Late Night Linux – Episode 236

Hello and welcome to episode 236 of Late Night Linux, recorded on the 3rd of July, 2023. I'm Joe, and with me, I'll fade in. Who are palm agrar? Crayon. Hello. And Will, I've only got 600 words today. Before we do news, just want to say thank you to all the new patrons that has ticked up significantly since the last time we recorded, and we've talked about the dynamic ads and stuff. And thank you, Trevor, who got in touch about that. That is still an ongoing situation, getting all that sorted out, so it's not going to start just yet, but it will do soon. And people had concerns about, are they going to be like, super loud? Well, have you heard how loud these shows are? No one is going to be able to do ads louder than I mix this. So don't worry about that. And they're not just going to be randomly in the middle of sentences. You get to say exactly where you want these ads to be. They were, I think, the main things that people were worried about. So yes, thank you, everyone. Right, let's do the news then. And we've got a few stories that we could talk about, but let's see how we get on with this first one. I would say this is probably the biggest Linux news since canonical dropped Unity. And that is that Red Hat won't be publishing the source code for Red Hat Enterprise Linux rel to anyone except their customer. So it's not going to be published. It's just people who are paying for subscriptions. So to be clear, the source for CentOS stream will still be public. It's just Red Hat Enterprise Linux that won't be. And you can still get access to the source code if you are a customer. And you then exercise your rights under the GPL to publish that source code, then your agreement may well be terminated with Red Hat and you won't get any updates or any more versions of REL. I think before we dive into the reactions to this news, I think it's worth just mentioning who this is going to affect. Like on the face of it, you think, OK, well, people can still get the source code. But we've got the likes of CentOS, but more specifically Rocky Linux, whose whole thing is that they are compatible with i.e. bug for bug compatible rebuilds of Red Hat, but then they give it away for free. So this move is effectively stopping those projects from ever being able to be compatible, truly compatible with Red Hat again. And so it's basically shutting down their entire project. Not only them, but Alma and the big one for me which annoyed me more than anything else was scientific Linux, because you mean you've got the likes of CERN using that and what you expect a scientific organization to shallow, quite large amounts of money. I mean, it's 350, I think per workstation, so I mean, not cheap. What about poor Oracle though? Who? I know we'll get into the details in the wording of this, but I just want to remind everyone of what Jim Whitehurst said when IBM bought Red Hat for $34 billion, you know, this business model that isn't working. Joining forces with IBM gives Red Hat the opportunity to bring more open source innovation to an even broader range of organizations. So that's what they're doing. It's going well. Can we just talk about the legal situation because I've heard people misunderstand this. So if you, as an entity, publish open source software under the GPL, you have to provide people with source code for that. You have to provide the people who you gave the binaries to access to the source code, but you only have to give them source code for that particular binary that you give them. So this is a real kind of loophole as it were that Red Hat are getting away with here in that if you do exercise that right, they can cut you off, but that doesn't mean that you don't have access to the source code for the version that you already have. It just means you won't get it for updates in the future. And that's totally legal as far as I can tell. And I'm not a lawyer and all the rest of it. But the people who are saying that this is somehow a GPL violation. Just don't seem to understand how all that works as far as I can see. I think the way that I can only speak and what I think they mean is the fact that you're not supposed to restrict what people do. And it does seem like you are restricting people. If by nature of acting on the license, you are then sort of you invalidate your entire business IT infrastructure by invalidating your support contract. I think it really goes against the spirit of what the thing is. I think that you are right that it very much goes against the spirit of open source. Well, spirit and may potentially be the legal issue too, because I mean, it's not been tested. So therefore in a case law based environment, it's not proven. Therefore, it's not sort of fully been tried out. Yes, it hasn't been tested in court. But do you really think that Red Hat, let's just forget IBM for a second, just Red Hat the entity within IBM. We don't think they've got enough expert lawyers in this field to know that they're going to get away with it. You think they just didn't think it through, so I'll be fine. I don't know. I think sometimes the way of a legal argument is enough to scare people away from trying it, because it would be a hell of a risk, a hell of an expansive risk, even if you won. So I think a lot of the time in a legal case, it's a bit of a mutually assured destruction sort of event. I think you've both dived into the details and the semantics too quickly. We'll talk about it. I'm sure, but Bradley Coon of the Software Freedom Conservancy, who's spent his entire career talking about the legal issues, looks at this, and it looks like it is legal. But the big thing for me, the big profound difference that I think is inarguable is that Red Hat has always been a beacon of open source software development and success in its commitments to the kernel and all the other things that it commits to GNOME. And continues to do so, most