WEBVTT 00:00:11.447 --> 00:00:16.131 Hello, friends, and welcome back to your weekly Linux talk show. My name is Chris. 00:00:16.131 --> 00:00:17.081 My name is Wes. 00:00:17.081 --> 00:00:17.978 And my name is Brent. 00:00:18.820 --> 00:00:22.651 Well, hello, gentlemen. Coming up on the show today, Linus put his foot down 00:00:22.651 --> 00:00:25.043 on AI in Linux kernel development. 00:00:25.496 --> 00:00:29.911 We'll cover his blunt message he had for developers, and then we're going to 00:00:29.911 --> 00:00:32.044 talk about reworking the Node stack. 00:00:32.293 --> 00:00:35.376 Brent and Wes are coming at me with a whole new hot stack. 00:00:35.834 --> 00:00:39.591 And then I'll do a quick remote Linux desktop showdown. I went through some 00:00:39.591 --> 00:00:44.164 of the more recent remote Linux desktop options, came away with what I think is a clear winner. 00:00:44.553 --> 00:00:47.961 And we're going to round it out with the great poos, I guess, 00:00:47.961 --> 00:00:51.581 the boost, the picks, and more. So before we get there, just sort of save me 00:00:51.581 --> 00:00:54.846 mumble room. Time-appropriate greetings, everybody. Hello, Virtual Log. 00:00:55.971 --> 00:00:56.711 Hello, there. 00:00:58.131 --> 00:00:58.471 Hello, Brent. 00:00:58.471 --> 00:01:00.391 Hello. A strong crew today. That's important. 00:01:00.391 --> 00:01:03.400 I'm going to pull on them. I'm going to get through it. I'm going to get through it. 00:01:03.778 --> 00:01:07.241 The other thing that's going to help me is my friends over at Defined Networking. 00:01:07.241 --> 00:01:11.498 Go to define.net slash unplugged and meet Managed Nebula from Defined Networking. 00:01:11.945 --> 00:01:15.231 It is the easiest way to build a fast, resilient private network on the open 00:01:15.231 --> 00:01:16.288 source Nebula platform. 00:01:16.909 --> 00:01:22.001 It is rock solid infrastructure. And now every new network includes a free managed 00:01:22.001 --> 00:01:24.287 lighthouse. So you could just get connected. Your first machines, 00:01:24.560 --> 00:01:27.485 I think they're using IPv6 too, real easy, smooth. 00:01:27.776 --> 00:01:31.811 And of course, you can still run your own lighthouses to have your own control 00:01:31.811 --> 00:01:33.732 too. So you can kind of take advantage of theirs if you want, 00:01:33.946 --> 00:01:36.367 use only yours or do a mix of both. 00:01:36.808 --> 00:01:40.371 That's really powerful. Owning your own infrastructure at the network level 00:01:40.371 --> 00:01:44.715 is ultimately important and powerful. So you can get started with 100 hosts for free. 00:01:45.104 --> 00:01:49.277 This is not like anybody else. No credit card required. Go to define.net slash unplugged. 00:01:50.194 --> 00:01:54.641 Check out a totally different way to do Mesh VPN. Go to define.net slash unplugged. 00:01:55.030 --> 00:01:58.355 And a big thank you to Define for sponsoring the Unplugged program. 00:02:00.409 --> 00:02:04.763 Speaking of sponsors, we probably should have addressed this earlier, 00:02:04.763 --> 00:02:08.463 maybe a week or two ago, but this is sort of – we don't really think often to 00:02:08.463 --> 00:02:09.993 talk about the mechanics of the show. 00:02:10.516 --> 00:02:13.070 But then we publish it and we go, oh, we should probably talk about this. 00:02:13.575 --> 00:02:19.217 So people that have been listening for the last few months know that we're experimenting with dynamic ads. 00:02:19.960 --> 00:02:24.683 And we're kind of doing that in place of when we don't have an ad that's been 00:02:24.683 --> 00:02:27.890 directly sold, like a deal where I'm coming on and I'm doing the host read. 00:02:28.575 --> 00:02:34.003 They don't, you know, my preference is the host read ads, but the ad world has 00:02:34.003 --> 00:02:38.593 changed. And dynamic is the way it goes because the buyers prefer it that way. 00:02:38.593 --> 00:02:39.564 The customers prefer it. 00:02:40.295 --> 00:02:45.643 So what we're going to experiment with is shortening the pre-show or like the 00:02:45.643 --> 00:02:48.703 pre-roll that you hear if you get a dynamic ad. We're going to, 00:02:48.703 --> 00:02:52.114 I think we're going to, we haven't done this yet, but I think we should shorten that to 60 seconds. 00:02:52.915 --> 00:02:57.833 And then, Roger, in the mid-roll ads, it'll be like ads that are in the show. Thank you. 00:02:59.327 --> 00:03:02.432 Those will be about 90 seconds because that's what the typical length is. 00:03:02.432 --> 00:03:06.404 And then there'll be ads at the end of the show. And it does help the show if you listen to those. 00:03:07.344 --> 00:03:10.682 However, if you are tired of ads, we do have the member version, 00:03:10.792 --> 00:03:14.589 which offers the same fully produced show by Drew, but with just no ads. 00:03:14.850 --> 00:03:19.002 And you also get the bootleg if you like, which typically doesn't have ads or 00:03:19.002 --> 00:03:20.365 it has like a shortened version of them. 00:03:20.481 --> 00:03:24.502 And we're going to try to dial this in. We do like your feedback on this. 00:03:24.502 --> 00:03:26.392 We're paying attention. And it's new. 00:03:26.992 --> 00:03:28.192 Yeah, very new for us as well. 00:03:28.192 --> 00:03:33.832 I resisted this for a very long time, so I'm now paying the tax of not really 00:03:33.832 --> 00:03:37.362 paying attention to this, and I'm learning on the go, on the fly. 00:03:38.652 --> 00:03:42.362 We have been really kind of tight, really since the beginning of the year, 00:03:42.362 --> 00:03:45.372 tighter than we really even talk about publicly. And so we're trying to find 00:03:45.372 --> 00:03:46.812 the right balance here to make this work. 00:03:47.584 --> 00:03:51.562 We're going to adjust some of it, and we encourage your feedback. 00:03:51.562 --> 00:03:53.516 We encourage you to listen because it does help the show. 00:03:53.934 --> 00:03:58.362 Or sign up for a membership if you're sick and tired of it. And if you have 00:03:58.362 --> 00:04:00.546 any questions, shoot it in and we'll try to answer. 00:04:01.393 --> 00:04:03.372 But we just wanted to mention it. We probably should have mentioned it a couple 00:04:03.372 --> 00:04:05.358 of weeks ago as we began experimenting more with this. 00:04:06.612 --> 00:04:09.682 And it's a learning process. So thank you, everybody. We really haven't had 00:04:09.682 --> 00:04:11.762 any complaints or anything like that, but we just wanted to talk about it. 00:04:11.762 --> 00:04:14.147 Now, before it became like an elephant in the room. 00:04:15.032 --> 00:04:17.812 And something that's exciting to see, because I'm hoping we're going to make 00:04:17.812 --> 00:04:21.102 it, is the Texas Linux Fest call for papers have reopened. 00:04:21.102 --> 00:04:24.922 Yeah, that's right. Just a short little window here. It'll close on July 22nd. 00:04:24.922 --> 00:04:25.304 Oh, yeah. 00:04:25.952 --> 00:04:30.342 But if you missed the first one, you got a shot now. They have also released 00:04:30.342 --> 00:04:34.162 the first round of approved talks, too. So you can kind of see what is already 00:04:34.162 --> 00:04:35.677 there if you need to calibrate off that. 00:04:35.968 --> 00:04:38.615 And specifically, they're looking for things in observability, 00:04:38.771 --> 00:04:41.982 automation, infrastructure, databases, or software development. 00:04:41.982 --> 00:04:44.602 That's pretty great. I'm hoping we can make it. We don't really have all our 00:04:44.602 --> 00:04:46.352 plans in yet, but we always try to make that one. 00:04:46.352 --> 00:04:48.483 Yeah, I believe it's at the beginning of November this year. 00:04:48.652 --> 00:04:50.092 Okay. Okay. All right. 00:04:50.092 --> 00:04:52.859 So get that on your calendars, and we'll try to join you. 00:04:56.302 --> 00:05:01.051 This week, Linus made news with a bomb he dropped on the Linux kernel mailing list. 00:05:01.051 --> 00:05:04.781 And I think to really kind of wrap our head around this, we should go back a 00:05:04.781 --> 00:05:07.061 couple of weeks, because what we're seeing right now are the headlines and the 00:05:07.061 --> 00:05:11.935 takeaways from this one post in the mailing list that has a lot of other stuff in it, too. 00:05:12.458 --> 00:05:17.264 But Linus has recently been on a bit of a world tour, actually, 00:05:17.723 --> 00:05:22.042 sitting down with Dirk and talking about his thoughts on AI. 00:05:22.518 --> 00:05:29.246 And what we have is a couple of series of clips that are over the last about 60 days. 00:05:29.640 --> 00:05:33.321 And you can actually hear Linus's thoughts evolve in that 60 days. 00:05:33.785 --> 00:05:39.041 And this is about 60 days, the timeline before this big bomb that was dropped 00:05:39.041 --> 00:05:41.111 on the mailing list that everybody's talking about this week. 00:05:41.575 --> 00:05:46.321 So we want to go back to that because it started – it really started with Linus 00:05:46.321 --> 00:05:50.321 being a lot more kind of contemplative on the pain points of AI and all this. 00:05:50.321 --> 00:05:53.027 So he sits down with Dirk Handel often at these open source summits, 00:05:53.463 --> 00:05:56.181 and there was just a couple that happened recently. 00:05:56.181 --> 00:05:59.141 And some of these clips our members have already heard in the bootleg because 00:05:59.141 --> 00:06:01.961 we played this a few weeks for them, but now it's relevant to this week's episode. 00:06:01.961 --> 00:06:06.565 Plus, there's been new stuff since then. So we're going to catch you up on all of it. 00:06:06.959 --> 00:06:11.691 So let's start with this. Linus starts with what he thinks right now is the 00:06:11.691 --> 00:06:15.818 most painful aspect of AI development. 00:06:16.869 --> 00:06:21.171 Again, I have a hard time seeing everybody because I have the lights in my face. 00:06:21.171 --> 00:06:25.431 Very bright. But how many people here use actually AI for coding? 00:06:27.236 --> 00:06:33.935 Yeah, so I have a love-hate relationship with AI. 00:06:34.080 --> 00:06:36.880 I actually really like it from a technical angle. 00:06:37.220 --> 00:06:42.051 I love the tools. I find it very useful and interesting. 00:06:42.445 --> 00:06:49.400 But it is definitely causing pain points. And the big pain points. 00:06:51.557 --> 00:06:55.162 In Linux traditionally, and I suspect in most projects, 00:06:55.916 --> 00:07:02.920 have not been so much the code itself, but has been when you're forced to change 00:07:02.920 --> 00:07:06.810 how you work and people get into a rut. 00:07:06.810 --> 00:07:13.030 People know this is how I work and this is how I like working. 00:07:14.376 --> 00:07:19.270 And then something comes along and for me it was like a long time ago, 00:07:19.270 --> 00:07:26.330 around 2000, I had to change how I worked because I did not scale anymore for 00:07:26.330 --> 00:07:28.819 the project as Linux was growing. 00:07:29.313 --> 00:07:35.080 And I still remember it being one of the more painful episodes in kernel development, 00:07:35.080 --> 00:07:37.399 and it was literally 25 years ago. 00:07:38.879 --> 00:07:42.450 And I think we're seeing some of the same effects now with AI, 00:07:42.450 --> 00:07:48.555 where it kind of forces people to get out of their comfort zone. 00:07:49.217 --> 00:07:57.773 so we've had a lot of social issues with the ai generated bug reports but also code and and there's, 00:07:59.846 --> 00:08:06.320 uh it's not a linus doesn't scale like it was 25 years ago now it's people don't 00:08:06.320 --> 00:08:10.573 necessarily scale and it takes a while to learn how to use the tools in a. 00:08:13.422 --> 00:08:20.082 Maybe responsible isn't the right word, but in a way that actually works with 00:08:20.082 --> 00:08:23.662 the community and works with the other developers. 00:08:24.069 --> 00:08:26.692 And we are definitely seeing some of the pain. 00:08:27.552 --> 00:08:32.472 One of the pain points recently has been the security mailing list where when 00:08:32.472 --> 00:08:40.131 people find bugs with AI, a lot of the time, I mean, a lot of bugs can be security issues. 00:08:40.952 --> 00:08:46.572 And I've said for years, decades, that a security bug is a bug, 00:08:46.872 --> 00:08:50.182 but the other side of that is a bug is often a security bug, 00:08:50.182 --> 00:08:51.786 and you may not even realize it. 00:08:53.232 --> 00:08:59.602 And people think that when they find a bug with AI, the first reaction seems 00:08:59.602 --> 00:09:05.434 to sometimes be, let's send it to the security list because this may have security implications. 00:09:06.055 --> 00:09:12.017 And the security list for the kernel was overrun by duplicate reports, 00:09:12.580 --> 00:09:18.396 by tens, hundreds of people running AI, often the same AI tool, 00:09:18.977 --> 00:09:20.898 and getting slightly different results, 00:09:21.653 --> 00:09:23.792 often with the same underlying cause. 00:09:23.792 --> 00:09:27.771 but we were flooded by people sending bugs. 00:09:28.079 --> 00:09:34.198 And then you have this list with fairly few people on it because it's all supposed to be super secret. 00:09:35.612 --> 00:09:39.428 And we spent all our time just forwarding these reports to. 00:09:41.599 --> 00:09:47.642 Other developers who knew that area better. And we made the policy change that 00:09:47.642 --> 00:09:55.692 basically if you find a security or any bug with AI, you should basically consider it to be public. And... 00:09:57.639 --> 00:10:02.677 Just because if you found it with AI, a hundred other people also found it with AI. 00:10:03.759 --> 00:10:08.522 And obviously that when it comes to things that really are security issues, 00:10:08.998 --> 00:10:13.236 you may not want to make the exploit public, right? 00:10:13.695 --> 00:10:18.319 Don't go and say to the world, hey, this is how, and this, by the way, 00:10:18.319 --> 00:10:21.427 I want to make it very clear. This is not just about the Linux kernel. 00:10:21.959 --> 00:10:25.003 When you find an exploit for any piece of software, 00:10:26.129 --> 00:10:32.159 don't be that guy or that gal who then crows about it publicly and say, 00:10:32.159 --> 00:10:37.199 look, hey, I could bring down this big company or whatever. 