For SBC, you can’t beat Raspberry Pi. The ecosystem is just there and the support outclasses every other board.
For hardware based on SBCs, Pine64 hands down. Devices like the Pinebook and Pinetab are SBCs in a hardware shell and as such should feel like cheap gadgets, but their build quality is excellent and these feel like premium devices. I have just started messing with the Pinetab 2 and it feels like a device 3x its price, to the degree that I don’t mind that the drivers and software for it are still a work in progress.
God, tell me about it. I did not fully appreciate the Pi until the Beagle, which has an ecosystem that seems to be following some branch of chaos theory when it comes to organization.
Pine64: I honestly regret I didn’t follow up on this more before now because I had no idea about the Pinebook and Pinetab and I’ve been thinking about diy tablets, since diy laptops are still–really not a thing and it occurred to me just recently to see what’s up with open source tablets. I use a kindle for reading but when I went back to school, most of my books aren’t really Kindle-compatible so I bought a Galaxy Tab Ultra (10 inch, as eyesight) both so I could use Kindle search functions and a readable text size and so I blow up the diagrams. It wasn’t as horrendously expensive as it could have been because, like my phone, I trade in yearly to upgrade, not because i need to but because–depressingly–it’s more affordable when I can get max trade-in value and watch carefully for Samsung’s random discounts.
So yes, I am excited about this. My tablet is a very different use case from my phone (which no, no way to switch to open-source or Linux there at this point); migrating to an open source tablet is actually a possibility. So very cool.
Do yourself a favor and nab Pinetab 2. The wifi and bluetooth drivers aren’t ready yet (you’ll need a dongle or to tether a phone,) but that’s part of the fun: you can join the Discord channel and watch the discussions and commits happening in real time.
That’s because they sell at community prices for little to no profit, either at cost or close to it. They’ve talked about eventually trying to get their prices into retail outlets with a retail markup, which would also pay for retail-level support rather than community support.
In other words, if you buy community, you’re buying just the hardware, and the community provided the software.
I use podman for almost everything. Especially since it’s working rootless. BUT I am also clearly swimming against the tide there. Everyone else in the company uses docker and I typically can’t just take their docker-compose setups 1:1 over to podman. First, because they often rely on having root and second, because they use docker specific hacks (like some internal hostname you can use to access the host from within docker). Since I am not a fan of docker-compose anyway, I don’t care that much … I would have built my own setup with docker as well.
On my server I have a lot less trouble with podman than I had with docker. I run quite a lot of services there, and the docker proxy (and sometime the daemon) always started to act up after a while, causing individual containers to no longer properly receive traffic and me no long being able to control them. With podman all of that just works. And I have systemd managing the container lifetimes instead of some blackbox.
I want full control over which containers launch when. I also typically have a different requirement in which network a container runs and I want to re-use existing databases instead of spinning up a new one for each service. I want specific container names. And so on.
You can tell which service depends on which in compose, you can create, specify and set networks and add containers to them, you can keep a central database and just add the network of it to your new services, and you can also specify a container name.
As I see it (and for my compose usage), everything you mentioned works in compose.
Besides, what is your alternative? Do you just use the docker cli? I personally found that to be way less flexible than compose.
You can tell which service depends on which in compose, you can create, specify and set networks and add containers to them, you can keep a central database and just add the network of it to your new services, and you can also specify a container name.
The point is, if I get a compose file, all of that is already wired up with expectations of the maintainer. When I start heavily modifying it, I end up with an unmaintainable mess. So I rather look into what the service(s) actually require and build it for my use case.
Besides, what is your alternative?
The CLI, yes. And for my own server Ansible. But the semantics of the ansible module are identical to the CLI. Knowing the CLI by heart gets me much further than knowing docker-compose by heart. (Actually, I would have to look into the manual for docker-compose all the time, while I can simply do podman --help to see what parameters it needs, if I forgot something.)
Compose is simpler, and has a much easier base use case, but we've found it more functional as a dev tool to get the service running before making a full deployment config, rather than as an effective production solution.