sailors, lots of software engineers who are continuing to do that. But it was always under the umbrella that it was going to do good, and this is something we've seen time and time again, especially when companies are bought by other larger companies, and it will happen again in the future. The spirit of Red Hat is dead. That's how it feels. That's what this move feels, regardless of the legality of what they're doing. Well, that's something that there's been a lot of debate about. Is this a Red Hat initiative or is this an IBM initiative? Is it IBM forcing Red Hat to do this? And from listening to what the Red Hat employees say, they don't get that impression at all. As far as they're concerned, this is coming from Red Hat and not IBM. And I don't see any reason to disbelieve that. If the people within the company are saying that publicly. Would they have a job if they disagreed? And I suspect there's very tight controls over what employees at Red Hat can say under IBM, because there should be, you know, I probably shouldn't be allowed to be saying this. Should I say that? I don't know. But why is there a red dot in your forehead, Grimm? But I mean, there will be tight controls. And maybe they've been told if they have something positive to say they can say something, but if they have anything negative to say, they really shouldn't be saying it. Well, that said, I've seen some negative takes from Red Hat people. I've heard them say, look, this is not ideal. I don't agree with this, for example, very publicly on Mustardon and Twitter and stuff. I think regardless of who is giving the message, it all must stem from the very, very top of IBM who are ultimately the bosses of Red Hat. If IBM are concerned about the bottom line, which they are, and they see Red Hat as a cash cow, they will be looking for ways to improve the income of Red Hat to prop up the roster of IBM while it's having troubling times. If the boss of IBM says to the boss of Red Hat, we need you to make more money. It doesn't matter what you do. We don't care. Just make more money. Then the CEO of Red Hat is obligated to find opportunities to bring in more money. And that's exactly what they've done. What about the argument that there's no value in the rebuilds? Because I've seen Red Hat employees say that they tried to make the case that Red Hat shouldn't do this. And they just couldn't gather enough evidence that it was worth Red Hat continuing to publish the source code and enable Alma Rocky at Al to keep doing their rebuilds. Because they don't do enough upstream work. It's sort of impossible for them to really do that as the argument because they have to be bug for bug compatible with REL. And so they are inherently downstream from it and can't really contribute upstream because they're not even really downstream. They're almost like side stream because they're just a complete copy of it. It's not like Ubuntu, which is downstream from Debian and they can contribute upstream to that. And Red Hat wants people to contribute to CentOS stream instead because that will feed into REL. Yeah, I think the problem there is the fact that this feels like a problem of Red Hat on making because they took over CentOS and because they did that, two things spawned from the side of it. And I think the problem there is the fact that they did it in the middle of a release and they've done it on this one in the middle of another release. So it was first in eight, which was supposed to be support for 10 years and then they help nail that and said, no, no, it's all gone downstream now and just tough. And then the same thing has now happened for nine. So it's kind of hard to believe in what they say with that because they're a company that was making billions and it's hard to feel too sorry for a company doing that in the fact that it seems like this is just a land grab maybe. Now they might very well have some other information that I don't about what maybe Alma or Rocky were doing. I have no clue where even Oracle Linux, it could be them. But the thing is they have existed sort of in peaceful coexistence with Oracle for so long that I don't think it was them. The original justification, if you like, for there being these other bug for bug compatible versions of Linux was that from Red Hat's perspective, you were building a community of Red Hat knowledgeable and experienced and even qualified engineers who could take that experience out to their workplaces and use competing products with Red Hat. But still the Red Hat ecosystem would build and the amount of people out there knew how to use it would build and that would reinforce the dominant position of Red Hat. Mike McGrath, a VP of platforms from Red Hat wrote a couple of blog posts talking about the decision generally and crying a little bit about how the community got upset about it. And what he really focuses on is that there is, as Joe said, there is no value in these rebuilds for them anymore. They've got CentOS, which is now different. But they don't value that community of Red Hat focused users to take that out to the wider world. They have now said we don't care about that anymore. And that's a real shame. Well, my understanding is that they believe it to be a myth that having all those people out there using the rebuilds will actually translate into paying customers. I think he's just sniffing glue if he believes that because where you get access to people at the low, low end beginning, when they're starting out, new company, university people, if those people are not using your product, they're not going to care less about it. Like I've not really grown up in Red Hatland, I'm never going to use it everywhere. Where I encounter it, I try to get rid of it because it's just not my thing. I run Debian or Ubuntu-based systems. If every company now looking at this goes, right, so you know, there we have 300 developer machines that we need to run. We