00:10:38.621 --> 00:10:43.739 Just let people know what the problem was and don't necessarily tell them exactly 00:10:43.739 --> 00:10:49.156 how to make somebody's life really miserable on a Friday afternoon. Right. 00:10:50.143 --> 00:10:56.799 So we are seeing some pain points with AI. And it's not that the technology 00:10:56.799 --> 00:11:00.766 is bad. And I personally really enjoy it. 00:11:01.283 --> 00:11:08.359 But there's like the social issues that it causes have yet to be all worked out, I would say. 00:11:08.968 --> 00:11:11.379 The conversation kind of goes on, and they talk a bit more about some of the 00:11:11.379 --> 00:11:15.662 tooling. In a more recent conversation that happened after that, 00:11:16.559 --> 00:11:21.269 Linus touches on the—he expands a little bit more on some of the pain points 00:11:21.269 --> 00:11:22.958 they've had with LLM development. 00:11:23.318 --> 00:11:28.209 I want to point out for the record, we made it 26 minutes into this conversation 00:11:28.209 --> 00:11:30.418 without talking about AI and LLMs. 00:11:36.716 --> 00:11:42.165 But I think this will be the only talk in this conference that gets to that. 00:11:42.165 --> 00:11:46.326 But of course, we should talk about this because we did the same session or, 00:11:46.825 --> 00:11:50.505 totally different, but same setup session a month ago in Minneapolis. 00:11:50.505 --> 00:11:54.735 And one of the things that a lot of people picked up on was that on stage you 00:11:54.735 --> 00:11:59.172 said that you see LNM as just a tool, which I completely agree with. 00:11:59.474 --> 00:12:07.352 And you made this interesting statement that you say it offers an incremental efficiency of a 10x. 00:12:08.019 --> 00:12:14.706 And a lot of people said, oh, Linus thinks that LLMs make everything 10x better. And so I was curious. 00:12:15.905 --> 00:12:21.899 That 10x was not scientific. Right. That was pulled out of my ass number, obviously. 00:12:22.601 --> 00:12:28.995 Just to clarify. That was two questions down. But OK, so this was a intuitive 00:12:28.995 --> 00:12:31.686 number that you came up with to rephrase what you just said. 00:12:32.104 --> 00:12:40.585 But so how much of this true added value that LLMs provide? 00:12:40.585 --> 00:12:44.304 We just talked about this. The core code gets more eyeballs, 00:12:44.832 --> 00:12:49.225 but the LLMs can apply the same scrutiny, the same algorithm, 00:12:49.225 --> 00:12:52.559 the same logic testing ability to all of the code. 00:12:52.884 --> 00:12:58.225 How much do you think that the LLMs really create productivity advancement for, 00:12:58.225 --> 00:13:00.925 not you personally, but for the Linux kernel community? 00:13:02.235 --> 00:13:04.095 So right now we're at kind of 00:13:04.095 --> 00:13:09.765 the point where hopefully it creates more productivity than it takes away. 00:13:12.085 --> 00:13:23.319 We certainly saw more junk being generated by LLMs than we saw useful code up until, 00:13:24.946 --> 00:13:30.780 the like early this year basically and even now uh, 00:13:32.486 --> 00:13:37.141 it's taking a lot of resources from developers because, 00:13:38.424 --> 00:13:43.425 you get these bug reports i mean i don't know how many of you have seen this 00:13:43.425 --> 00:13:49.385 pattern in your own work uh i bet it happens or i know for a fact that it happens 00:13:49.385 --> 00:13:52.866 outside the kernel too where you get bug reports that, 00:13:53.993 --> 00:14:00.842 look entirely valid and it actually can take quite a lot of effort to figure out that, no, 00:14:02.299 --> 00:14:04.145 that was just hallucination. 00:14:06.317 --> 00:14:13.355 And it can actually be a huge drain on resources when it takes humans a lot 00:14:13.355 --> 00:14:17.845 of effort to figure out that, hey, this machine-generated report was not true. 00:14:18.491 --> 00:14:22.215 And so Linus goes on to say that he does think there is some long-term good, 00:14:22.655 --> 00:14:26.895 and he kind of articulated this in the previous chat as well. 00:14:26.895 --> 00:14:30.245 Don't get me wrong. This may sound negative to a lot of people, 00:14:30.245 --> 00:14:34.235 but I actually think it's all very good. 00:14:35.485 --> 00:14:39.340 When AI finds a bug in any source code, 00:14:41.383 --> 00:14:44.628 that is a short-term pain. 00:14:45.841 --> 00:14:49.609 But long-term is you found a bug. We fixed it. 00:14:50.989 --> 00:14:55.179 The end result is better for it. So I'm actually, this is part of why I actually 00:14:55.179 --> 00:15:01.069 really enjoy AI. And again, I want to stress that the conflict is not that AI is bad. 00:15:01.409 --> 00:15:05.829 The conflict is then that there are some social choke points and social pain 00:15:05.829 --> 00:15:12.409 points that come with this new tool that we have 35 years of code, 00:15:12.409 --> 00:15:15.853 that it sometimes finds issues that we did not find. 00:15:16.749 --> 00:15:23.742 and it's going to take us some time to just work through the new issues that are found. 00:15:23.986 --> 00:15:28.889 And so I'm actually very positive about this whole thing. I think finding bugs 00:15:28.889 --> 00:15:34.009 is great because the real problem is all the bugs you didn't find. 00:15:35.096 --> 00:15:38.519 I also picked up that you seem to note that there has been some improvements 00:15:38.519 --> 00:15:40.523 in some of the bigger, more important areas. 00:15:40.860 --> 00:15:45.159 At the same time, what has definitely happened in the last couple of months 00:15:45.159 --> 00:15:50.096 is that we are getting a lot of really good bug reports. 00:15:51.419 --> 00:15:55.459 Most of the good ones require more than just the LLM, though. 00:15:56.569 --> 00:16:04.469 They require a person that then goes through the report that was generated by 00:16:04.469 --> 00:16:07.853 the AI and makes sure it was valid. 00:16:08.161 --> 00:16:13.443 and then also acts as the go-between between developers who fix the problem. 00:16:14.128 --> 00:16:20.389 And we've had to push back quite a bit on people in the sense that if you find 00:16:20.389 --> 00:16:27.949 a bug with an LLM, it's not enough to just ask the LLM to make a bug report and then throw it over, 00:16:28.819 --> 00:16:31.972 the fence to us. We want to see... 00:16:33.342 --> 00:16:40.771 a suggested patch. We want to see the human who ran the LLM act as a kind of 00:16:40.771 --> 00:16:46.771 back and forth, that we have somebody who we can talk to and discuss the issues with. 00:16:46.771 --> 00:16:50.095 Because sometimes the patch is a simple one-liner that is obvious, 00:16:50.571 --> 00:16:52.899 but quite often it ends up being, hey, 00:16:53.392 --> 00:16:58.881 we didn't think about that issue and we will need to reorganize some code and 00:16:58.881 --> 00:17:02.425 fixing this properly is going to take a bit of effort. 00:17:03.510 --> 00:17:08.723 Then you really want to have the person who sent the original report involved in. 00:17:10.156 --> 00:17:15.022 The developer is coming back and saying, hey, this is a suggested fix. 00:17:15.571 --> 00:17:18.428 What does your LLM think about this? 00:17:18.562 --> 00:17:22.562 OK, I think that gives us kind of a more expanded view on Linus's thoughts around 00:17:22.562 --> 00:17:31.900 AI. So everything kind of shifted on July 14th in a Linux media mailing list discussion. 00:17:32.713 --> 00:17:37.122 And it wasn't like a like a discussion around a conventional release or anything like that. 00:17:37.122 --> 00:17:43.122 But it was, I guess, a thread that concerned connecting a patchwork of open 00:17:43.122 --> 00:17:47.732 source reviews, like a different set that had been like agentically reviewed, like a whole patch set. 00:17:47.732 --> 00:17:51.122 Yeah. So they're connecting some existing tools, right? 00:17:51.122 --> 00:17:53.922 So there's a bunch of tools in the kernel so that when you submit a patch, 00:17:53.922 --> 00:17:57.274 there's automated review of it, tests get run, things get built, 00:17:57.465 --> 00:17:58.642 and then you get feedback 00:17:59.002 --> 00:18:02.502 sent to you and also to the mailing list, maybe to the maintainer that says 00:18:02.502 --> 00:18:05.502 like, hey, the program found these things, you know, kind of like linting or 00:18:05.502 --> 00:18:06.802 running a basic set of unit tests. 00:18:06.802 --> 00:18:10.712 There's various versions of analyzing the software automatically to see if, 00:18:10.712 --> 00:18:14.212 like, it is, you know, before a human even has to get to it to try to lower 00:18:14.212 --> 00:18:15.052 the maintenance burden. 00:18:15.052 --> 00:18:15.382 Okay. 00:18:15.382 --> 00:18:19.212 Especially when you consider, like, Linux tree maintainers, their job is ingesting 00:18:19.212 --> 00:18:22.782 a giant stream of potential patches and trying to make sure that they're doing 00:18:22.782 --> 00:18:26.762 so in a well-done, rigorous way. And so, Sashiko. 00:18:27.902 --> 00:18:31.902 Sashiko, okay. Yeah. So, it's not a proposal. It's operational already. 00:18:31.902 --> 00:18:35.472 It's just this is more about more plumbing on the actual, you know, 00:18:35.472 --> 00:18:39.632 connecting it more to the, like, making it more permanent fixture of the ongoing 00:18:39.632 --> 00:18:42.572 development process. Right now it's been a tool sort of more on the side. 00:18:42.572 --> 00:18:44.402 And so this is a discussion around that. 00:18:44.402 --> 00:18:44.850 Yes. 00:18:45.106 --> 00:18:46.313 And Linus decides to chime in. 00:18:46.441 --> 00:18:47.822 Well, it starts... 00:18:50.017 --> 00:18:54.289 It starts because people are talking about, let's add this to the thing. And then someone... 00:18:55.224 --> 00:18:56.097 Objects. 00:18:56.097 --> 00:18:56.537 Yes. 00:18:56.537 --> 00:18:56.887 Yeah. 00:18:56.887 --> 00:19:03.527 So a developer comes in to, and we can get the dudes here, but yeah, 00:19:03.527 --> 00:19:08.482 a developer comes in to basically say like, shouldn't this be an opt-in thing? 00:19:08.715 --> 00:19:10.613 Isn't that the normal default of what we do? 00:19:10.868 --> 00:19:14.758 And then that gets into a whole discussion around, is that a viable model? 00:19:15.181 --> 00:19:19.030 and then ethical concerns get brought in and it becomes a question of, 00:19:19.750 --> 00:19:26.958 you know, if I don't want to be made, I don't want to be forced to use an AI tool. 00:19:27.289 --> 00:19:29.677 And then the whole debate expands around there. 00:19:29.677 --> 00:19:31.277 Yeah. And then that's when Linus chimes in. 00:19:31.597 --> 00:19:31.837 Yes. 00:19:32.717 --> 00:19:37.077 Linus writes, I realize some people really dislike AI, but this is an area where 00:19:37.077 --> 00:19:40.705 I'm willing to absolutely put my foot down as the top level maintainer. 00:19:41.355 --> 00:19:44.188 Linux is not one of those anti-AI projects. 00:19:45.047 --> 00:19:48.457 And if somebody has issues with that, they can do the open source thing and 00:19:48.457 --> 00:19:50.074 fork it or just walk away. 00:19:50.753 --> 00:19:55.449 AI is a tool and like other tools we use, and it's clearly a useful one. 00:19:55.850 --> 00:20:01.597 It may not have been that clearly even just a year ago, but that no longer is in question today. 00:20:02.328 --> 00:20:06.769 There are other questions around AI, like what the economy of it will actually look like in the end. 00:20:07.216 --> 00:20:11.207 But, quote, is it still useful? That's no longer one of those questions. 00:20:11.207 --> 00:20:14.699 Anybody who doubts that clearly hasn't actually used it. 00:20:15.279 --> 00:20:19.567 Yes, it can also be somewhat of a painful tool, both for maintainer workloads 00:20:19.567 --> 00:20:22.158 and from it just keeps finding embarrassing bugs standpoint. 00:20:22.667 --> 00:20:25.937 But the solution is not to put your head in the sand and sing la la la la la 00:20:26.257 --> 00:20:29.894 I can't hear you at the top of your voice like some people seem to do. 00:20:30.817 --> 00:20:34.555 The solution is to make sure those LLM tools help maintainers, 00:20:34.985 --> 00:20:39.101 that's his italics, help maintainers instead of just causing them pain. 00:20:39.565 --> 00:20:41.591 There is no question on that side. 00:20:42.479 --> 00:20:46.573 We're not forcing anyone to use it, but I will very loudly ignore people who 00:20:46.573 --> 00:20:48.540 try to argue against other people using it. 00:20:48.894 --> 00:20:53.323 And no, AI isn't perfect, but Christ, anybody who points to the problems with 00:20:53.323 --> 00:20:57.363 AI had better be looking in the mirror and pointing at themselves at the same 00:20:57.363 --> 00:21:01.963 time, because it's not like the natural intelligence is always all that great either. 00:21:03.708 --> 00:21:09.683 The kernel project has been and will continue to be about the technology sure 00:21:09.683 --> 00:21:13.893 the social angle of working on open source is important and often very much 00:21:13.893 --> 00:21:15.341 a motivating part of the project, 00:21:16.102 --> 00:21:22.033 but in the end it is a side benefit not the point of the project again his emphasis 00:21:22.033 --> 00:21:23.490 not the point of the project, 00:21:24.274 --> 00:21:28.099 this is not his emphasis, some kind of social warrior project. 00:21:28.523 --> 00:21:30.247 Never has been, never will be. 00:21:30.770 --> 00:21:34.996 In the kernel community, we do open source because it results in better technology, 00:21:35.379 --> 00:21:36.888 not because of religious reasons. 00:21:37.265 --> 00:21:42.443 And so we make decisions primarily based on technical merit, not fear of new tools. 00:21:43.884 --> 00:21:47.568 That is a pretty strong, clear stance. What's your reaction to that, 00:21:47.568 --> 00:21:50.960 Brent? Is that stronger than you expected after hearing the comments from Linus? 00:21:51.727 --> 00:21:54.608 Well, I think it's more clearly laid out here. 00:21:54.608 --> 00:22:00.660 Like I found in his clips on stage, he was a bit more exploring with the crowd, 00:22:00.916 --> 00:22:02.378 like saying, oh, yeah, yeah, we're using it. 00:22:02.378 --> 00:22:07.098 These are the experiences the project's having. But here he clearly states both 00:22:07.098 --> 00:22:14.528 his personal kind of view, but also just a very clear stance from the project. 