Developing against k8s would kill me. I want my services and debugger running locally and don’t want to deploy shit first. In my current local setup it doesn’t matter if I spin up a service as container (because I just need it doing its thing) or if I spin it up with debugger attached in the IDE because I am developing (or debuging) it. I can fully mix-and-match at nearly every layer of the system.
Our shared dev, stage and prod systems are also fully k8s. Not with helm though. For our own stuff we have an operator with CRD, so we can easily define our business services without much boilerplate and still be consistent across the teams. The different configs are built using kustomize as part of our CD pipeline.
I must say, I have mixed feelings about it. When Gnome 3 replaced Gnome 2, I just hated it. It was missing features in every aspect, and the ergonomics were… questionable at best. Over time, modern Gnome evolved and since version 42, I think it’s a modern, pretty desktop environment. It is clean and readable on the eyes, looks fancy with all those animations, and there are amazing apps with almost minimalist approach, really useful, nicely integrated into a unified design language. I ran Fedora Silverblue for almost a year now, and it took me about 6 weeks to get used to the modern Gnome workflow. It’s just that different. And for a while, I even began to think that I really like it and that it might be my favourite desktop environment now. But lately… I just start to think that with a simple, traditional DE like XFCE, it would be way easier to manage many open applications and windows, and those fancy animations start to really annoy me. I think I have explored Gnome enough to now think that I prefer the oldschool way. I’ll be on something with XFCE soon.
I manage them by not. My configs are gone when I wipe my drive and I simply recreate them from memory. Things get forgotten, new things get changed. Holding on to the past too tightly will make you unable to leave it.
I’ve struggled to put in words my stance on this, but you said it well. If I backed up my configs, I would get stuck doing things the exact same way for ever. If I backed up my configs I’d be still using Vim with Vungle plugins, now I use Neovim with Packer plugins. I would be still using urxvt with powerline-status bar, now I use Alacritty with starship status. I’d be still using my old favourite Inconsolata font, now I use Fantasque for everything.
There are always newer (and sometimes better) and certainly different ways of tweaking your PC to suit your needs. If you hold on too tight to your old configs, you might miss out on discovering the next cool thing to enhance your experience.
Note: there are of course some home dir things I definitely keep backed up that are irreplaceable, like SSH private keys, GPG keyrings and private keystores, and even my Firefox profile directory.
There should be no dotfiles outside of home directories so I assume you mean a config file. In those cases, git and symlinks are a great option. Make a config directory in your home dir and organize it however you want. Include config files for the tools you’re interested in, commit them to git for backups and then symlink/hardlink the file to the expected path for the application.
Whole situation is ridiculous. People can’t expect enterprise features and support infrastructure for free. But enterprises need to offer more price tiers.
Does Red Hat have anything you couldn’t install in any linux distro?
Can you install Satellite servers on your fleet of Ubuntu machines? OpenShift isn’t free. I don’t think there’s anything that RHEL does that any other enterprise vendor can’t do. And I don’t support Red Hat (IBM) closing access to the source RPMs. But it costs money for vendors to develop their enterprise management platforms, the storage and bandwidth for geo-cached mirrors of updates, and all that. And if you’re in an organization with a fleet of thousands of installations you need enterprise management platform.
Alma sells support IIRC don’t they?
Exactly. It costs Alma money to have the resources to do that. So customers will need to pay the support costs to keep Alma viable. Just like with RedHat. But enterprises a freaking out about needing a new free enterprise distro, because RH is too expensive to license on thousands of machines. So RH should be finding more flexible price models, instead of trying to squeeze out competition.
Use Rauncher from SUSE instead, they may be a corp but they’re committed to Free Software at the moment.
The free stuff is subsidized by enterprise subscriptions (and YaST sucks). That’s all I’m saying. Alma has a free option and paid subscription. So does Rocky. So does Ubuntu. So does Suse. RedHat has free stuff too. (CentOS Stream, Fedora, and free RHEL developer license, and ubi). If you want the enterprise features of RedHat, pay the enterprise price. And if you don’t want to (I sure don’t), then use something else, because like you said we have choices.
capitalism propaganda won’t fly this time around
You’re way off the mark here. I haven’t used RH in like 20 years, since they first introduced RHEL and killed its predecessor because screw that greedy shit. But I also haven’t been trying to use 1:1 rebuilds of RHEL. Employers have made us use CentOS to because customers use RedHat but no we won’t pay for RedHat but also no we can’t use CentOS because no enterprise management to push security updates without the application updates but also no we won’t pay for RedHat. It’s stupid. Either pay for RedHat because you need it, or shut up and move onto something that isn’t RedHat.
