:25 - Rob Campbell (Co-host)
have to plug. All right, so my usual, except for I've added a new icon Come connect with me on my website, robertpcampbellcom. On there's my links to my LinkedIn, links to my Twitter. Newly added is a link to my Blue Sky and I don't know why, but my follower count has been going up like crazy. I've noticed I hadn't even logged in for months and I logged in and have over 500, no, almost 500. And I don't know why it just keeps going up and up. I don't know if it's you guys or if it's bots or what, but I figured there's a follower base there. I am going to try to be active on there Also. Yeah, it's like way surpassed my Mastodon somehow. But the next one there is my Mastodon, which I will be in there also. And finally, this nice little coffee cup if you want to donate a coffee to me in increments of $5. Just because you like what I do and you want me to keep doing it.

01:15:37 - Jonathan Bennett (Host)
Yeah, you know, I've threatened for a while now that I would go and set up a blue sky account, um for floss as well as for me, just because there are people that are there and it would be nice to be able to have it for them.

01:15:49 - Rob Campbell (Co-host)
Um, yeah, I had an invite to that when it was invite only from our very own mashed potato. Thank, you. But now it's not invite only, so I'm I'm less special than I was then.

01:16:10 - Ken McDonald (Co-host)
You're less special, but thank you, yeah, all right, ken. Yes, I do want to uh share with you uh link, I've got it in the show notes that talks about the end of an error for a certain type of device.

01:16:19 - Jonathan Bennett (Host)
Yes, sony's blu-ray media production I am not happy about that. I think that is, an overall, a bad thing for the world. Um, I like being able to hold on to my media so that I actually own something instead of just perpetually renting from good news there's still somebody out there making blu-ray.

01:16:39 - Ken McDonald (Co-host)
Uh, media pioneer ah there you go.

01:16:44 - Jeff Massie (Co-host)
Well, and it just to just clarify that that's for the recordable. So they're still going to sell movies on the blu-ray discs. It's just you won't be able to use the read, write or write once material. And but, like ken said, pioneer is still making it. They've kind of found a niche market that they've been living in japan.

01:17:09 - Ken McDonald (Co-host)
Uh, probably some of it, I'm sure because japan's got the requirement that you are able to archive and they're providing archival quality blu-rays which they say will last 100 years.

01:17:27 - Jeff Massie (Co-host)
And they're looking at having some that will last even longer.

01:17:31 - Jonathan Bennett (Host)
You have to ask yourself how do they know that it will last 100 years? They've tested it, duh.

01:17:39 - Jeff Massie (Co-host)
Think about that for a while Rob, actually, I know how there's ways you can accelerate where. So, instead of having to?

01:17:48 - Rob Campbell (Co-host)
have time machines, so right.

01:17:52 - Jonathan Bennett (Host)
That's actually an interesting rabbit hole. We, which we are not going to go down right now, but so what they do is they simulate accelerated wear by doing things like exposing them to wet conditions and then dry conditions, and then hot conditions and cold conditions and abrasive conditions, like. So there's these tests that get designed, and you, you put the media in these tests and you see how it behaves. Um, I've just I'll just throw out there, though, that that is not actually entirely the same thing as letting 100 years go by, and so we may get 50 years down the road and go oh, there's an interaction here that we didn't even know about to be able to make a test for.

01:18:27 - Ken McDonald (Co-host)
There is a way to adjust time now. You move deeper into a gravity well, and then back out.

01:18:36 - Jonathan Bennett (Host)
Yeah, good luck with that.

01:18:37 - Jeff Massie (Co-host)
Well, I mean, I will say that I understand what you're saying, jonathan, but there's an actual kind of engineering, scientific study in all this and there's there's a lot of data and testing that has been done to uh, make sure that you can accelerate where and it it will behave like actual long-term, where I mean there's a lot of long-term, where I mean there's a lot, of a lot of science and engineering that goes into it, and in 70 years we'll find out how accurate it was yeah, no, that's, that's definitely true.

01:19:16 - Jonathan Bennett (Host)
Like I said, we're not going to go down that rabbit hole right now, because that's a deep rabbit hole. Uh, it is, jeff knows.

01:19:21 - Jeff Massie (Co-host)
Jeff knows more about another episode uh, that would be probably a few episodes, but I just wanted to basically just point out that it's not kind of like, hey, let's do this, let's do that, it's there's scientific engineering.

01:19:35 - Jonathan Bennett (Host)
It is real engineering, that is for sure. Yeah, yes, yes, that is for sure. It is real rigorous engineering. Yeah, for sure. Yeah, okay, rigorous engineering.

01:19:44 - Ken McDonald (Co-host)
Yeah, for sure, yeah, um, okay. Well, one other thing I wanted to add is this week's uh command line tip is building up for next week's uh pipewire command line tip interesting, all right.

01:19:57 - Jonathan Bennett (Host)
Nice, cool, all right. Uh, jeff, anything you wanted to plug uh, I really don't have anything.

01:20:04 - Jeff Massie (Co-host)
So this week it's poetry corner again. So the user didn't like how his office was set. He felt a rearrangement would be best. Upon completion, to his dismay, the computer would not respond in any way. Frantically he called the help desk, submitted an emergency request. The deployed tech knew just what to expect. Into the receptacle the power cable was set. Have a great week everybody.

01:20:33 - Jonathan Bennett (Host)
I love it. I love it All right. Thank you guys for being here. I sure appreciate it and I sure appreciate everybody that watches us and listens gets us live or on the download. If you want more of me, you can check out Hackaday. We've got Floss Weekly is now over there at Hackaday, as well as my security column, which goes live pretty much every Friday morning and we don't miss many beats with that. I have a lot of fun there, yeah, and if you're interested, if you want more, if you want to get the show uninterrupted by ads, there is, of course, club Twit, which you should definitely take a look at. It's about the price of a cup of coffee per month and it is a way to support and show your love for us and the network as a whole, and we sure appreciate everybody there at Club Twit. We will be back next week and we will see you then on the Untitled Linux Show.