now have to either switch them to stream, which is a completely irrelevant product. If you have to go to a fully supported thing, I don't know what they're thinking that people are going to do here. Who is going to be testing an upstream, almost rolling release for a stable, real release? It's absolute nonsense. No one wants that. People want is a free version of REL, and if they don't like people who have no free version of REL, they shouldn't be surprised when people no longer even talk about their product because nobody can afford to use it, because you just can't roll that out in a dev environment. And they say, oh, yeah, you can get so many ex licenses as a test or whatever. That's just not relevant in a business thing. You've got like developing QA pre-production and then production, productions a bit where you might have REL. But all those four of their steps that all have to mirror that identical setup, you're no way you're going to run REL on them. That's just absolute lunacy. Nobody would do that. Yeah, I agree. And I think the thing that made CentOS so successful before Red Hat bought it was that you could learn CentOS knowing that you were learning relevant skills. And surely to Red Hat, they knew that they were the upstream developers. They were in charge of where CentOS was going through REL. In the same way that they sponsor kernel development, you know, they're in charge of those features that land in the kernel and they've got significant influence over that. And so they understand that kind of symbiotic relationship or they used to understand that. And I do think it did pay off maybe not in the clients that they're expecting because this is a generational thing. And one of the things that I've brought up time and time again, I think is so important to like an open source free software mentality is that you're buying into something that you have some degree of control over. So in using the old CentOS and in using the modern rebuilds of REL, you knew that or even REL, you knew that you always had a backup. You had a way of keeping Red Hat honest if they made bad decisions. And I think that's so vital to the success of Linux and open source. And that's what changes this, this is where things change here. Yeah. Plus, how else are you going to get to be an absolute expert on REL 5 unless you've got a CentOS VPS from all that time ago, A.C. Yeah. I'm still working on it. It's all your fault. Great. They don't think they can get a license out of your toll. It's all those ePEL dependencies. What are they again? Okay. This episode is sponsored by HelloFresh. With HelloFresh, you get FarmFresh pre-proportioned ingredients and seasonal recipes delivered right to your doorstep. To eat well this summer, HelloFresh's menu features calorie smart and protein smart lunch and dinner options, plus new vegan dinners to choose from. HelloFresh makes it easy to reach your food goals with flavorful recipes that leave you feeling satisfied. Did you know HelloFresh offers more than just delicious dinners? It's now easier than ever to skip that extra grocery store run by adding snacks, sides, and more to your weekly order. Simply shop HelloFresh Market and take your pick from a curated selection of over a hundred items. My car host Dalton from Linux After Dark tried HelloFresh and said having all the ingredients together and correctly portioned is super convenient and the great meal selection made it tons of fun to try out new ingredients and techniques. So support the show and go to hellofresh.com slash latenightlinux50 and use code latenightlinux50 for 50% off plus free shipping. That's hellofresh.com slash latenightlinux50 and code latenightlinux50 for 50% off plus free shipping. Onto a bit of admin then. First of all, thank you everyone who supports us with PayPal and Patreon. We really do appreciate that. If you want to join us, people can go to latenightlinux.com slash support and remember for various amounts on Patreon, you can get an advert free RSS feed of either just this show or all the shows in the latenightlinux family. And if you want to get in contact, you can email show at latenightlinux.com. The wizard wheeze in all of this is that Red Hat are still publishing images to the public clouds where you can spin up a Red Hat VM with all of the Red Hat binaries on it. And the GPL gives you access to the source. So what Rocky, for example, have said that they might do, and I hope I'm not giving you any secrets here, but what they expect to do is spin up a machine in AWS. They now pay for that machine. They have a legal obligation or a legal right to get access to the source. They ask for the source then they switch the machine off. They have access to the source that was used to build the exact packages that they need. And there's nothing that Red Hat could do about it. So all of this is for nothing, lulls. Well, it's whack-a-mole, isn't it? It's what Red Hat are trying to set up here because Alma and Rocky and the others are going to continue to be able to operate. It's just going to be harder. It's just going to be more of a ballake. And the Rocky blog post about this that said they might well go that AWS route suggests that that's probably the best way because they can automate that a lot more easily than some of the other ways they can do it. But worst case scenario, they could just pick through the central stream source, which is just that's not sustainable as far as I can see. But that's the kind of worst case scenario. The end game is what concerns me the most here because if Rocky do spin up a VM in AWS and get access to the source that way, how do Red Hat stop them doing that? Well, they have to stop being the GPL at that point. And will that happen really? Will they really be able to say to justify not being GPL anymore, just to stop people like Rocky from getting access to the source? Will they have special builds that go up onto AWS that