Not sure what direction you’re leaning with this one. From here:
OKD is the upstream project of Red Hat OpenShift, optimized for continuous application development and deployment.
So it’s the CentOS Stream of OpenShift. And just like CentOS Stream is openly available while Red Hat Enterprise is not, OKD is openly available while OpenShift is not. So revenue from OpenShift is used to support the development of OKD, just like with RHEL and CentOS Stream.
Raphael is blindly ignoring that I’ve literally said I don’t support RedHat closing access to their sources and that I’m in here applauding Alma for moving away from their dependence on a greedy corporation. Somehow my acknowledging that enterprise support costs money to provide, and that the resources to develop and distribute FOSS aren’t free, means to him that I’m just blindly opposed to FOSS and that I’m pro-corporation.
In this thread I’ve said don’t use RedHat because they’re being dickbags, also maybe don’t use clones of RHEL because they then see you as a customer who isn’t paying them, and also if you need enterprise support it costs money so pay for it (because it also pays for the FOSS projects that these companies foster and contribute to).
Please link/quote to me defending what RedHat is doing with access to their source repos. I’ve said repeatedly in this thread (and in your other satire s/RedHat/Linux/) post that we should all stop using RedHat and stop creating a market for Red Hat as the de facto standard, because I do not support what they’re doing with access to their source repos.
I always thought the Red Hat business model was based around service and support with the OS being a secondary product which is why the free forks existed. When did the OS become the product?
When other companies made a business out of building a clone distro from the source RPMs with trademarks removed.and selling support contracts for it. Oracle being the absolute worst about it. Fuck Oracle.
What even is the point you think you’re arguing against with me? Someone asked when RedHat decided to change aspects of their business model and I provided an answer. I didn’t say I agree with it. Even in the face of me saying literally “I don’t support RedHat” and “I haven’t used RH in like 20 years” you seem really dedicated to convincing yourself that I just love RedHat and think they can do no wrong. Geerling is right. RedHat is stupid, and IBM is killing whatever was left of the brand. There are many, many alternatives to RedHat. Both free and commercial. Lets use them instead of clinging to RedHat-but-not-RedHat-because-we-don’t-want-to-pay-RedHat.
This is an ethical problem. You can rebrand RHEL as Oracle Linux, copy Red Hat’s business model, and dump the work on Red Hat, but should you do that? Nobody at Oracle stopped to think about that question. According to the software license, sure. But is it ethical to exploit the good will of foss to these extremes? I would personally say no, especially considering Red Hat reinvests money into foss. Money now in the pockets of Oracle.
Imagine you develop a foss and license the software under MIT like everybody else does. Now a company roams around and makes your foss a core part of their business after a massive success. Perhaps they’re the next FAANG company, who knows. You generated billions of dollars in revenue for that company, and never saw any of that. You regret your careless licensing decision, yeah. But at the end of the day, you will feel exploited.
The ability to do that is literally one of the core purposes of the license.
You don't and can't own derivative works of GPL projects. Oracle has the exact same right to resell an exact copy as red hat does of the original project.
I agree. That’s why I said I don’t support RedHat’s choice to close off access to their source to non-customers. RedHat is still complying with their end of the license though, by keeping source access open to the recipients of their binary distribution. This is how Rocky is aiming to maintain 1:1 binary. RedHat is still publishing their Universal Base Image Docker image, so they need to keep source for that open, and Rocky will be using that method to get sources.
My stance is that we as users should be moving on from RedHat and RedHat derivatives, or just pay for RedHat if that’s what we want. Continuing to use derivatives will just convince RedHat we’ll all pay up if they can just get rid of those other options.