00:22:14.528 --> 00:22:19.601 And I think a stance that you haven't seen very often in other projects, 00:22:20.170 --> 00:22:25.768 especially not written in Linus-like words as he does. 00:22:26.433 --> 00:22:32.108 But the thing that really hit me here is one of these last sentences in the 00:22:32.108 --> 00:22:35.868 kernel community, we do open source because it results in better technology, 00:22:35.868 --> 00:22:38.270 not because of religious reasons. 00:22:39.024 --> 00:22:45.168 And we've seen a couple of projects in the last few years fall victim, 00:22:45.967 --> 00:22:51.878 maybe that's a strong word, but fall victim to some infighting in the community 00:22:51.878 --> 00:22:56.695 just because of differences in philosophy of how the project should be run. 00:22:57.826 --> 00:23:02.175 So in a way, having Linus say, yes, this is exactly how it is, 00:23:02.175 --> 00:23:08.695 and this is how the project believes this should be implemented is a very clear 00:23:08.695 --> 00:23:11.645 voice about how everyone should be following. 00:23:11.645 --> 00:23:16.019 And there's not much room for pushing back, I think. 00:23:16.958 --> 00:23:21.035 Yeah, we've seen Linus take – I don't know what the right – but we've seen Linus 00:23:21.035 --> 00:23:27.627 take a technical over religion argument when it came to GPL2 versus GPL3. 00:23:28.179 --> 00:23:30.420 It's been a consistent position for him for a long time. 00:23:30.774 --> 00:23:35.385 Wes what is your reaction here to really framing this as look the usefulness 00:23:35.385 --> 00:23:39.742 is proved if you don't think these things are useful now then you're just clearly not using them um. 00:23:40.195 --> 00:23:43.965 It came up again in the thread some folks pushed back on or you know we're saying 00:23:43.965 --> 00:23:47.505 like you know we shouldn't compare these things you know these existing tools 00:23:47.505 --> 00:23:50.865 we have are deterministic you can debug it these llms are a black box and they 00:23:50.865 --> 00:23:52.995 can hallucinate and then linus agreed 00:23:53.315 --> 00:23:56.049 that they were different but he kind of flips it and he says 00:23:56.374 --> 00:24:01.337 you are correct the ai generate reports are usually much superior and more flexible 00:24:01.761 --> 00:24:04.715 they actually find real non-trivial problems, 00:24:05.157 --> 00:24:09.525 yeah so there's sort of even more um i will say it's worth noting that like 00:24:09.525 --> 00:24:13.376 this is a big thread with lots of maintainers it's not just these folks right so there's um, 00:24:13.887 --> 00:24:18.495 like there's uh roman gushkin who's a google kernel engineer who built shishiko 00:24:18.835 --> 00:24:22.125 and was sort of you know working on it lauren pinchart who sort of started the 00:24:22.125 --> 00:24:26.355 objection part of this a linux media contributor um and then linus but ted you, 00:24:26.875 --> 00:24:29.925 know a whole bunch of regular kernel contributors are in there and that's where 00:24:29.925 --> 00:24:33.500 it's becoming you know like they've been talking about around does, 00:24:34.040 --> 00:24:40.185 it does this particular tool work even about a 20 false positive rate 80 to 00:24:40.185 --> 00:24:45.315 90 percent true positives on real patches and only two maintainers seemingly 00:24:45.315 --> 00:24:49.015 in the whole thread reported real output problems everyone else seems to, 00:24:49.535 --> 00:24:53.435 is using it and likes it ted steven rostet John Cara, Bottomley, 00:24:53.435 --> 00:24:59.042 Bonzini, they all say it catches real bugs and that the noise is manageable. So we're seeing... 00:25:00.489 --> 00:25:03.578 It's important to remember that the kernel is like Linus and then a whole bunch 00:25:03.578 --> 00:25:05.968 of lieutenant tree maintainers. 00:25:05.968 --> 00:25:06.188 Right? 00:25:06.188 --> 00:25:10.083 And so what's happening is different maintainers are already enabling this and 00:25:10.402 --> 00:25:14.088 thinking that they may have different rules around do you have to use the tool 00:25:14.088 --> 00:25:17.598 or not. But they are already seemingly to commit that this is worth it for them 00:25:17.598 --> 00:25:20.897 as a tool that they want to see used in the code they're maintaining. 00:25:21.779 --> 00:25:26.440 Yeah, that decision has been made. And so the debate becomes should we be using 00:25:26.557 --> 00:25:29.035 AI-assisted development in free software? 00:25:31.079 --> 00:25:34.178 And Linus had an analogy. He had a vegetarian analogy. He says, 00:25:34.178 --> 00:25:37.968 being vegan because you think meat is murder is fine, but vegans don't expect 00:25:37.968 --> 00:25:40.622 the rest of the kernel community to stop eating meat. 00:25:41.237 --> 00:25:44.610 I think it comes down to, like, I can understand the position of, 00:25:44.738 --> 00:25:47.478 I don't want to be forced to use this tool that I disagree with, right? 00:25:47.478 --> 00:25:48.888 Yeah, and it's not mandatory. 00:25:49.004 --> 00:25:53.978 And so that's where what use, I think, matters a lot in the structure of what 00:25:53.978 --> 00:25:54.888 is happening here, right? 00:25:54.888 --> 00:25:58.738 So it's like, I think folks can have sympathy for your boss makes you use a tool. 00:25:58.738 --> 00:26:00.958 and you're supposed to hit a velocity target and you've got to ship, 00:26:00.958 --> 00:26:03.918 ship, you know, there's all kinds of stupid stuff that is just bullshit. 00:26:03.918 --> 00:26:07.468 But in this case, that's not what's happening, right? 00:26:07.468 --> 00:26:11.108 Like what's happening is an automated tool that is open source. 00:26:11.108 --> 00:26:13.608 I don't know what inference provider they're using or whatever, 00:26:13.608 --> 00:26:16.288 but it's a Rust tool, it's open source. All of the prompts are there that, 00:26:16.288 --> 00:26:19.944 you know, the normal open source process sort of applies there. 00:26:20.048 --> 00:26:24.428 And then it's finding to, at least by the people that are going to ultimately 00:26:24.428 --> 00:26:28.831 decide if your code is accepted, real bugs that humans don't find our quality 00:26:29.104 --> 00:26:33.667 and then it's sending it to you. So it's not you have to use it to make your code. 00:26:33.893 --> 00:26:33.978 Right. 00:26:34.038 --> 00:26:38.334 It's not you have to have a certain tool on your system that you authored the commit with. It's, 00:26:39.083 --> 00:26:44.479 you may be forced by the maintainer to respond to the output of this tool. 00:26:44.479 --> 00:26:44.957 Yes. 00:26:45.283 --> 00:26:48.609 And then where that gets tricky is you can imagine the case where, 00:26:48.609 --> 00:26:53.699 like, I don't want to have to sort through this slop that has stuff that's false 00:26:53.699 --> 00:26:55.249 positives, and I can understand that. 00:26:55.714 --> 00:27:00.489 But the flip side is where it gets harder, because what happens if the tool 00:27:00.489 --> 00:27:03.429 finds a real bug that you now refuse to engage with? 00:27:03.429 --> 00:27:07.945 Which is going to become more and more likely as they improve, as they get better. 00:27:08.125 --> 00:27:11.868 Which I think connects to Linus's point. Firstly, I think folks, 00:27:12.060 --> 00:27:17.239 like, there is very much a sort of a moral and a social and an altruistic sense 00:27:17.239 --> 00:27:20.703 to open source. I don't want to negate that at all. That's some of the stuff that drew me to it. 00:27:21.069 --> 00:27:25.769 But there is also this technical side, but it's also sort of almost a game theory 00:27:25.769 --> 00:27:29.349 side, right? It's like a strategic bet on how to build software that you yourself 00:27:29.349 --> 00:27:33.196 need, right? Like Google and Microsoft are not altruistic entities. 00:27:33.579 --> 00:27:36.011 they're doing it because there's a strategic thing and that's sort of a bet, 00:27:36.487 --> 00:27:41.539 and that ties into sort of it's also a methodology to develop software as much 00:27:41.539 --> 00:27:43.949 as it is a philosophy around software and that is the side i think linux is trying, 00:27:44.749 --> 00:27:49.984 or linus is trying to uh articulate and then he's also saying that, 00:27:50.785 --> 00:27:54.779 we have methods to adjudicate things like you know the kernel community does 00:27:54.779 --> 00:27:58.389 have a code of conduct now so we can sort of talk about intrapersonal conflict 00:27:58.726 --> 00:28:01.829 things like that and of course there's the there's a whole the whole 00:28:02.209 --> 00:28:06.197 system of maintainership is set up to adjudicate code quality on that sense 00:28:06.487 --> 00:28:10.329 but there is no system in place to evaluate individual ethical claims, 00:28:10.962 --> 00:28:13.919 once once that's in there like how do you that's an individual thing there is 00:28:13.919 --> 00:28:17.788 no system in place to do that there's kernels not set up to do that so we can't right, 00:28:18.485 --> 00:28:22.467 it that that's where a lot of this is coming from i think it's it's really a mechanism thing. 00:28:22.816 --> 00:28:29.352 Touching on that it Makes me remember, do you remember a bunch of Russian nationality, 00:28:30.124 --> 00:28:32.969 developers were booted off the project for... 00:28:34.809 --> 00:28:36.728 kind of shady reasons well. 00:28:36.728 --> 00:28:42.288 It was just as the sanctions or whatever went in during the ukraine war and yeah we were. 00:28:42.288 --> 00:28:47.518 Kind of arguing well maybe it's not the project's place to say who can and can't 00:28:47.518 --> 00:28:53.332 contribute uh because it should fall on more technical merits than it should on social ones. 00:28:53.971 --> 00:28:57.098 This is just another example of there's these sort of social lines yeah. 00:28:57.098 --> 00:28:59.388 And that one i think you could argue was partially imposed. 00:28:59.388 --> 00:29:04.124 On them right in a sense so there's some research out there that shows that 00:29:04.373 --> 00:29:08.708 going by numbers that are reported by developers i think it's interesting that 00:29:08.708 --> 00:29:13.079 there is sort of this false positive rate it's like an lm thinks it found a bug. 00:29:13.415 --> 00:29:17.458 There's also in this specific tool a case where because it sees patches sometimes 00:29:17.458 --> 00:29:21.948 it might find real bugs but in code that wasn't what you just added or touched so there's some noise. 00:29:21.948 --> 00:29:25.938 In that sense too yeah but it looks like like a lot of like the so the rate 00:29:25.938 --> 00:29:30.008 of is about a 20 false positive but what is interesting is it's like what they're 00:29:30.008 --> 00:29:34.788 developing like what the developers call a gray zone nitpicky small theoretical 00:29:34.788 --> 00:29:36.920 stuff that nobody really kind of tends to care about. 00:29:37.402 --> 00:29:40.438 And when you look at the stuff that's like the real important stuff, 00:29:40.844 --> 00:29:45.830 the hit rate is more like 80-90% of true positives for the actual critical things that get submitted. 00:29:46.150 --> 00:29:48.898 So an interesting thing is, if you look at what's coming in, 00:29:48.898 --> 00:29:51.130 according to self-reported numbers by these developers... 00:29:52.869 --> 00:29:56.449 The stuff that is important tends to be accurate. And the stupid stuff that's 00:29:56.449 --> 00:29:58.706 just like low-hanging stuff tends to be where the slop is at. 00:29:58.955 --> 00:30:01.289 The challenge is really separating the two, isn't it? 00:30:02.375 --> 00:30:04.899 And tuning it and seeing it. It'll be interesting because this will, 00:30:05.125 --> 00:30:09.792 in effect, provide some real data out there if you want to mine it to see how this changes over time. 00:30:10.059 --> 00:30:13.449 I did think there was some signal here in just seeing that, like, 00:30:13.449 --> 00:30:18.399 you know, Chris Mason, the ButterFS developer, has been contributing to some 00:30:18.399 --> 00:30:20.999 of the prompts that are getting used here. 00:30:20.999 --> 00:30:27.209 And then Jonathan Corbett at LWN took a look and called them some of the prompts 00:30:27.209 --> 00:30:32.339 that get fed into this 11-stage super—it's quite the process—called them some 00:30:32.339 --> 00:30:35.392 of the best documentation on kernel patch review ever written. 00:30:35.889 --> 00:30:37.987 He expressed a little, you know, sadness, 00:30:40.135 --> 00:30:43.529 around that we couldn't write it until we were writing it for a machine. 00:30:43.529 --> 00:30:47.059 Yeah. But seemed to think that there was a lot of really good stuff in there. 00:30:47.059 --> 00:30:51.949 Huh. That is funny how that works, huh? Here's the question I have for anyone 00:30:51.949 --> 00:30:55.750 in the audience who is on the side of we shouldn't be using AI in free software. 00:30:56.592 --> 00:31:00.589 I'd like to hear from you because I think, A, that's a point that should get voiced more, but B... 00:31:02.779 --> 00:31:08.785 I'd like to know where you stand on if an LLM-assisted tool finds a legitimate 00:31:08.785 --> 00:31:13.826 security issue, submits it to a project that is anti-AI, anti-LLM, 00:31:15.300 --> 00:31:16.215 what should that project do? 00:31:16.215 --> 00:31:19.835 Should they just ignore the security patch? Should they go try to recreate it 00:31:19.835 --> 00:31:23.387 from scratch? That's a ton of work. That's probably more work than going through their submission. 00:31:24.043 --> 00:31:27.655 So I'd like to know if you're out there and you're not really pro-AI use and 00:31:27.655 --> 00:31:29.205 free software, how would you 00:31:29.205 --> 00:31:32.437 solve that little particular problem that's obviously going to come up? 00:31:32.942 --> 00:31:37.335 What do you do there? What happens if a real bug is found, somebody sends in 00:31:37.335 --> 00:31:41.748 a patch that's AI-assisted or LLM-generated, and that project, 00:31:41.928 --> 00:31:44.070 do they refuse to engage with it? What should they do? 00:31:45.103 --> 00:31:47.895 I don't know what the options there are for them. Because that seems like that's 00:31:47.895 --> 00:31:52.725 going to be a thing is the projects that close the door to this, two things happen. 00:31:53.288 --> 00:31:56.965 Number one, they're going to get folks that are submitting bug reports that 00:31:56.965 --> 00:31:59.081 may be actually useful in some cases that they're going to ignore. 