are not bound by the GPL? Can they even do that? I don't know. But it just seems like a real mess and it's going to be quite ugly over the next couple of years. Well, no, they can't do that because so much of the software that Relis built from is GPL. The kernel, for example, is GPL too. So what they're going to do right there on fucking kernel, I don't think so. So they just can't get away with that. They can try and change the license of stuff that they have written, but that just doesn't seem likely to me. I suppose another side of it is that I do have a lot of sympathy for the situation that Red Hat finds itself in and people using their work without contributing back. I understand the natural frustration that brings and it's a conversation that I think we've all had a time and time over when we talk about how do you make money out of open source? I think they've made quite a bit. I know they have. And of course, it's crazy when they've made so much money and sold for that amount to IBM. But this is the same old argument we've had ever since the beginning when you meet people that don't know anything about the next open source that's like, oh, you give it all away for free. So how do you make money out of it? That's still unresolved. It's still unresolved for Red Hat. And I think that's, that is a problem that isn't going to go away as much as we like to frame Red Hat as the bad people in this. That's the saddest part about this, though, that Red Hat was always the poster chart for look, this is how you make money and stay totally open source. And now they're tightening the screws on the open source side of things because they're not making quite enough billions. It's just galling really that, you know, maybe it's the sort of slightly anti-capitalist whatever side of me, even though I am a small business person. So I shouldn't be very anti-capitalist, should I? But nevertheless, I don't think anyone should be a billionaire. I think it's unethical to be a billionaire and companies shouldn't get that big. This idea of chasing growth constantly is just a fallacy like on a planet that is finite in its resources. Yeah. Have you been reading my manifesto again? No, I think you're absolutely right. I think you've seen in everything from Google to Twitter to Reddit recently, you know, this is exactly the same problem of chasing growth when Twitter was fine just as it was. Why did it need exponential growth? The problem is public companies realistically because of shareholders who don't give a shit about, you know, the values of a company or the values of a community, they just want to return on their investment and it needs to be X percent and to the moon and all that nonsense. So it's sad that, as you say, this is them sort of, you know, that thing of, you become the villain after you're the hero sort of thing, it's like, it's a very sad sort of turn of events because they might think they're totally in the right of this, but it makes them look so bad, I think. Well, it's inshitification of open source, isn't it? Yeah. In episode two, three, one, we talked about things that we were worried about and a couple of us talked about open source becoming a cost center instead of a, I don't know, instead of just being overlooked or being ignored and that seems to be on the money. Like we were right, this should be a worry, red hat, I've worried about it as well. But this is all part of a bigger picture. You mentioned Twitter and companies are generally laying off people, certain cloud companies might not be advertising on certain podcasts anymore, now that we're into Q3. Thanks for that. And things are tightening up, aren't they? The end of free money and all that. Like the walls are closing in on everything, seemingly, you know, like the very reason that people are having to go to extreme measures like getting the automated programmatic ads, for example. Who are you thinking of? No idea. Even on crypto or doing all sorts of out there stuff to think outside of the box to continue because I don't know whether the world was going to shit before COVID and COVID kind of accelerated it and, you know, it's just been a really strange decade so far, hasn't it? It's the bottom line and it feels like this is just part of that broader picture to me that Red Hat had to do something or IBM or whoever it is who made this decision. And it's not just some overnight thing. It's been trending this way over the last few years, of the last five years, arguably. Late stage capitalism seems to be where we're headed. And if we continue on this road and now we are left wing bias because we're in open source to begin with, but if things continue on this growth above all other things pattern that we see in general life reflected in technology, reflected in the supermarket prices, reflected in energy prices, absolutely everything is going to shit. And it's going to have to end. And I don't think it's going to end like peacefully. I think it's going to be people taking to the streets, not necessarily about Red Hat licenses, I admit. But this is the, this is the thin end of the wedge. Right. Now be careful, everyone, because we're about to spray orange paint out of your phone as you listen to this forecast. It does make you think that maybe those just up oil people have got a point, right? And the other protest groups that do extreme shit, it feels like, I don't know, are we making too much out of just a simple business decision by an IT company? Never. Don't be ridiculous. I think the problem here is the fact that it does so much damage at the low level that I don't think we'll see the results of this for maybe five years. But then it'll very much be a case of, do you remember Red Hat? And there'll be just a niche player for banks or something like that. And every other place that uses this is going to be using Debian or Arch or even Ubuntu where, you know, the advantage of an Ubuntu