Having a prerequisite contract that allows them to punish customers who exercise their rights to the software is not complying with the license. Selling the code is allowed (though if it were written in the modern era where distribution costs are negligible I'm not sure it would be. Predicating distribution on other contracts that limit your rights is not.
You don’t have a right to their sources until they distribute to you. And they have the right to choose to whom they do business (as long as they’re not violatong discrimination laws). If they’ve distributed their product to you they have to give you the source, and they will. And if you distribute that source, they won’t distribute the next release to you, so you won’t have license to those subsequent sources. Compliant with the letter, not the spirit. It’s shitty. And I think we should accordingly not do business with RedHat. That’s what Alma is chosing here, by pivoting to no longer being 1:1 source rebuild distribution. Rocky is trying to hold onto the model that RedHat is trying to kill, by finding ways to still be a non-paying recipient of an RH distribution, requiring they be given access to source. I think we can expect RedHat to try and find a way to cut that off. Then Rocky will either pivot or die. But I wouldn’t want to wait and see and then be screwed. I would want to break all dependence on an entity intent on breaking me. And I’d be wary of recommending Rocky as a migration from CentOS because of RedHat’s actions.
It's not compliant with the letter. The GPL doesn't allow you to place other limitations on someone to receive the source. "You have the rights the GPL grants you, but we can punish you for exercising it" is a blatant and egregious breach of the GPL.
They're not betting that there's a 1 in a billion chance that they're right. They know with absolute certainty that they aren't even in the neighborhood of complying with the license. They're betting that no one is willing to spend the massive amount of money it would take to punish them for their stolen code.
Just to reiterate, I don’t think RH is in the right here.
They’re “punishing” you by not taking any more of your money for future versions. Maybe we’ll see a court case out of it to settle the question but I doubt it. But consider you are a customer, and you have to ship RH binaries with your application. In order to comply with the license you must also make the source available. RedHat can’t stop you from doing so, they just won’t give you access to any more updates (and stop taking your money). So now you can’t ship security updates to your customer. So now you have a legal liability by being a RedHat customer. Either you fail to comply with GPL yourself for the sake of updates, or you expose your customer to known security risks because you compiled with the GPL. So … why do business with RedHat anymore? Explain this problem to your customers why you can’t certify on RedHat anymore.
"If you distribute the code you're entitled to distribute we can terminate your contract" is identical to "if you distribute the code you're entitled to distribute we can charge you money". They're additional restrictions that are unconditionally not under any circumstances allowed by the GPL. You cannot restrict redistribution in any way for any reason outside of the GPL terms.
The second you do so, you are no longer covered under the GPL and everything you're distributing is copyright infringement.
“If you distribute the code you’re entitled to distribute we can terminate your contract” is identical to “if you distribute the code you’re entitled to distribute we can charge you money”
I’m not a lawyer, but I categorically disagree that those two statements are the same. If someone takes RedHat to court and wins, fantastic. But as I’ve said, I wouldn’t make business plans that rely on
Maybe you should read the rest of my comments in this post.
You mean enterprise features mostly developed by the community under the GPL?
Enterprise features like the update management server to keep a fleet of thousands of machines patched with only security updates. Infrastructure like geo-located mirrors of the update repositories (not volunteer mirrors like universities around the world mirroring kernel.org and centos.org and eclipse.org etc). Support service like on-call staff to pick up the phone whenever you call. Those things cost money to provide. If you know of a distribution which provides all that for free, please let me know. If you need that level of support, pay for it instead of trying to find a freebie around it.
Why shouldn’t they be free?
I assume you like to be paid for your work. You might be surprised to learn that revenues from that commercial support pay for the free stuff.
Red Hat is not owed our money just because they’re a business, they however do owe the community strict adherence to the GPL and if they’re not downright violating it here, they are most certainly trying to do an end-run around it.
I agree with you, and everyone else who thinks I need to be told this. Which is why I’ve been advocating in this thread for users to drop RedHat like I did 20ish years ago when they first replaced their free desktop with RedHat Enterprise. And further to move away from source-rebuild distributions because RedHat has clearly stated that they see these users as lost revenue and are taking these actions as a way to “claim” those customers by removing the options. And I certainly wouldn’t pay RedHat after shitting on at least the spirit of the GPL (and I’ll be happy if someone sues them successfully to set a precedent about the letter of the GPL).