00:32:00.689 --> 00:32:03.245 but number two and i've talked about this before on the show and i'll just so 00:32:03.245 --> 00:32:05.977 i'll make a brief is they're not going to participate in the dialogue, 00:32:06.429 --> 00:32:12.455 and the state of free and open source ai is at a critical junction right there 00:32:12.455 --> 00:32:18.085 is and it could really use our community out there advocating for what we believe 00:32:18.085 --> 00:32:20.025 in and what we want to see and 00:32:20.286 --> 00:32:24.988 instead a lot of our community is like linus says putting their heads in the sand as a result they're. 00:32:26.614 --> 00:32:32.218 A lot of money from Anthropic and OpenAI and others has knives out for open 00:32:32.218 --> 00:32:36.708 source AI. And we're going to see a concentrated effort as the open source AI 00:32:36.708 --> 00:32:39.153 models catch up to the frontier models, quote unquote. 00:32:39.617 --> 00:32:43.988 They have such a financial position to protect that they're going to turn to 00:32:43.988 --> 00:32:45.794 lobbying efforts and things like that. 00:32:46.148 --> 00:32:50.333 And it's going to be like the 90s again with free software, but now in the intelligence field. 00:32:50.688 --> 00:32:53.399 And only this time, we don't have a Stallman out there. 00:32:53.700 --> 00:32:57.520 We don't have a free software foundation out there for AI. In fact, 00:32:57.636 --> 00:32:59.893 for the most case, they're all just ignoring and rejecting it. 00:33:00.125 --> 00:33:04.318 And so ironically, they're going to end up with worst-case scenario AI because 00:33:04.318 --> 00:33:06.191 they're not participating in the dialogue. 00:33:07.109 --> 00:33:09.953 And I've said it before, so I don't need to go on about it. 00:33:10.417 --> 00:33:16.860 But I hope some people have been realized or come to realize this since Linus made that post. 00:33:17.383 --> 00:33:20.808 Because I think what Linus is trying to at least say for the Linux kernel is 00:33:20.808 --> 00:33:22.984 the decision is made. these are useful. 00:33:23.531 --> 00:33:26.948 Now it's on you if you want to use it or not. But from a project standpoint, 00:33:26.948 --> 00:33:29.609 we're not debating the usefulness anymore. It's a tool. 00:33:29.963 --> 00:33:35.058 And our sovereign individual developers on their own prerogative are implementing 00:33:35.058 --> 00:33:37.178 and using these tools because they want to. 00:33:38.298 --> 00:33:43.975 And that is now the official state, the official position of the Linux kernel. 00:33:44.979 --> 00:33:48.588 You have to wonder if that does move some people and maybe get some people more 00:33:48.588 --> 00:33:51.208 actively involved in the free software side of ai. 00:33:51.626 --> 00:33:55.168 I do think i even saw in that mailing list read some folks saying you know like 00:33:55.168 --> 00:33:58.128 i had folks who had their own ethical concerns but also sort of were forced 00:33:58.128 --> 00:34:02.548 to reach a conclusion of like well at this point you know not using it is worse 00:34:02.548 --> 00:34:05.158 than using it even with those concerns it's. 00:34:05.158 --> 00:34:09.648 Like that that game theory scenario yeah i know it sucks it's the welcome to 00:34:09.648 --> 00:34:12.169 the world of technology is this is not the first time we've been here. 00:34:20.685 --> 00:34:24.714 Well, that's strange. The title of this little section says notes with Wes and Brent. 00:34:25.674 --> 00:34:26.334 Hey, I know that guy. 00:34:26.334 --> 00:34:31.504 And that's because it's been a, what would we say, multi-week journey of reassessing 00:34:31.504 --> 00:34:35.081 what notes mean to us in our life and times. 00:34:35.789 --> 00:34:40.381 And we touched on this a little bit in the members feed last week. 00:34:40.956 --> 00:34:46.104 And I think it went into a great bit more detail than we even expected. 00:34:46.104 --> 00:34:49.574 We were just like, I think Chris was like, hey, I heard you talking about notes 00:34:49.574 --> 00:34:54.330 recently. And then there went 10 minutes worth of really interesting discussion. 00:34:54.835 --> 00:34:59.184 So, but it's been another week. And Wes, you've been diving into notes. 00:34:59.184 --> 00:35:02.624 I've been diving even more into my notes. So where should we start? 00:35:02.624 --> 00:35:07.419 Should we give like maybe a detailed summary of what we were talking about last week? 00:35:08.621 --> 00:35:12.694 I wanted to start, if you guys don't mind, because I think the real story is 00:35:12.694 --> 00:35:15.094 why this all started was because you had a NextCloud disaster. 00:35:16.760 --> 00:35:21.572 And I also recently experienced a NextCloud disaster. I was viewing a very important text file. 00:35:22.129 --> 00:35:24.817 I clicked it just to view it in the text editor. I didn't edit, didn't save. 00:35:25.595 --> 00:35:30.308 And it wiped the file out just by viewing it. And then it immediately closed and saved that. 00:35:31.011 --> 00:35:35.854 And it was a very important file. I had it in backup, but just viewing it, 00:35:35.854 --> 00:35:40.404 wiped it out. It turned out to be like a little bug where basically... 00:35:40.414 --> 00:35:41.414 Some kind of race condition. 00:35:41.414 --> 00:35:45.864 A race condition, yeah. And wiped out one of my most important notes. 00:35:45.864 --> 00:35:48.704 And I realized right then and there, I'm not using NexCloud for notes anymore. 00:35:49.395 --> 00:35:53.054 but I want to know what happened to you. And I think cause that is what caused 00:35:53.054 --> 00:35:55.194 you to spiral into redoing your entire notes. 00:35:57.314 --> 00:36:02.113 Oh my gosh. Why do I do this? Uh, so I've been using next cloud. 00:36:02.368 --> 00:36:05.973 I looked at the calendar for about like something like 10 years now and notes, 00:36:06.606 --> 00:36:10.634 for that amount of time. And it's been the number one application. 00:36:10.634 --> 00:36:14.762 I would say definitely on next cloud that has been important for me over the years. 00:36:15.997 --> 00:36:23.242 But even more broadly, it's probably the ecosystem I use the most every single day. 00:36:24.113 --> 00:36:28.722 So if you took away a bunch of tools of mine, notes would be the one thing that 00:36:29.042 --> 00:36:32.492 is a capture of my knowledge, the place where I think about things, 00:36:32.492 --> 00:36:35.582 the place where I keep tricks and tips and stuff. 00:36:35.582 --> 00:36:39.763 So some people don't use notes as much as that, but I'm a note fiend. So, you know. 00:36:41.800 --> 00:36:51.882 Kind of like tabs. And what I have noticed happening recently is a slight erosion 00:36:51.882 --> 00:36:56.492 of my trust in the sinking process of my notes over the last... 00:36:57.862 --> 00:37:02.507 Many months, I would say. But it's just like keeps suggesting. 00:37:02.507 --> 00:37:05.267 So it wasn't as drastic as your experience, Chris. But for me, 00:37:05.267 --> 00:37:10.104 it's these little tiny paper cuts and these little micro warnings I kind of see them as. 00:37:10.604 --> 00:37:12.857 Because I would have like one note over here that would be like, 00:37:12.857 --> 00:37:16.827 oh, I thought I wrote a little detail there. And it's kind of missing on my laptop. 00:37:17.349 --> 00:37:21.777 And then I go to use my phone and, you know, I'm out in the world. 00:37:21.777 --> 00:37:24.997 And I want to refer to something that I left on there 30 minutes ago. 00:37:24.997 --> 00:37:29.467 And all of a sudden, it's not syncing. And it hasn't synced for many hours. 00:37:29.767 --> 00:37:34.707 And so the things I was relying on having with me at all times are just kind 00:37:34.707 --> 00:37:37.144 of not there. And so you got to find some other way to get it. 00:37:38.487 --> 00:37:45.607 So I had enough of those experiences that my otherwise deep trust in this note 00:37:45.607 --> 00:37:48.411 system, as PJ will definitely, 00:37:49.407 --> 00:37:53.047 back me up on, as I've been trying to convince him to use this thing forever 00:37:53.047 --> 00:37:56.641 because I trusted so much in it, has now been eroded to the point where I was like, well, 00:37:57.471 --> 00:38:00.931 I think I have to solve this a different way because I looked for, 00:38:01.610 --> 00:38:06.607 you know, was it a config set and was it some weird files I had in my infrastructure 00:38:06.607 --> 00:38:07.577 that are causing problems? 00:38:07.577 --> 00:38:11.437 Is it a bug I can report or something like that? And for months I've been looking 00:38:11.437 --> 00:38:20.847 and trying to try different methods and play around and I came to no clear explanation. 00:38:20.847 --> 00:38:22.982 So that's been a little frustrating. 00:38:22.982 --> 00:38:25.016 You were giving this a shot. you were trying not to move. 00:38:25.236 --> 00:38:29.852 Well 10 years if you've been using for 10 years and all of a sudden you know 00:38:30.061 --> 00:38:34.257 a key binding goes away and you just can't live without it so. 00:38:34.769 --> 00:38:39.046 This is after a long journey how you reluctantly ended up with obsidian. 00:38:40.905 --> 00:38:44.407 Don't give all the sauce away right and and i don't even know if that's the 00:38:44.407 --> 00:38:47.365 final landing point because i'm a little hesitant so. 00:38:47.592 --> 00:38:50.807 You started flirting with obsidian you're a little nervous because it's not 00:38:50.807 --> 00:38:53.343 free software but it is markdown based so you. 00:38:53.795 --> 00:38:55.554 And you're going to look at it. So, you know. 00:38:55.769 --> 00:38:56.680 Yeah. All right. 00:38:57.847 --> 00:39:00.088 I was basically looking for a couple... 00:39:01.563 --> 00:39:10.190 things maybe absolute needs one is markdown backed because i i can't do databases 00:39:10.190 --> 00:39:10.985 i don't know how those work, 00:39:12.770 --> 00:39:16.770 but also i'm using it on various devices right and i want to use kate on my 00:39:16.770 --> 00:39:20.880 desktop and i want to use something that's just like fairly fast on the phone 00:39:20.880 --> 00:39:23.930 so it has to have a phone right right phone fast, 00:39:24.511 --> 00:39:28.783 way to use it that that's super important for me maybe like 50 of the importance, 00:39:29.822 --> 00:39:31.710 but also like a really good search, 00:39:32.829 --> 00:39:37.580 that's surprisingly difficult to find in some of these apps so i scour the landscape 00:39:37.580 --> 00:39:41.400 trying to find you know something that's open source something that's quite 00:39:41.400 --> 00:39:47.470 mature it's still developed actively something that uh you know hits all these other points and um, 00:39:48.443 --> 00:39:52.370 once you put all those filters on and there's not much out there and turns out obsidian kind of falls, 00:39:53.517 --> 00:39:59.350 in that category quite nicely and i had hesitated to even look at it or even 00:39:59.350 --> 00:40:02.190 consider it meanwhile everybody else is raving about it so, 00:40:03.182 --> 00:40:04.527 I reluctantly installed it. 00:40:04.527 --> 00:40:05.219 Everybody's doing it. 00:40:05.219 --> 00:40:07.787 Yeah, I know. Everybody's doing it. Well, I, contrarian, you know, 00:40:07.787 --> 00:40:10.667 I don't want to be everybody, but it turns out. Well, they're kind of right. 00:40:10.867 --> 00:40:11.537 Like, it's got some great features. 00:40:11.747 --> 00:40:15.107 I mean, there's a reason why people like it. However, it did provoke feedback 00:40:15.107 --> 00:40:16.187 from listener Kai, right? 00:40:16.187 --> 00:40:20.087 Indeed. Dear Lupteam, and specifically, dear Brent. 00:40:21.267 --> 00:40:24.577 I'm a semi-long-term listener and core contributor. 00:40:24.577 --> 00:40:25.027 Thank you. 00:40:26.187 --> 00:40:32.062 I just finished episode 675, where you were considering LogSeek and Obsidian. 00:40:32.492 --> 00:40:35.897 But it sounds like you've missed that Logseek has moved away from Markdown files 00:40:35.897 --> 00:40:37.907 and are switching to a graph database approach. 00:40:38.703 --> 00:40:43.247 The Logseek Markdown version, OG as they call it, will only receive security updates in the future. 00:40:43.247 --> 00:40:44.183 Oh, wow. 00:40:44.374 --> 00:40:47.235 I am currently also reconsidering my note-taking application. 00:40:47.444 --> 00:40:48.971 Obsidian just doesn't feel right for me. 00:40:49.273 --> 00:40:53.156 And I'm currently taking a closer look at Silver Bullet. 00:40:53.626 --> 00:40:57.747 The beauty. It runs on your server in a centralized place, and you access it 00:40:57.747 --> 00:40:59.751 from your devices by connecting to the server. 00:41:00.018 --> 00:41:02.857 The files are cached and synced locally, so you can use it even when there is 00:41:02.857 --> 00:41:03.895 no internet connection. 00:41:04.255 --> 00:41:06.647 Though they are starting to work on local applications as well. 00:41:06.987 --> 00:41:10.317 I haven't tested it in depth yet, though I believe that you and I share the 00:41:10.317 --> 00:41:13.967 same taste. So maybe this is something you could check out as well. 00:41:13.967 --> 00:41:18.677 You know, Wes, I believe you and I came across Silver Bullet not too long ago. 00:41:18.677 --> 00:41:22.977 Yes, and are kind of looking around for various note-taking applications for show notes things. 00:41:22.977 --> 00:41:25.837 Yeah, for show notes. We were looking at this as a potential way to do our show notes. 00:41:25.837 --> 00:41:29.267 Yeah, so it's MIT-licensed, server-hosted, browser-based. Your data stays on 00:41:29.267 --> 00:41:31.827 your machine. It's plain Markdown files, live preview. 00:41:32.267 --> 00:41:33.402 I'm liking this. 00:41:33.710 --> 00:41:39.197 By-directional links, a built-in query language, SLIQ, and a Lua scripting environment 00:41:39.197 --> 00:41:41.897 that lets you program your notes, which is very cool. Run queries, 00:41:41.897 --> 00:41:44.964 build widgets, automate workflows, all from inside your pages. 00:41:45.225 --> 00:41:49.807 And the pages or the plugins run in sandboxed web workers, so you can have extensions, 00:41:50.407 --> 00:41:52.394 that can be secure without sacrificing power. 00:41:52.551 --> 00:41:53.044 Okay. 00:41:53.567 --> 00:41:54.223 All right. 