system is it's the exact same whether you get a license or whether you don't, you know, and there's a lot to be said for having to do bloody license management and automating your updates through their, you know, special software to allow you to have the thing. If you can just go out with a frictionless ability to do open source development or hosting or work or whatever the project you have to work on, be it the large Hadron collider to be your local small office IT shop without that friction of having to go, Oh God, I want to have a license for this thing. That is the entire key sourcing of fast, if you ask me, and Red Hat have shot themselves in a foot with this. And I think in five years time they might realize that. You think that Red Hat is going to end up a punchline like Oracle? Yeah, I really do. And I feel sorry for people who work in there who believe in fast and, you know, they're doing great work. I mean, look, they do huge amounts of stuff all the way through the pipeline. But if that dries up and if they're purely driven by this whole profit above all else, then they have died internally. Well, I'm glad you brought up a Ubuntu, right? And Graham, you might want to go and make a cup of tea or something because here it comes. Canonical is not immune from this same shit. And it's not to the same extent, but with their Ubuntu pro updates thing, their paywalling the updates. And as far as I can tell, I may be wrong and please correct me if I'm wrong, either Graham or one of you or someone listening. But as far as I am aware, they are only offering the source code for those updates to paying customers. Your ancient software, though, that's the difference of this. This is not 20204, it's not even 20404, it's 1804 which is just gone dead and beneath. So you know what, I don't have a problem with that. That's fine because why are you using it? Update it. Well, no, I'm talking about things like image magic and stuff. That was the first, that's only stuff that's old though. So if you keep up with the, like, free LTS versions, you're, you're golden. You don't have to worry about this. But like I said, it's nowhere near on the same scale as Red Hat, but it is the same sort of thing. And it worries me that if, let's just say the people at Canonical, could be any of the people at Canonical, see this, see Red Hat, pull it off, get away with it, revenue goes up as people say, our bollocks to it will have to just pay for Red Hat licenses then. And what is that going to make the space man think, you know, maybe he might be inspired to have a bit of this is my worry. The only thing I would say against that is that at least there's Debian, at least there is arch. I mean, Debian realistically arched, lunatics, Graham, but yeah, I mean, at least that does exist, at least there is that ability to do that where you can fund, you can pay people to work on the stuff and it goes back to everybody and it's a simpler contract to have with them. But you go to arch before our Pinsuzer, who? Just alienate like a third of our audience well done. I forget about them, okay, yes, you are right, they do exist and I did forget and I use them for years so I should not forget them, but I do. Well, and they have been making the point very much that they are committed to our Pinsuz and not going to do this shit. It does feel a bit harsh though, given a company that has been sold as many times as they have. To be fair to Suzer, they do seem pretty committed to our Pinsuz, it doesn't seem like just rhetoric, but for some reason we just don't care about Suzer, do we? Honestly, forget about them too. Yeah, I know and we really shouldn't because- No, we shouldn't, you're right. Look, when we do episodes about them, and you know, I've done a lot of episodes of a lot of different shows over the years and I've looked at the numbers and everything and when you do an episode about Suzer, open Suzer, whatever, you get a lot of downloads because people just don't talk about them, at least in the English speaking world. So there's going to be loads of people who are right to us now saying, yes, I use up Suzer. It's great. I mean, Adam picked for one, who's just fucking obsessed. Don't encourage the man. God, sake. Yeah, if it's not sailfish, it's bloody- Yeah, I'm Suzer. We love you, Adam. Anyway, so what's next? What is Red Hat or IBM's Red Hat, as people like to say, going to do next, like what about Ansible and loads of other stuff that they do, are they going to do the same shit or is it just going to be real? I mean, I'd be okay with them taking out Ansible because I don't like it and I like Suzer Stacks. So I mean, if they could hobble that, that'd be great. I think this is probably a change of attitude for Red Hat. And I know everybody knows, but I mean, everything changes and I think this is one of those moments where things are changing for Linux and open source and will readjust and it also provides a huge number of opportunities for people who want to kind of fill the void and do other things and maybe fill the gap that Red Hat had. I mean, I mean, new projects as well as more established ones. I mean, that's what's always happened. So I think it may end up creating a more healthy and diverse kind of open source community, especially for kind of enterprise level Linux and maybe Red Hat will just become one of those kind of, I don't know, guardians in the distance, like Oracle. I mean, God. And you said that about Oracle before being a joke, but Oracle is huge, they're hugely successful. Larry, I listened to one of the most richest people on the planet. So, you know, whatever they're doing is working, even if we don't find it exciting in the world of capitalism that is. Yeah. We better get out of here then. We'll be back next week when we'll have some discoveries and who knows what. But until then, I've been John. I've been Salem. I've been Graham. And I've been Will. See you later.