It seems that you, and many other corpo-apologists, have been brainwashed into a commercial software mode of thinking where you get the basic software for free, and then pay for extra features. Your “price tiers” remark certainly indicates that you don’t really understand what open source software is about.
I’m so apologetic to these corporations that I’m literally commenting in here to stop buying from them! Such an apologist! When RedHat killed CentOS, I recommended at my office that we switch all CentOS usage to Ubuntu. When they announced this last move of closing the RHEL source to non-customers and the user agreements that they’ll terminate your contract if you distribute the sources, I recommended we don’t even consider a source rebuild distribution either, because I don’t want us to be caught with having to transition to another distribution if RedHat finds a way to kill off the source for UBI to non-customers (how Rocky is planning to stay compatible as a source-rebuild distribution). And it seems Canonical is killing their free distribution too, for organizations of more than 5, so I have to reconsider Ubuntu now (which sucks because WSL was really helping my case to use Ubuntu) Maybe now that Alma is moving away from the RHEL source rebuild model I can recommend Alma, maybe can get a WSL package of Alma. If the other distributions stop caring about RHEL compatibility, then RHEL will cease to be the de facto standard. And we can all rejoice. Seriously why would anyone want to sell a product they built on RHEL now. If they have to redistribute a library they got from RHEL, then they are faced with either being in violation of GPL or losing access to security updates from RHEL (meaning they’ll be exposing their own customers to security risks). It’s a legal lose-lose to be a RHEL customer now.
As for support infrastructure: nobody is expecting Red Hat to give tech support to AlmaLinux and RockyLinux users.
Fucking duh. I never implied that. I said if you’re trying to make use of enterprise features that cost money to provide, you should pay for them. I personally get by just fine with support from GitHub issues/discussions, Gitter/Slack channels, IRC, and Usenet.
The generally accepted position that these free rebuilds are just funnels churning out RHEL experts and turning into sales just isn’t reality. I wish we lived in that world, but it’s not how it actually plays out. Instead, we’ve found a group of users, many of whom belong to large or very large IT organizations, that want the stability, lifecycle and hardware ecosystem of RHEL without having to actually support the maintainers, engineers, writers, and many more roles that create it. These users also have decided not to use one of the many other Linux distributions.
This is the perspective that is informing RedHat’s decision making on the matter. It doesn’t matter that you and I know the people using Alma and Rocky, and previously CentOS, won’t switch to paid RHEL users if those options are gone.
You conflate two things: on one hand there is A: “being able to use a Linux distribution that’s binary compatible with RHEL”, on the other hand there is B: “having a support contract and access to technical support”.
I can see how you would see my comments as conflating the two. It was not my intention to do so.
I see no issue with A, “the software”, being free, and I see no issue with B, “the support”, being not free. This is how it has been since Red Hat came into existence, yet you’re telling me here that A shouldn’t exist.
I’m not saying they shouldn’t exist, RedHat is saying that. I’m saying given RedHat’s actions, I wouldn’t want to be in the business of trying to fight with them to maintain a source-rebuild distribution or base my own business continuity on them being able to out-maneuver RedHat and continue to exist.
That’s a broken analogy. The existence of a free and legal alternative to RHEL doesn’t mean that Red Hat doesn’t get paid, it just means that a free alternative exists. But big businesses do love support contracts from big reliable vendors, so Red Hat does in fact get paid and their model is quite profitable.
On the other side: is Red Hat cutting a paycheck to all the contributors of the thousands and thousands of tools and utilities that go into RHEL?
It is a fact that big corporations like Canonical, RedHat, and Suse have historically paid full time developers to contribute to and maintain FOSS code. They have to have money to pay those developers. They can’t make a reliable and predictable revenue stream on just the existence of the software itself, so they sell support contracts to pay for it.
On the other side: is Red Hat cutting a paycheck to all the contributors of the thousands and thousands of tools and utilities that go into RHEL?