00:41:54.449 --> 00:41:58.119 So, to me, it seems a little complicated, but... 00:41:59.018 --> 00:42:01.908 Maybe not when you actually look at what it's doing. Silver Bullet. 00:42:02.804 --> 00:42:05.864 By the way, silverbullet.md if you'd like to check it out, listener. 00:42:06.224 --> 00:42:06.942 Ah, that's clever. 00:42:07.667 --> 00:42:08.068 Yeah. 00:42:08.765 --> 00:42:13.544 I kind of felt it landed somewhere in the middle. Like, it's extremely powerful 00:42:13.544 --> 00:42:16.043 if you know what you're doing. You don't have to use those features. 00:42:17.624 --> 00:42:23.164 It has some nice easy wins, like bi-directional wiki links and those kind of 00:42:23.164 --> 00:42:28.034 things that I haven't been using that were kind of like on my mandatory list. 00:42:28.034 --> 00:42:29.204 even though that's not what I was using. 00:42:29.791 --> 00:42:31.370 Built-in to-do items is nice too. 00:42:31.752 --> 00:42:36.224 Yeah, it has all these well-thought-out quick access features. 00:42:37.364 --> 00:42:40.804 There's a nice video where one of the main developers goes through kind of a 00:42:40.804 --> 00:42:45.172 tour, and you can tell they're passionate about it, and there's some cool stuff. 00:42:46.507 --> 00:42:51.017 I thought of Wes right away with some of the power of using, 00:42:52.044 --> 00:42:57.437 scripting languages, Lua specifically, to just make your notes programmatic, 00:42:57.734 --> 00:43:01.849 and that seems like a wonderful thing for us nerds around here in this room. 00:43:01.954 --> 00:43:07.794 Yeah, it almost makes it closer to something like a Jupyter notebook or something, which is interesting. 00:43:08.397 --> 00:43:11.996 Reminded me a lot of text-based accounting. 00:43:13.006 --> 00:43:15.416 Mm-hmm. I like that comparison. 00:43:16.240 --> 00:43:18.594 Okay, so here's what I want to know, Brent, is what is it not... 00:43:18.594 --> 00:43:19.944 It doesn't sound like it worked for you. 00:43:19.944 --> 00:43:21.004 You're clearly not using it. 00:43:21.004 --> 00:43:21.164 Right? 00:43:21.164 --> 00:43:24.349 Yeah, there's like a big butt hanging here and I want to know what it is. 00:43:25.638 --> 00:43:31.884 Well the big one here is but it's only a web app like it runs in a browser I 00:43:31.884 --> 00:43:33.804 don't know if people like and you want a mobile. 00:43:33.804 --> 00:43:35.604 App you want a mobile app too. 00:43:35.604 --> 00:43:40.022 Yeah it's a big mobile app guy and you know is this the kind of thing people 00:43:40.306 --> 00:43:44.554 pin is this the kind of app you would pin you can pin. 00:43:44.554 --> 00:43:45.284 A tab on that. 00:43:45.284 --> 00:43:46.837 You'll have to teach me. 00:43:54.708 --> 00:43:59.017 Now, what I'm wondering is, are you willing to hear the LogSec sales pitch? 00:43:59.017 --> 00:44:01.287 Because I think, Wes, you've stuck with it, right? 00:44:01.287 --> 00:44:01.645 Oh, yeah. 00:44:01.994 --> 00:44:04.297 That's saying something, because it's been a couple of years. 00:44:04.297 --> 00:44:04.667 Yeah. 00:44:04.667 --> 00:44:05.389 A few years, yeah. 00:44:05.627 --> 00:44:10.137 I don't use it for, like, absolutely everything, but, yeah, I got some real data in there. 00:44:10.137 --> 00:44:14.117 So, Brent, I mean, could we be interested in putting you in a nice LogSec today? 00:44:14.917 --> 00:44:18.907 I also thought it would be a good, just because there is a big change going, 00:44:18.907 --> 00:44:20.617 I thought it was just a time to check in as well. 00:44:20.617 --> 00:44:20.977 Yeah. 00:44:20.977 --> 00:44:23.547 You know, it's not up front. It's not going to be for everyone. 00:44:23.547 --> 00:44:25.077 One, it's a very different model. 00:44:25.077 --> 00:44:27.977 Can you tell me about the change in your words and what you think about this 00:44:27.977 --> 00:44:29.117 database change they're doing? 00:44:29.117 --> 00:44:33.007 Yeah. So one way to think about it is it's more of almost like less of an upgrade 00:44:33.007 --> 00:44:34.207 and more of a different product. 00:44:34.207 --> 00:44:35.137 Yeah. It seems like it. 00:44:35.597 --> 00:44:37.597 It's a block-level graph database 00:44:37.597 --> 00:44:41.861 with typed properties. You get data log queries and an AI agent API. 00:44:42.134 --> 00:44:43.788 It also happens to be a note-taking app. 00:44:44.357 --> 00:44:48.057 Okay. I see. So they're pivoting for the agentic future. 00:44:48.057 --> 00:44:51.422 Well, I mean, in some sense, but this was sort of always an evolution because 00:44:51.793 --> 00:44:56.357 like in the Obsidian model, just as an example, but just sort of like you inherit 00:44:56.357 --> 00:44:58.457 a hierarchy from the operating system, right? 00:44:58.457 --> 00:44:58.724 Yeah. 00:44:58.991 --> 00:45:02.237 You're sort of like you have a folder tree, you have Markdown in there, 00:45:02.237 --> 00:45:03.627 you probably have it all organized. 00:45:03.627 --> 00:45:05.023 Yeah. It's all folders and whatnot. 00:45:05.255 --> 00:45:05.497 Yeah. 00:45:06.631 --> 00:45:08.465 Which I kind of, I do like to some degree. 00:45:08.628 --> 00:45:11.861 Yeah, that's not bad. I'm not saying anything's wrong with that. 00:45:12.012 --> 00:45:16.017 But this is a different way to think about how you might want to organize things. 00:45:16.397 --> 00:45:21.957 And so the old version of LogSeq sort of was a mix where it had markdown files 00:45:21.957 --> 00:45:27.057 on disk, but it was never a great fit for really what the best representation 00:45:27.057 --> 00:45:30.066 and use of what LogSeq was good for was, I thought. 00:45:30.269 --> 00:45:34.397 So now with the LogSeq DB version, the SQLite database, I would also argue I 00:45:34.397 --> 00:45:37.106 get why people don't like databases for storing stuff. 00:45:38.116 --> 00:45:42.977 if you are using agents I think that's perhaps less of a concern because they're 00:45:42.977 --> 00:45:46.037 pretty good at getting data in and out of things and they have no problem manipulating 00:45:46.037 --> 00:45:53.052 SQLite for you so or turning a SQLite dump back into structured text files or things like that but, 00:45:55.001 --> 00:45:59.671 you're no longer a prisoner of the file system. Pages are UUID-addressed entities 00:45:59.671 --> 00:46:01.521 with titles. No folders, no paths. 00:46:02.801 --> 00:46:06.841 Basically, the hierarchy is at the block level, including create parent-child, 00:46:06.841 --> 00:46:10.431 but then any other hierarchy you want, you're not limited to a tree. 00:46:10.431 --> 00:46:14.030 You build whatever graph you want with the graph primitives that are built right in. 00:46:15.087 --> 00:46:17.781 And then, of course, it does come with an MCP. It comes with a CLI. 00:46:18.501 --> 00:46:19.721 Of course, it has really nice search. 00:46:20.081 --> 00:46:24.881 Plus, it has the query language. It includes data log, lets you do stuff. 00:46:24.881 --> 00:46:28.461 So like if you want to write a query that isn't supported by however the search 00:46:28.461 --> 00:46:32.589 works as it stands now, you have full access to like the underlying data model. 00:46:32.960 --> 00:46:35.919 It does provide a read-only markdown mirror. 00:46:36.680 --> 00:46:37.251 Oh, that's interesting. 00:46:37.251 --> 00:46:39.811 So like you don't have to worry about that side of it. You can't edit those 00:46:39.811 --> 00:46:40.841 files and have it sync back. 00:46:40.841 --> 00:46:41.351 Oh, well, geez. 00:46:41.351 --> 00:46:43.001 But you can totally read. Yes. 00:46:43.001 --> 00:46:44.110 Yeah, you can view. Yeah. Yeah. 00:46:45.220 --> 00:46:45.958 It has a mobile app. 00:46:46.259 --> 00:46:49.524 Yep, mobile app that's new. It's all very new. So a lot of the biggest critiques 00:46:49.524 --> 00:46:53.594 right now is like, it's moving fast, it's a big change, and I think maybe the 00:46:53.594 --> 00:46:57.604 folks behind it haven't super articulated what I think is kind of the exciting 00:46:57.604 --> 00:46:59.864 vision for where it could be going. 00:46:59.864 --> 00:47:03.804 It seems notable that in May, they raised $4.1 million. 00:47:04.104 --> 00:47:04.514 Wow. 00:47:04.514 --> 00:47:04.805 Yeah. 00:47:05.311 --> 00:47:07.661 So I think they're going to be around for a minute working on this, 00:47:08.264 --> 00:47:10.953 and that knowledge graph seems to be the thing that got them the funding. 00:47:11.395 --> 00:47:15.784 It is AGPL. So you can pay them $5 a month if you want them to handle the sync, 00:47:16.304 --> 00:47:17.164 but you can host it yourself. 00:47:17.164 --> 00:47:18.004 That's pretty reasonable. 00:47:18.004 --> 00:47:22.046 And the new stuff, which I haven't tried this yet, but the new sync is real-time sync. 00:47:22.545 --> 00:47:29.031 Oh. Huh. That is really nice. It does come up when I'm moving between machines sometimes. 00:47:29.244 --> 00:47:32.294 And it is super new. So some folks with super giant libraries have had some 00:47:32.294 --> 00:47:35.350 issues importing them, but they does support importing from the previous version. 00:47:36.308 --> 00:47:36.662 Okay. 00:47:37.138 --> 00:47:39.784 Yeah. So I think it's sort of like, as we were talking about in the, 00:47:40.164 --> 00:47:42.194 in the pre pre the members show, we, 00:47:42.809 --> 00:47:47.535 some of the biggest, maybe one of the competitors to a personal knowledge system, 00:47:48.220 --> 00:47:53.494 is like the fact that if you have something like Hermes, it's building a thing 00:47:53.494 --> 00:47:56.524 that's like has Markdown files that it's indexing to search and injecting. 00:47:56.524 --> 00:48:00.634 and a lot of these agent memory systems or QMDs or whatever are sort of half 00:48:00.634 --> 00:48:04.044 of part of building structure on top of markdown. 00:48:04.604 --> 00:48:08.684 And so I think where this is interesting is LogSync presents another point on 00:48:08.684 --> 00:48:13.504 that continuum, especially like I was trying it out last night and the MCP is 00:48:13.504 --> 00:48:16.943 really nice. It's very well done. It's not perfect yet, but it has pretty much everything you need. 00:48:17.308 --> 00:48:21.169 And so you can read the files if you want. You can read it and do searches through the MCP. 00:48:21.511 --> 00:48:26.094 It's just that the main thing that changes is the right path goes through the database. 00:48:26.094 --> 00:48:26.347 Yeah. 00:48:27.489 --> 00:48:30.858 Which is, I mean, for a long-term tool, probably not a bad idea. 00:48:30.858 --> 00:48:31.418 More efficient. 00:48:31.418 --> 00:48:33.136 Also, I noticed they have a CLI as well. 00:48:33.357 --> 00:48:34.297 A CLI, yep. 00:48:34.907 --> 00:48:39.528 A Mac client and a Windows. Do they have a Windows? Yeah, they have a Windows client, I think. Okay. 00:48:39.528 --> 00:48:43.158 I got a couple of questions here, Wes, since you, over the years, 00:48:43.158 --> 00:48:48.299 seem to have really connected with this ethos, if I say that. 00:48:49.356 --> 00:48:54.148 What do you find is the perfect kind of information to store in this? 00:48:54.148 --> 00:48:57.368 Because I realized when we were having these discussions about like, 00:48:57.368 --> 00:49:01.198 well, what is notes even, that I was trying to store very different types of 00:49:01.198 --> 00:49:03.858 information in the same place to have it accessible in the same ways. 00:49:03.858 --> 00:49:09.668 And maybe that's not a very good idea. So perhaps there's a way to store the 00:49:09.668 --> 00:49:14.068 right kind of information in something like LogSeq and store the other kind 00:49:14.068 --> 00:49:15.778 of information somewhere else. It's more suitable. 00:49:15.778 --> 00:49:17.948 Oh, boy. That sounds complicated. 00:49:17.948 --> 00:49:20.938 I know. That's why I was leaning away from it. But Wes seems to think this is 00:49:20.938 --> 00:49:23.258 really good for something. So what is that? 00:49:23.258 --> 00:49:24.197 Why not everything? 00:49:25.074 --> 00:49:28.968 Well i think i tend to keep like some stuff like per project stuff tends to 00:49:28.968 --> 00:49:32.678 live in like a repo that i have for a project so i mean not that there's not 00:49:32.678 --> 00:49:35.738 notes or thoughts or ideas about that that might be in there but it tends to 00:49:35.738 --> 00:49:37.344 be that kind of stuff it's sort of 00:49:37.681 --> 00:49:39.728 it's ideas maybe things i want 00:49:39.728 --> 00:49:44.055 to investigate you know things i've learned about stuff notes for a future 00:49:44.444 --> 00:49:49.918 me going through or redoing something and i find that i what is really nice is that you can just, 00:49:50.394 --> 00:49:55.618 enter information and then start building adding links to it and you don't it's 00:49:55.618 --> 00:50:00.866 very easy to go navigate that so i don't worry as much about having a strict hierarchy because it 00:50:01.261 --> 00:50:04.878 was always felt like i had part of what limited me was classifying it up front 00:50:04.878 --> 00:50:07.498 like that and this means i don't really have to i can just 00:50:07.898 --> 00:50:10.888 put information in and sort of the structure emerges. 00:50:10.888 --> 00:50:14.078 Yeah i do sometimes spend a lot of time figuring out okay where does this live 00:50:14.078 --> 00:50:15.758 what yeah like classifying. 00:50:17.018 --> 00:50:21.638 But I wondered if there's some kind of upfront discipline every time you're 00:50:21.638 --> 00:50:24.578 making a note to try to tag it or tag a block. 00:50:25.558 --> 00:50:26.948 Can you explain a little bit of the... 00:50:28.009 --> 00:50:29.246 I don't know, daily workflow. 00:50:29.246 --> 00:50:31.646 Hygiene to make sure the notes actually link and stuff. 00:50:31.646 --> 00:50:33.576 What's the thing I need to learn to make this work well? 00:50:35.619 --> 00:50:38.756 Well, I think you just kind of play with it. 00:50:38.756 --> 00:50:43.336 Like, you could have an orphaned thing, but, you know, there's autocomplete 00:50:43.336 --> 00:50:47.196 on the tags. So as long as you have, as long as your variance in the tags that 00:50:47.196 --> 00:50:50.996 you do isn't so crazy, and then you can kind of go check, and especially now 00:50:50.996 --> 00:50:53.366 if you had an agent in there, you could go do some sort of a quick audit. 00:50:53.366 --> 00:50:57.246 Can I ask you about that, Brent? Sure. Is that is your intention to have an 00:50:57.246 --> 00:50:59.170 agent have access or edit these? 00:50:59.710 --> 00:51:03.656 Well, when we first started this discussion, it wasn't, I think, 00:51:03.656 --> 00:51:06.989 partly because I wanted to have some protected space, 00:51:07.576 --> 00:51:11.266 where this was like blessed information that I had, you know, 00:51:11.266 --> 00:51:15.592 cross reference a documentation or had used in anger and was like, 00:51:16.388 --> 00:51:20.323 the golden command or whatever it is that solved that problem. 00:51:20.863 --> 00:51:27.116 And I feared that letting an agent kind of have access to that would mix the 00:51:27.116 --> 00:51:29.580 information in a way that I felt was impure somehow. 00:51:31.891 --> 00:51:35.056 But you guys challenged me. And so I've been using agents like crazy to make, 00:51:35.896 --> 00:51:41.136 a Carpathia LLM wiki thing. And I've been freaking, oh, why did you guys get 00:51:41.136 --> 00:51:42.821 me into this? So for weeks now. 00:51:44.175 --> 00:51:48.709 been playing with this and it's totally amazing so i'm starting to change that. 00:51:49.248 --> 00:51:53.311 Person with that consideration to me it seems like logsec looked at what obsidian 00:51:53.311 --> 00:51:56.681 is doing for agents and said oh we can we can do better than that, 00:51:57.301 --> 00:52:02.001 i mean this is with the mc i'm thinking with an mcp and a cli you really could 00:52:02.001 --> 00:52:06.521 truly brain dump and then have the agent start connecting the dots right they're 00:52:06.541 --> 00:52:09.200 actually really good at that kind of context stuff so you could have it, 00:52:09.961 --> 00:52:13.896 post facto and you know obviously propose it before it makes the changes. 00:52:14.355 --> 00:52:16.981 But post-factor, you could say, go through and link things. And I've done this. 00:52:17.401 --> 00:52:18.802 That's why I can speak to this. 00:52:19.643 --> 00:52:21.727 I wasn't very good at linking my stuff together. 00:52:22.901 --> 00:52:25.615 Some people, they go into Obsidian, and they have these awesome, 00:52:26.132 --> 00:52:30.370 I don't know, point graphs or whatever, this incredible cluster of notes. 00:52:30.747 --> 00:52:35.089 Mine are just like dots all disconnected, just like stars in space. 00:52:35.391 --> 00:52:39.951 And so I was very bad at that. But the agents are able to go through and very 00:52:39.951 --> 00:52:42.841 effectively notice the commonalities. 00:52:42.841 --> 00:52:44.569 They're like your librarian almost. 00:52:45.155 --> 00:52:49.891 I had a thought on your question too about the tagging versus – and as it relates 00:52:49.891 --> 00:52:54.362 to directories. I think – so one version is you can just do the same thing where 00:52:54.710 --> 00:52:57.311 you were going to say where in the tree does it go. 00:52:57.361 --> 00:52:59.941 It could be which is the first tag I had. 00:53:00.261 --> 00:53:03.161 But then the thing you don't have to do is fight about the relative importance. 00:53:03.161 --> 00:53:05.681 You don't have to be like, oh, well, should it be in this one or this one? 00:53:05.681 --> 00:53:06.721 Because you just add both. 00:53:08.340 --> 00:53:11.214 Okay alright yeah and then the other thing too is like, 00:53:12.873 --> 00:53:16.501 It's sort of nice, if you have an agent in front of it, you could capture notes 00:53:16.501 --> 00:53:20.628 in Telegram, you could capture them in Kate, wherever you want, and then, 00:53:21.702 --> 00:53:25.051 the agent could go through and save them, you know, using this database, 00:53:25.051 --> 00:53:27.912 this block system. The agent could be the one that files it for you and then retrieves it for you. 00:53:28.504 --> 00:53:32.331 The one limitation right now is there isn't a headless MCP, so it requires the 00:53:32.331 --> 00:53:34.601 window to be running, but there's no, I was taking a look at the code, 00:53:34.601 --> 00:53:36.401 and there doesn't seem to be any technical limitation there. 00:53:36.401 --> 00:53:37.955 It just isn't plumbed up that way yet, so. 00:53:38.994 --> 00:53:43.221 So check on Wes's GitHub to see next week if we got something? 00:53:46.441 --> 00:53:48.642 Something to consider. I mean, you've got me thinking about it. 00:53:50.021 --> 00:53:53.371 Like, it might not be for anyone, but I think it's a fun way to at least make 00:53:53.371 --> 00:53:58.447 you rethink some of how you think about and relate to the knowledge you're trying to store. 00:53:59.213 --> 00:54:01.431 Yeah, and if you're looking for a knowledge system, that's, yeah, 00:54:01.431 --> 00:54:05.261 that is... All right, so LogSec, L-O-G-S-E-Q. 00:54:05.947 --> 00:54:08.681 But it definitely now, it maybe also makes it a little clearer because it's 00:54:08.681 --> 00:54:12.831 definitely now not, decisively not competing for, I have an existing markdown 00:54:12.831 --> 00:54:15.629 infrastructure that I want a thing that can upgrade it to a graph. 00:54:15.943 --> 00:54:17.580 That is totally separate now. 00:54:27.784 --> 00:54:32.364 And we've got some feedback from our dear Olympia Mike. Howdy, JB team. 00:54:32.898 --> 00:54:36.730 I've been quiet since LinuxFest Northwest, but definitely not idle. 00:54:37.330 --> 00:54:41.090 Lots of stuff has been happening thanks largely to people in this community. 00:54:41.930 --> 00:54:45.420 First, Nixbook 2605 is officially out. 00:54:45.420 --> 00:54:45.810 With Flakes? 00:54:45.810 --> 00:54:47.790 With a new installer. 00:54:47.790 --> 00:54:49.430 Hey, new installer's cool. 00:54:49.770 --> 00:54:55.040 Yeah, Flakes? Good question, bro. And that inherited upstream changes that allows 00:54:55.040 --> 00:54:58.370 new users to choose their file system on install Oh, hey. 00:54:58.670 --> 00:54:59.170 Oh. 00:54:59.550 --> 00:55:04.280 And it has a progress bar that doesn't appear to get stuck at 46%. That's a big upgrade. 00:55:04.280 --> 00:55:05.597 That's nice. Yeah, I like that. 00:55:06.021 --> 00:55:10.090 In Nick's book itself, the channel, okay, that answers it, has been bumped to 00:55:10.090 --> 00:55:14.070 2605, a new version of Cinnamon, a collection of default-free wallpapers, 00:55:14.070 --> 00:55:17.810 and fixes to Etsy scale so multi-user systems have more consistency. 00:55:17.810 --> 00:55:18.540 That's a nice touch. 00:55:18.540 --> 00:55:19.150 Good, good. 00:55:19.150 --> 00:55:22.880 This has all been aggressively tested and working well. If you have a Nick's 00:55:22.880 --> 00:55:26.157 book, you'll automatically get this within a week, or if you want it Now, 00:55:26.493 --> 00:55:28.920 you can just run Upgrade and Reboot from your start menu. 00:55:28.920 --> 00:55:29.410 Love that. 00:55:29.410 --> 00:55:34.630 So cool. The new installer can be downloaded from nixbookos.org. 00:55:35.584 --> 00:55:40.410 Also, the Computer Upcycle Project has officially won its first grant as well. 00:55:40.950 --> 00:55:45.570 The city of Olympia has chosen the Computer Upcycle Project as one of 60 nonprofits 00:55:45.570 --> 00:55:49.070 to fund this year and awarded us $15,000. 00:55:49.070 --> 00:55:50.241 Congratulations. 00:55:52.269 --> 00:55:55.158 Very well deserved. This will be used to purchase chargers, SSDs, 00:55:55.158 --> 00:55:58.776 and branded laptop bags, along with continuing to fund the development of Nixbook. 00:55:59.003 --> 00:56:02.128 We're also committed to hosting monthly workshops in Olympia aimed at the tech 00:56:02.128 --> 00:56:07.318 curious and youth to show them hands-on, and how to take that old laptop and turn it into a Nixbook. 00:56:08.018 --> 00:56:11.158 All this makes me super excited and happy about the future of Linux and getting 00:56:11.158 --> 00:56:13.707 more people involved and aware of what's possible. 00:56:14.032 --> 00:56:17.158 And of course, I'll urge them all to be listeners of this here podcast. 00:56:17.658 --> 00:56:21.193 Thanks again to everyone who has donated, supported, or simply cheered on Nixbook, 00:56:21.733 --> 00:56:24.996 and the non-profit. This community really does make anything possible. 00:56:25.901 --> 00:56:26.898 That is so sweet. 00:56:26.898 --> 00:56:32.288 And it feels with the price of hardware right now more important than ever that 00:56:32.288 --> 00:56:33.430 there's something like this out there. 00:56:34.214 --> 00:56:36.304 And I wonder if it could be a model for other areas because, 00:56:36.907 --> 00:56:38.852 you know, Mike's doing a lot. 00:56:39.125 --> 00:56:42.108 But it'd be great to see if like another organization like this could pop up, 00:56:42.108 --> 00:56:45.078 say like on the East Coast and then pop up in the UK and pop up in Canada. 00:56:45.638 --> 00:56:46.579 You know, just sort of sprout. 00:56:47.496 --> 00:56:50.398 I think that'd be really neat to see. But in the meantime, Olympia Mike is doing 00:56:50.398 --> 00:56:52.598 incredible work and I do appreciate the update. 00:56:52.598 --> 00:56:53.098 We sure do. 00:56:53.098 --> 00:56:57.418 It's nice to hear. Thank you, sir. And keep up the good work and keep us posted 00:56:57.418 --> 00:57:01.045 on it. With that, gentlemen, let's do a couple of boosts. 00:57:06.020 --> 00:57:11.528 And Greg the Lawyer is back. He's our baller booster with 100,000 sats. 00:57:11.958 --> 00:57:12.338 Wow. 00:57:23.572 --> 00:57:26.619 Well, Greg, the lawyer did. He said, love the discussion on making a Clanker 00:57:26.619 --> 00:57:30.279 OS. I've been using Claude to maintain my flake for a while, 00:57:30.279 --> 00:57:31.443 and it's nearly flawless. 00:57:31.861 --> 00:57:37.237 I now get a dedicated Hermes profile set up for the task. My Hermes runs on a Debian VM. 00:57:37.596 --> 00:57:41.429 I would love a Clanker stream on how you have yours running on Nix. 00:57:41.429 --> 00:57:42.059 Uh-huh. 00:57:42.571 --> 00:57:44.289 I really— Oh, a Nix combo stream. 00:57:44.289 --> 00:57:50.589 I feel like I missed the boat. I set up just the most amazing EV camping routing. 00:57:50.589 --> 00:57:53.769 Like, you know, all the years I've been road tripping, I've been collecting 00:57:53.769 --> 00:57:57.909 like what my idea of a perfect app would be, you know, wildfire, 00:57:57.909 --> 00:58:03.169 cellular signal, closures, everything I've ever wanted to know in one thing. 00:58:03.169 --> 00:58:05.838 And I finally built it. I should have streamed it. Should have streamed it. 00:58:06.441 --> 00:58:07.409 I could stream a tour. 00:58:07.409 --> 00:58:10.519 I feel when we maybe get back and we get our feet underneath us, 00:58:10.519 --> 00:58:13.089 Greg, I feel like it's going to be time for another clanker stream. 00:58:13.089 --> 00:58:15.329 We've probably gone too far, but we just we got stuff coming up. 00:58:15.329 --> 00:58:19.749 We're doing a little trip to Canada, up to Brentlandia. So when we get back from that. 00:58:19.749 --> 00:58:20.899 I think we got to pin some tabs. 00:58:20.899 --> 00:58:23.599 We're going to start cooking. We're going to pin some tabs, and we're going 00:58:23.599 --> 00:58:26.819 to start cooking on a clanker stream. That's the mission. Thank you, 00:58:26.819 --> 00:58:28.360 sir. Appreciate that. Baller boost. 00:58:29.538 --> 00:58:34.641 BTC is my 401k boost sin with 3,998 sats. 00:58:37.230 --> 00:58:40.608 Oh, no. Seems my lightning node can't reach Brent or the JB node. 00:58:40.736 --> 00:58:42.219 But here's some more for you. 00:58:42.219 --> 00:58:42.879 No, thanks. 00:58:42.879 --> 00:58:47.219 And a catch-up boost to provide value while my LB hub was down and stream and sats were depleted. 00:58:47.219 --> 00:58:51.519 Well, thank you. So, BTC, I saw that boost come in live. and I was like, 00:58:51.519 --> 00:58:52.503 oh, what's going on? Went over. 00:58:53.199 --> 00:58:56.259 Yeah, Tor was totally down, and that node's on Tor. Totally offline. 00:58:56.259 --> 00:58:57.089 Yeah, that'll make it a little hard. 00:58:57.089 --> 00:59:00.489 Won't it? Yeah. I was like, oh! It was like, I wasn't monitoring Tor. 00:59:00.489 --> 00:59:02.709 I was monitoring the lightning daemon. I was monitoring the Bitcoin daemon. 00:59:02.709 --> 00:59:03.659 I was monitoring the boost. 00:59:04.339 --> 00:59:09.604 Not monitoring Tor. So I got that fixed. Thank you very much. BTC is my 401k. 00:59:10.492 --> 00:59:14.021 Odyssey West is here with 2,550 sats. 00:59:16.726 --> 00:59:21.934 Hey, could you guys look into Odysseus, that PewDiePie thing he made? 00:59:22.333 --> 00:59:26.090 I know he's probably a little more jank compared to something like Hermes, 00:59:26.090 --> 00:59:27.750 but, you know, he's still worth 00:59:27.750 --> 00:59:31.930 looking into since the popular YouTube is building it and deploying it. 00:59:32.573 --> 00:59:34.862 It might be something worth looking into. It could be fun. Yeah. 00:59:35.570 --> 00:59:38.286 We've been talking about building an agentic test harness bench. 00:59:39.465 --> 00:59:42.920 Something like that. That could be a good one. Thank you, Odyssey. 00:59:42.920 --> 00:59:46.222 Nice to hear from you. Autobrains here with 5,000 sats. 00:59:49.060 --> 00:59:52.699 I did it. I took the plunge and I signed up for an intro computer programming course. 00:59:53.135 --> 00:59:58.860 Oh, until now, I've just been a dabbler in writing scripts. The course is about Python. Nice. 00:59:58.860 --> 00:59:59.340 Nice choice. 00:59:59.340 --> 01:00:01.160 It's asynchronous online at my university. 01:00:01.160 --> 01:00:01.761 Cool. 01:00:02.232 --> 01:00:05.640 It's well done. It's a good blend of bite-sized chunks of written and video 01:00:05.640 --> 01:00:08.150 content with homework and quizzes. Nice. 01:00:08.850 --> 01:00:12.530 Have you guys taken lots of computer programming courses? What worked especially 01:00:12.530 --> 01:00:16.170 well for you and what inspired you? Oh, man. 01:00:17.131 --> 01:00:22.600 The only official computer course I've taken, I think, other than like training 01:00:22.600 --> 01:00:25.694 courses on like, here's a crash course on PHP. I did one of those. 01:00:26.187 --> 01:00:27.847 I did a crash course on Python. 01:00:28.790 --> 01:00:34.900 But the only actual official training I took was basic. I'm an old man. 01:00:34.900 --> 01:00:37.170 And yet I've never seen you write basic. How is that? 01:00:37.170 --> 01:00:42.310 Well, I wrote some cool basic programs. Let me tell you what. QBasic was my friend. 