No, and I never claimed anything close to that. But RedHat is among many Linux distributors who employ developers full time to contribute to and maintain FOSS projects.
Come on now, it’s the other way around. The enormous amount of free software development they have received from the community is what allows them to have this profitable commercial support model in the first place.
Indeed, hence why I think RedHat is ethically in the wrong here.
Yet you provide not a single convincing argument why that should be the case. What kind of artificial bs label is “enterprise” anyway? It’s just software, and whether it has a label of “enterprise” or “consumer” is irrelevant
I gave examples of what I perceive as enterprise support, you’re free to think those things don’t matter, but maybe tell me who does those things for free. Alma Foundation isn’t some group of benevolent billionaires paying for everything out of their own pockets. If they weren’t receiving donations (be they monetary or services) or revenue, they wouldn’t be able to do what they’re doing.
the only thing that determines whether or not it’s ok for it to be free is the license of the software, and so far the license says that it can be free.
Again, I agree. All the source-rebuild distributions have the right to exist. And if they feel it’s worthwhile to pursue still , good for them and good luck.
I mean … we all agree that RedHat is in the wrong here because the actions of the source-rebuild distributions are protected under the FOSS licenses. We have different reactions and hopes, but we all agree that RedHat is doing wrong. So I don’t understand why you and Raphael are out here calling me an apologist who doesn’t understand OSS.
edit: I do support Linux distribution vendors having the option to do freemium if that’s how they feel they can best deliver, just not the way that RedHat is now trying to do it. And I support people trying to do it in a way that is completely gratis to the users.
Seems like an odd software choice to create actual dotfiles under /etc. Often files there have much the same name as a user's dotfiles but without the dot. Thinking of things like /etc/profile vs ~/.profile and so on.
Without knowing precisely why those decisions were taken (good reason? ignorance? insanity?) it's not clear what steps to take.
Vaguely leaning towards symlinking or hardlinking but precisely what to do and which way around is still unclear.
Do those dotfiles have your user permissions or are they owned by a system account like root?
Then they shouldn't have dots on them (bad design decision) and should be backed up by the system backup/restore mechanism, whatever that might be, not the user's own homedir backup, which is what I assume OP is talking about.
If they're talking about a full system backup including home directories, that's a moot point because I'd expect they'd be included anyway, dot or not.
ODroids are massively underrated. The first Pi we bought is dead. The second one is now so underpowered for what we want that it's been turned into retro arcade machine. It still finds ways to cause problems too. Whereas our first ODroid is still going strong after many years of faithful service. We added an ODroid toaster to the mix a couple of years ago that's also given us zero issues and works wonderfully.
I bought Home Assistant Blue from Ameridroid, which was Home Assistant’s first (and happily still continuing) jump into making Home Assistant more accessible and easy if you weren’t a hobbyist or tinkerer: Odroid N2+ preloaded with Home Assistant OS, a super adorable blue case, and power supply. That was my first experience with that board and with eMMC; 128 GB of it, talk about turning my head (also 4 GB RAM). Honestly, the only reason I didn’t get another is I didn’t have a project that required it; the reason I even found out the Beagles existed was the Open Source Border Router project I wanted to do had it as an option for the walk-through and gave me a reason to test drive.
But I have to agree: I’ve been running it straight for three years now and the Odroid does its job with zero issues. Home Assistant and its parts have given me problems, but Blue (yes, it’s name is Blue, it was just there) never does.
I’m using fedora as my work system, because I have a relatively new laptop that needs the new kernels. Haven’t experienced anything you’re describing. Are you on fedora regular or on sliverblue (the immutable version)? If you’re having issues running the newest kernel, follow the fedora documented way to build and run your own. I did just that when needed a prerelease kernel and it worked out fine. I usually upgrade to a new release by the end of the cycle, so that the new version had 6 months to mature. I never immediately upgrade.
I always found Fedora to be a little unstable for my work use. I switched to CentOS because of that, and that was truly rock solid. I even used CentOS Stream for a while (but switched to Alma and Rocky eventually).
linux
Active
This magazine is from a federated server and may be incomplete. Browse more on the original instance.