01:00:42.310 --> 01:00:44.890 If it helps, Visual Basic was my first programming language. 01:00:45.450 --> 01:00:48.484 I hated Visual. Okay, so, what about you? 01:00:49.150 --> 01:00:52.070 Yeah, well, before I ended up getting a physics degree, I was a CS major, 01:00:52.070 --> 01:00:56.590 so I took the first two or three years of most of those classes. I would say... 01:00:56.610 --> 01:00:57.940 Everything Simpson's been self-taught? 01:00:59.200 --> 01:00:59.660 Yeah. 01:00:59.660 --> 01:00:59.990 Wow. 01:01:00.330 --> 01:01:05.539 I think, no, I did, I think I did one MOOC-style... 01:01:06.828 --> 01:01:08.675 uh functional programming course 01:01:08.675 --> 01:01:11.835 and a machine learning course but yeah that was a long time ago yeah. 01:01:12.475 --> 01:01:13.810 Back when we called it machine learning yeah. 01:01:13.949 --> 01:01:19.355 Yeah that's how long ago it was what usually helps me is sort of trying to chase 01:01:19.355 --> 01:01:22.595 the thing that i'm most curious about like that doesn't work you know some things 01:01:22.595 --> 01:01:25.339 you just have to grind through that's always gonna be the case but you know 01:01:25.524 --> 01:01:27.916 finding things that motivate you that you're curious that make it, 01:01:28.444 --> 01:01:33.675 feel more like play than like work and if you can try to keep do whatever you can 01:01:33.971 --> 01:01:37.325 sometimes it's worth a little like you don't want to over fixate on your setup 01:01:37.325 --> 01:01:40.765 and not the actual work but it is worth a little bit of investment just to try 01:01:40.765 --> 01:01:44.965 to keep the feedback loop as quick as possible because that helps keep it feeling like play. 01:01:44.965 --> 01:01:47.299 Yeah i agree that's good that's good advice. 01:01:47.612 --> 01:01:52.303 I was doing exactly that last night where i will confess that i stayed up, 01:01:52.773 --> 01:01:58.215 all night never went to bed at all because i was so deep in the flow of like 01:01:58.215 --> 01:02:02.365 playing and chasing curiosities and like learning all sorts of stuff and breaking 01:02:02.365 --> 01:02:03.635 things and fixing things. 01:02:04.015 --> 01:02:07.842 You still showed up late for the stream. How does that work? You're up. 01:02:08.172 --> 01:02:12.895 Well, I don't know. I was moving slower, and I'm on a different time zone. 01:02:12.895 --> 01:02:14.975 There wasn't even any squirrel-based excuses. 01:02:14.975 --> 01:02:15.305 I know. 01:02:15.305 --> 01:02:16.805 Usually there's no squirrel-based reason. 01:02:16.805 --> 01:02:21.595 We were all up late last night. I was up late last night. I replumbed. 01:02:22.315 --> 01:02:26.195 I have Blue Bubble set up, and I was replumbing it, and then I was working on 01:02:26.195 --> 01:02:29.035 some of the EV road trips. I was up way too late. 01:02:29.035 --> 01:02:29.744 Too. 01:02:30.535 --> 01:02:31.235 We all know we have a show. 01:02:31.255 --> 01:02:32.835 Your version of late and my version of later. 01:02:33.235 --> 01:02:35.479 Yeah, that's true. My version of late's like midnight. 01:02:35.700 --> 01:02:36.315 Mine's like 4 a.m. 01:02:36.715 --> 01:02:42.885 Yeah, yeah, yeah. Marvin42 came in with $42 in Fiat Funk coupons. 01:02:42.885 --> 01:02:43.844 Nice. 01:02:45.092 --> 01:02:47.641 Here's a few bucks to help pay for refactoring Hypervibe. 01:02:50.941 --> 01:02:55.971 I just found out that the Dendritic Nix idea a few days ago and have been trying to get Hermes to use it. 01:02:56.354 --> 01:02:59.616 I'm still struggling with the concept, and it wants to build everything monolithic 01:02:59.924 --> 01:03:03.465 or make each setting a file. I'm for better skill shopping. 01:03:04.021 --> 01:03:05.751 Ooh, fun update. Thank you, Marvin. 01:03:05.751 --> 01:03:06.241 That's so good. 01:03:06.261 --> 01:03:09.781 The Dendritic Nix stuff is interesting. I think that might be something I have to poke at more. 01:03:10.261 --> 01:03:11.387 Thank you. That's a great boost, too. 01:03:11.811 --> 01:03:15.759 Anonymous boosts in with 25 Fiat Fun coupons. 01:03:18.290 --> 01:03:21.741 A long-time listener, first-time booster. Thank you from Austria. 01:03:21.741 --> 01:03:25.691 Awesome. Thank you for the first-time boost. Appreciate that. 01:03:26.167 --> 01:03:29.244 Jem Lepore is here with 10 Fiat. 01:03:31.322 --> 01:03:35.234 Thanks for all the hard work you guys put into this show week after week. 01:03:35.728 --> 01:03:41.353 Next time you're headed out east, come have a good barbecue with us in KCUS at Kansas City. 01:03:42.001 --> 01:03:42.931 We would love that. 01:03:42.931 --> 01:03:43.849 That sounds ideal. 01:03:43.971 --> 01:03:45.608 No complaints here. 01:03:45.887 --> 01:03:48.127 Someone just needs to get a certain van working. 01:03:48.255 --> 01:03:51.211 Oh, I can do that. I just need a reason, you know. Just give me a reason. 01:03:51.211 --> 01:03:51.641 A little bit of a kid. 01:03:51.681 --> 01:03:52.041 Barbecue. 01:03:52.641 --> 01:03:52.801 I'm in. 01:03:54.083 --> 01:03:56.731 I don't think I've ever really properly had it. My biggest exposure has always 01:03:56.731 --> 01:03:58.511 been to Texas barbecue. I'm open-minded. 01:03:58.511 --> 01:03:59.662 We need to do a barbecue tour. 01:04:01.461 --> 01:04:04.182 High five connoisseurs here with $5 in Fiat Fun Coupons. 01:04:06.498 --> 01:04:10.145 How are you all thinking about the secession planning now that there are agents? 01:04:10.145 --> 01:04:10.765 A whole good question. 01:04:11.125 --> 01:04:14.735 I imagine my family could get an agent to handle much of the home lab maintenance after I'm gone. 01:04:14.735 --> 01:04:15.475 Fascinating. 01:04:15.475 --> 01:04:17.965 Wow. Do you remember that episode we made? And then there was like a couple 01:04:17.965 --> 01:04:21.893 projects trying to tackle this, but we came to the conclusion that it's just a very hard problem. 01:04:22.461 --> 01:04:24.075 A secession agent. 01:04:24.075 --> 01:04:27.320 This is cool. Start building, I say. 01:04:28.376 --> 01:04:29.245 I mean, there's something there. 01:04:29.245 --> 01:04:32.785 Just imagining the chat GPT thing behind a window they break in case of. 01:04:33.305 --> 01:04:38.745 Break in case or like a or telegram chat thread that's like set up but don't use I mean it gives. 01:04:38.745 --> 01:04:40.315 A whole new meaning to soul.md. 01:04:40.315 --> 01:04:45.395 Doesn't it oh okay we gotta move on it's getting dark Wes you wanna take our. 01:04:45.395 --> 01:04:51.825 First one member free booth yes I bear comes in with two member free booths oh by the way had to re, 01:04:52.625 --> 01:04:56.965 sign up for Jupiter Party not sure what email I used oh yeah that's that happens 01:04:56.965 --> 01:05:00.635 that happens for sure thanks for having the web booths with USD I'm not into 01:05:00.635 --> 01:05:03.375 Bitcoiner stats but I love the shows Thank you very much. 01:05:03.385 --> 01:05:07.558 Here's a little utility that I have been vibe coding for Linux-based routers and firewalls. 01:05:07.878 --> 01:05:12.887 I'm using it on my ViOS boxes, but it works on any Linux instance using IP chains. 01:05:13.311 --> 01:05:15.125 Cool. FW Top. 01:05:15.125 --> 01:05:17.961 This looks really good. It's got a great TUI. 01:05:18.152 --> 01:05:22.195 Python, MIT licensed. Oh, very colorful and pretty TUI. You are right. 01:05:22.825 --> 01:05:26.745 A real-time router firewall traffic visualizer for the terminal inspired by 01:05:26.745 --> 01:05:29.325 BTOP. You nailed it. It looks so good. 01:05:29.733 --> 01:05:35.469 This is impressive. huh and a release three days ago very nice. 01:05:36.844 --> 01:05:38.505 That's really sweet that is so 01:05:38.725 --> 01:05:43.505 so what we need is uh you need to vibe up a flake.nix we'll give it a shot. 01:05:45.494 --> 01:05:45.955 Thanks for the. 01:05:45.955 --> 01:05:49.406 Support yeah thank you very much and thanks for signing up again appreciate that. 01:05:50.823 --> 01:05:54.689 Well, Mr. Pickles is here with a free members boost. 01:05:56.343 --> 01:06:05.224 You should provide API keys to members. We have three 3090s and use about 2 billion tokens a month. 01:06:05.584 --> 01:06:09.009 Oh, 3090s as in NVIDIAs. 01:06:09.310 --> 01:06:13.309 Oh, yeah, right. I don't even, you know, I've just blocked this out of my mind. 01:06:14.146 --> 01:06:19.580 I also had to pick Hisler, a full text indexer for every website you visit. Whoa. 01:06:19.841 --> 01:06:23.609 Great for finding a thing you read about a month ago, but can't recall. Holy smokes. 01:06:23.609 --> 01:06:24.519 This is like recall. 01:06:24.519 --> 01:06:27.179 But just let's read it. It indexes their full content, so you can find information 01:06:27.179 --> 01:06:30.603 again from the web interface terminal or an AI assistant connected through MCP. 01:06:31.782 --> 01:06:31.989 Huh. 01:06:31.989 --> 01:06:38.389 I guess it's by Adam Tauber, Ascii Moo, who also made CRX or Chex. 01:06:38.974 --> 01:06:40.569 Maybe not the NG one, but the one before that. 01:06:40.589 --> 01:06:41.009 No kidding. 01:06:41.009 --> 01:06:43.329 Kali and a bunch of other smaller projects. Yeah. 01:06:43.769 --> 01:06:48.964 No kidding. That's such a neat find. Thank you, Pickles. Appreciate that. 01:06:49.597 --> 01:06:52.830 I'll take the king. King Arthur is back with a member boast. 01:06:55.332 --> 01:06:59.132 Uh, is not a king voted into parliament as one doesn't vote for one's king. 01:06:59.132 --> 01:07:04.068 Rather, the lady of the lake bestows Excalibur to the one worthy to rule over Britain. Of course. 01:07:04.317 --> 01:07:07.762 I do not require a sound clip or a member boost. Only a specific boost for those 01:07:07.762 --> 01:07:13.437 who can bestow one, two, five, five, three sats to you in the future. Oh. 01:07:13.965 --> 01:07:16.345 What is that? Twelve? 01:07:16.757 --> 01:07:20.442 Does it require a map or just some kind of culture? 01:07:20.462 --> 01:07:22.862 I think it's related to the overall boost. 01:07:22.862 --> 01:07:24.142 It must be. Yeah, we have some homework. 01:07:25.022 --> 01:07:27.201 Thank you, Cain. Appreciate you. 01:07:28.199 --> 01:07:33.222 LP Boost Sin, I'm just starting to dig in for Nix OS, but I got confused from the start. 01:07:33.222 --> 01:07:33.731 It's okay. 01:07:33.975 --> 01:07:37.580 What's the difference between declarative Nix OS versus the community one? 01:07:38.143 --> 01:07:42.602 Which one is the best for a VPS versus desktop laptop? Oh, you must mean determinate Nix. 01:07:43.202 --> 01:07:43.382 Ah. 01:07:43.862 --> 01:07:44.737 Autocomplete, maybe. 01:07:46.122 --> 01:07:50.202 Okay, so determinate Nix is, yeah, it's like, yeah, okay. That's like, 01:07:50.202 --> 01:07:52.412 I guess you would call it a distro of Nix? 01:07:52.888 --> 01:07:54.722 a spin of Nix a fork of Nix. 01:07:55.262 --> 01:07:56.202 Yeah somewhere in there. 01:07:56.202 --> 01:08:00.362 Yeah with some tools and installers that are designed for you know enterprise 01:08:00.362 --> 01:08:04.902 use and what not in there perfectly fine I think though probably for most people 01:08:04.902 --> 01:08:08.382 unless they know they need determinant Nix they probably just want to use Nix yeah. 01:08:08.602 --> 01:08:12.752 It offers some extra stuff but you you probably won't need or necessarily even 01:08:12.752 --> 01:08:15.702 care about a lot of that until you are a more experienced Nix user or you're 01:08:15.702 --> 01:08:18.510 trying to use it for some like production sort of purpose. 01:08:18.759 --> 01:08:23.592 What about Wes what about if you what about out if he wanted to install it on 01:08:23.592 --> 01:08:24.749 an existing distribution. 01:08:25.016 --> 01:08:27.292 Yes so for that the determined nix installer is great. 01:08:27.292 --> 01:08:27.913 I think so too. 01:08:29.156 --> 01:08:33.412 And then i also it also means like this decision doesn't matter for the most 01:08:33.412 --> 01:08:35.742 part as much as you think it is so if you're going to install it there just 01:08:35.742 --> 01:08:37.162 reach for whatever as long as you get nix, 01:08:37.722 --> 01:08:40.612 it'll work uh if you're going with nix os then you're probably just going to 01:08:40.612 --> 01:08:43.182 get like the standard nix os install media and then you're just going to get 01:08:43.182 --> 01:08:46.861 the plain upstream nix os anyway uh and either way, 01:08:48.080 --> 01:08:51.454 They'll all build your next packages. So for that, you'll be just fine. 01:08:51.454 --> 01:08:57.314 The amount of overlap is overwhelming. Like there's a very tiny slice of notable 01:08:57.314 --> 01:09:00.884 differences, but most of it as a beginner, you just learn. 01:09:00.884 --> 01:09:04.364 All right. Help me break this down. Dance Tango writes, Chris, 01:09:05.044 --> 01:09:08.194 you considered a dedicated camera, a camera without AI that fixes the image. 01:09:08.194 --> 01:09:10.980 Is that really something for you? Do you remember what they're speaking about? 01:09:11.229 --> 01:09:14.864 Yeah. I think we had a caller mention how. 01:09:14.864 --> 01:09:16.404 Oh, we're talking about dedicated pocket cameras. 01:09:16.904 --> 01:09:19.144 Or dedicated like a music player or. 01:09:19.144 --> 01:09:19.937 Something like that. 01:09:21.864 --> 01:09:27.764 So the question is a camera without ai that fixes oh the camera without ai that 01:09:27.764 --> 01:09:30.061 fixes the image is that something for you. 01:09:30.833 --> 01:09:34.654 I would think if you're going for like a pocket camera you probably want you 01:09:34.654 --> 01:09:38.334 don't want something manipulating the image right you do that after the fact if you want to do it. 01:09:38.334 --> 01:09:39.114 Right yeah that's what i would. 01:09:39.114 --> 01:09:40.864 Think how much of a purist are you that's a. 01:09:40.864 --> 01:09:43.924 Scale that's that is a good point though like with the camera phones you're 01:09:43.924 --> 01:09:45.034 not really getting a true picture 01:09:45.034 --> 01:09:48.014 you're getting like an interpretation that's pretty close you always. 01:09:48.014 --> 01:09:49.024 Did you always. 01:09:49.024 --> 01:09:50.256 Did that's true. 01:09:51.214 --> 01:09:56.554 Well uh paul comes in with our last member boost half listened to the bit about analog. 01:09:56.554 --> 01:09:57.264 Cameras two. 01:09:57.264 --> 01:10:01.464 Years ago i've started taking pictures of concerts with a nintendo 3ds awesome 01:10:01.464 --> 01:10:04.484 it's far more engaging for me than taking them with my phone if you're interested 01:10:04.484 --> 01:10:06.864 you can see them on 3d stage vibes dot. 01:10:06.864 --> 01:10:08.154 80 stop it. 01:10:08.154 --> 01:10:09.484 Let me know if you want to know more. 01:10:09.484 --> 01:10:11.258 That's that is definitely a vibe. 01:10:12.557 --> 01:10:15.346 So you were only half listening, and yet you sent us some cool photos? 01:10:15.346 --> 01:10:16.388 Thank you. That's pretty great. 01:10:16.533 --> 01:10:20.606 That's pretty great. I'm looking at them right now. 3DS stage vibes. 01:10:20.606 --> 01:10:24.676 They're surprisingly good. Wow. Yeah. 01:10:24.676 --> 01:10:27.731 Yeah, for a 3DS, right? They're not better than you. Yeah. 01:10:28.213 --> 01:10:31.466 There's a philosophy in the photography world that says the best camera you 01:10:31.466 --> 01:10:32.927 have is the one you have with you. 01:10:33.054 --> 01:10:36.316 So if you've got a 3DS, you start shooting with that thing. 01:10:36.316 --> 01:10:37.675 And why don't you have a 3DS? 01:10:38.644 --> 01:10:40.978 I've actually got a dead one I need to replace the battery of. 01:10:41.251 --> 01:10:42.708 I really would like to get it working again. 01:10:42.818 --> 01:10:43.306 Jeff? 01:10:43.306 --> 01:10:46.976 Thank you, everybody, who boosted the show. We had 12 of you stream sats as 01:10:46.976 --> 01:10:52.226 you listened. You sat streamers. You collectively stacked 25,726 satoshis. 01:10:52.226 --> 01:10:53.347 Not too shabby. 01:10:53.533 --> 01:10:57.286 Not too shabby at all. When you bring it together with the fiat boosts and all 01:10:57.286 --> 01:11:01.526 the traditional boosts and convert it all to sats, our grand total for this 01:11:01.526 --> 01:11:06.002 week's episode is 265,638 satoshis. 01:11:06.344 --> 01:11:09.915 We tell you the numbers so that way we're transparent. You know what the show makes. 01:11:10.448 --> 01:11:13.256 you can decide what you think the value of the show is we put it out there and 01:11:13.256 --> 01:11:16.266 we ask if you get value from it and want to keep it going, prevent the fate 01:11:16.266 --> 01:11:20.276 like the Linux magazines have seen you can do it at boost.jupiterbroadcasting.com 01:11:20.276 --> 01:11:22.204 or with a podcasting 2.0 app. 01:11:28.224 --> 01:11:31.526 If your boost is above 2,000 sats we'll read it on the air you can also just 01:11:31.526 --> 01:11:35.366 stream sats you can boost below 2,000 we still put in our show notes and we 01:11:35.366 --> 01:11:38.546 read it ourselves and have a nice little chuckle usually because they're usually 01:11:38.546 --> 01:11:40.686 quite nice Well, we really appreciate it, of course. 01:11:40.686 --> 01:11:44.556 You can get a membership at linuxunplugged.com slash membership or support the 01:11:44.556 --> 01:11:46.073 whole network at jupyter.party. 01:11:46.649 --> 01:11:50.686 Keep the thing going and get the bootleg or the ad-free version of the show 01:11:50.686 --> 01:11:54.166 and have that support on autopilot, and then you get access to the free member 01:11:54.166 --> 01:11:56.586 boost at boost.jupyterbroadcasting.com. 01:11:57.131 --> 01:11:57.689 Thank you, everybody. 01:12:02.798 --> 01:12:05.053 All right, so we're doing something a little different for the pick segment, 01:12:05.053 --> 01:12:08.487 and I'm going to make it quick because I'll tell you right up front, 01:12:08.887 --> 01:12:11.320 Moonlight and Sunshine are the way to go. 01:12:11.842 --> 01:12:15.133 Sunshine runs on the remote computer, captures the desktop, and encodes using 01:12:15.133 --> 01:12:19.267 your GPU or CPU if you don't have a GPU. And it's more like a video stream. 01:12:19.843 --> 01:12:24.153 It's really designed for remote gaming, which means a modern composited desktop 01:12:24.153 --> 01:12:26.685 actually performs almost natively. 01:12:27.092 --> 01:12:30.573 You can even do things like use OBS, and if you had to, you could even edit 01:12:30.573 --> 01:12:32.943 video. Even over a relatively weak connection. 01:12:32.943 --> 01:12:34.533 And that's kind of where you were playing with some of this, 01:12:34.533 --> 01:12:38.283 right? The idea of like, well, we've evaluated a ton of different ways to handle 01:12:38.283 --> 01:12:39.793 remote studio production over the years. 01:12:39.793 --> 01:12:44.813 Rust Desk has been a real solid go-to, but on Wayland, it requires someone at 01:12:44.813 --> 01:12:46.751 the console to authorize the connection. 01:12:46.896 --> 01:12:47.883 Which is kind of a no-go. 01:12:47.883 --> 01:12:54.203 Cannot do that. MeshTastic is a community favorite, but I don't need fleet inventory, 01:12:54.203 --> 01:12:58.872 file transfer, process management, a dashboard. I don't need any of that. 01:12:59.943 --> 01:13:03.723 And I also don't need it to solve the proxying the remote because I'm going 01:13:03.723 --> 01:13:06.693 to use mesh networking. So I've solved the network aspect already. 01:13:06.693 --> 01:13:08.943 And there's a lot of tools that bundle a bunch of that stuff together. 01:13:08.943 --> 01:13:13.883 Yes, yes, yes. So what I really like about Sunshine is that the hardware video 01:13:13.883 --> 01:13:20.153 encoding, if you have it, even with just a basic Intel laptop GPU, is really impressive. 01:13:20.362 --> 01:13:23.711 It's fine on the CPU, too. I have one system here we have to do on the CPU. 01:13:24.216 --> 01:13:28.103 But everything works. hdr sound all comes through so you can hear the desktop 01:13:28.103 --> 01:13:32.693 audio as you're in the remote session it works with plasma and gnomes animations 01:13:32.693 --> 01:13:38.317 hyperlands animations you can watch videos it has multi-monitor support, 01:13:38.926 --> 01:13:43.013 so all of it's really really nice um and of course it has wayland support as 01:13:43.013 --> 01:13:44.689 well as x11 support which matters a lot. 01:13:44.852 --> 01:13:46.043 What's the downside what's the. 01:13:47.011 --> 01:13:51.168 Well, the downside is that you have to set up Sunshine on one end. 01:13:51.168 --> 01:13:53.748 You have to set up Moonlight, which is a desktop application. 01:13:53.748 --> 01:13:54.878 And then you pair them once. 01:13:55.518 --> 01:13:58.128 And once they're paired once, you can connect. But the downside is you do have 01:13:58.128 --> 01:13:59.218 to do that one-time pairing. 01:13:59.218 --> 01:14:02.768 So there's a little bit of setup, especially before maybe you are in a remote situation. 01:14:02.768 --> 01:14:06.768 But the handy thing is the pairing UI is a web interface. So you can probably, 01:14:06.768 --> 01:14:09.018 if you're on the mesh network, just bring up the remote web UI, 01:14:09.018 --> 01:14:11.758 put in your pin, and pair that way from the remote system. 01:14:12.318 --> 01:14:15.988 So that's not so bad. So really the downsides are pretty minimal unless you 01:14:15.988 --> 01:14:20.068 need the other things that a MeshTastic team viewer type things offer with file 01:14:20.068 --> 01:14:24.953 transfer and process management and maybe even user management and inventory. I don't need any of that. 01:14:26.198 --> 01:14:29.018 So I was really impressed because I wanted to revisit and I'll actually tell 01:14:29.018 --> 01:14:32.435 you the strongest contender doesn't come up very often. 01:14:32.830 --> 01:14:37.658 The strongest contender is not MeshTastic, is not RustEssle. Those are great. 01:14:38.159 --> 01:14:38.818 MeshTastic? 01:14:39.538 --> 01:14:46.500 Yeah, that's what it's called, right? Oh, I'm sorry. MeshCentral. Thank you. Thank you. 01:14:47.138 --> 01:14:49.518 I mean, video over MeshTastic is impressive. 01:14:50.906 --> 01:14:52.328 Honestly, if you could do remote 01:14:52.328 --> 01:14:56.380 desktop, even process management over MeshTastic, I would be thrilled. 01:14:56.978 --> 01:14:59.851 No, I meant Mesh Central. I just got the meshes meshed up. 01:15:00.988 --> 01:15:04.005 But I actually think the one that we really should give more credit to and the 01:15:04.005 --> 01:15:09.005 one that people should consider besides the sunshine and moonlight combo, 01:15:10.003 --> 01:15:12.385 the folks over at No Machine are still cooking. 01:15:12.385 --> 01:15:12.575 Wow. 01:15:12.595 --> 01:15:13.605 They've been going for so long. 01:15:13.605 --> 01:15:14.295 They've been going. 01:15:14.295 --> 01:15:15.555 Which is why I think folks forget. 01:15:15.555 --> 01:15:20.795 Yes. They've been going since the very early days. And they really have pioneered 01:15:20.795 --> 01:15:23.755 hardware H.264 encoded remote desktop streams. 01:15:24.434 --> 01:15:32.550 And it still kicks butt. I think no machine is better if you need a separate remote session. 01:15:33.537 --> 01:15:36.815 So if you want to be able to log in and have a desktop session running and have 01:15:36.815 --> 01:15:40.505 a whole other user session that's in a remote session, no machine's great at 01:15:40.505 --> 01:15:44.835 that. Sunshine, you'd have to set that up completely. Sunshine is really for 01:15:44.835 --> 01:15:46.480 taking control of the console. 01:15:46.881 --> 01:15:48.071 No machine lets you do both. 01:15:48.576 --> 01:15:53.135 And it supports USB and device redirection. So you can have printers and certain 01:15:53.135 --> 01:15:56.122 mouses and keyboards. So it does a lot more nice stuff. 01:15:56.749 --> 01:16:00.275 But if you just need something quick, you've already got the networking taken 01:16:00.275 --> 01:16:03.875 care of, like a mesh network or something like that. You don't need a central 01:16:03.875 --> 01:16:05.538 proxy or inventory or fleet. 01:16:06.368 --> 01:16:10.744 I really think for the 10 minutes it takes to get Moonlight and Sunshine going 01:16:10.744 --> 01:16:12.554 again, it's, like, super. 01:16:12.554 --> 01:16:15.704 And these are pretty easy. Like, I mean, they're now, it's been out for a while, 01:16:15.704 --> 01:16:17.804 right? So they're, like, pretty available in repos or Flatpak. 01:16:17.804 --> 01:16:20.794 Sunshine is actively still being developed. 01:16:21.274 --> 01:16:24.124 Moonlight hasn't had an update for a while, but it's a pretty stable. 01:16:24.124 --> 01:16:26.964 I mean, it's just an H.264 streaming client that supports input. 01:16:26.964 --> 01:16:28.224 So it's not very complicated. 01:16:28.567 --> 01:16:31.944 There are a few forks out there, but Moonlight, it's packaged in Flatpak. 01:16:31.944 --> 01:16:35.300 It's packaged in just about every distro now because it has been pretty stable for a while. 01:16:35.742 --> 01:16:38.384 same with sunshine but sunshine is still moving they're still progressing with 01:16:38.384 --> 01:16:42.428 it still getting better and better and it's was really nice been a couple years to check in, 01:16:43.113 --> 01:16:46.354 i did all the remote testing we'll see though we're going to go to canada and 01:16:46.354 --> 01:16:52.074 next sunday's episode will be a remote episode and if we're live that meant that uh, 01:16:52.911 --> 01:16:55.680 sunshine and moonlight worked and if we're not live. 01:16:55.930 --> 01:16:56.924 Keep on refreshing that. 01:16:56.924 --> 01:17:03.514 Rss yeah exactly yeah exactly uh it should be good though we're We've got some 01:17:03.514 --> 01:17:06.474 good projects planned when we get up to Brent's house that's going to make, 01:17:06.474 --> 01:17:07.374 I think, for a great episode. 01:17:07.374 --> 01:17:08.591 I've got a whole list. 01:17:09.026 --> 01:17:12.344 Yeah, I know you do. We're looking forward to it. So, of course, 01:17:12.344 --> 01:17:17.217 we also want to know your feedback on Linus's position in the kernel now, how you feel about that. 01:17:17.426 --> 01:17:20.684 And particularly, I'd like to reach out to those of you who are against AI use 01:17:20.684 --> 01:17:24.554 with free software and what you'd like to see happen if a project receives a 01:17:24.554 --> 01:17:29.084 legitimate concerning bug that's generated or, you know, the patch, 01:17:29.084 --> 01:17:31.520 the fix is generated by an LLM. What would you like to see that project do? 01:17:31.991 --> 01:17:34.954 A couple of range of questions there and also your thoughts and feedback on 01:17:34.954 --> 01:17:38.914 LogSec, Obsidian, and free software notes in general. I think all of those would 01:17:38.914 --> 01:17:42.234 really fill out the feedback and the boosts, and we'd really appreciate it. 01:17:42.652 --> 01:17:47.366 So why not join us? See if we make it live. See if it actually happens. 01:17:51.046 --> 01:17:54.522 Make it a Tuesday on a Sunday. Join us at 10 a.m. Pacific, 1 p.m. 01:17:54.522 --> 01:17:56.302 Eastern. Of course, we've got chapters. 01:17:56.302 --> 01:17:57.772 Oh, yeah, and transcripts. 01:17:57.772 --> 01:17:58.972 Those are all available in the feed. 01:17:58.972 --> 01:18:01.632 Magic text files just waiting for you to download them. 01:18:01.632 --> 01:18:04.392 I don't know if we'll have video next week, but we usually have an MP4 hidden 01:18:04.392 --> 01:18:07.502 in that feed as well, if you can believe it. If you could even believe it. 01:18:07.502 --> 01:18:10.052 If you find yourself a fancy client, they can make it even easier. 01:18:10.052 --> 01:18:14.353 Yeah, they'll just show it to you. Yeah. How about that? Podcasting 2.0. 01:18:14.696 --> 01:18:18.332 If you want more show, remember that lug gets together in the mumble room before 01:18:18.332 --> 01:18:20.083 we get going. They're cooking all along. 01:18:20.541 --> 01:18:23.792 members you get the full bootleg version with twice the content clocking at 01:18:23.792 --> 01:18:27.702 almost two hours right now and then links to everything we talked about today 01:18:27.702 --> 01:18:32.922 how to contact us mumble our matrix chat all of it at linuxunplugged.com and 01:18:32.922 --> 01:18:35.432 this was linuxunplugged 676, 01:18:35.906 --> 01:18:39.192 thank you so much for joining us on this week's episode of your unplugged program, 01:18:39.668 --> 01:18:44.428 we'll see you right back here next tuesday which of course is next sunday.