637
this post was submitted on 08 May 2025
637 points (100.0% liked)
Technology
70199 readers
3385 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Man could you imagine what proxmox would be if that project got just a tenth of the money VMware got?
Classic prisoners dilemma. Nobody wants to invest in proxmox because not enough people invest in proxmox.
Honestly I think if Proxmox got VMWare money then they’d become stuffed to the gills with business sharks and probably go the same route eventually.
That is not a Proxmox problem, that is a capitalism problem.
*Humanity problem.
There are some solutions invented, but they require work and revolutionary wars. And the functioning system, I think, will be as close to ancap as to Trotskyism. Won't be clearly "socialist".
No, this is not a humanity problem. This is a capitalism problem. Companies are not beholding to their customers, they are beholden to stock owners. It is no longer in their best interest to make customers happy, it’s in their best interest to provide ROI for their investors. Every software product hits a point of diminishing returns. There are no new amazing features to woo new customers, it is a mature product that only has incremental features. When this happens, you either flip to a subscription model and parasitize your user base, or sell to another vendor, management group, or some other entity who does it after you’ve been paid out. If we had better controls on mergers and buyouts there would be active competition to foster diversity and keep prices down, but when companies buy all their competition and all of the small companies who make products and enhancements for their base, it’s a lose lose situation for the end users. This is my jaded two cents after a quarter century of being in the IT/AEC field in the direct line of this enshittification process from multiple companies across the spectrum.
I don't think you realize how much of an improvement this is over other really existent options.
One can be a serf, or a slave, or a city dweller in a privilege-based society, or a peasant in some despotic kingdom. The list of options is long, none are good.
That's a very eurocentric view. Most of the world outside of Europe and imperial Asia was communist or socialist tribes until Europe went and colonized everything. And they were doing pretty damn good at colonizing.
Remember the reason colonists conquered lands so easily was largely because they out-armed them. The tribes had no need for such advanced weapons until colonists arrived with them.
Name one for this wild statement.
That's another aspect - building something and not getting overrun by those more considerate.
Proxmox is already perfect (for my use case)
Yes, that’s a more correct use of “prisoners dilemma:” a choice to either cooperate or defect. Origin below, for the curious.
The dilemma
Two prisoners are interrogated in separate rooms. Each is asked to snitch in exchange for a reduced sentence.
Because they’re separated, the prisoners can’t coordinate, but each knows the other is offered the same deal and the interrogator will only offer bargains that increase their combined years of imprisonment.
For example, “house wins” if snitch gets -2 years and snitchee gets +3 years, since interrogator would net +1 year from the deal.
So what will each prisoner do?
The result
Of course, the best outcome overall is for neither to snitch, and the worst is for both to snitch.
The Nobel-Prize-winning observation was that any prisoner faced with this dilemma (once) will always net a lesser sentence if they snitch than if they don’t, no matter what the other decides.
In other words, two perfect players of this game will always arrive at the worst result (assuming they only expect to play once). This principle came to be known as the Nash equilibrium.
Applications
The result above sounds bleak because it is, but real-world analogs of this game are rarely one-offs and thus entail trust, mutuality, etc.
For example, if the prisoners expect to play this game an indeterminate number of times, the strategy above nearly always loses (the optimal strategy, in case you’re wondering, is called “tit-for-tat” and entails simply doing whatever your opponent did last round).
The study of such logic problems and the strategies to solve them is called game theory.
Edit: fixed typo, added headings and links
Everybody is moving to Openshift or public cloud
Openshift is a kubernetes platform isn't it?
There's still a need for real VMs, and I didn't think openshift filled that.
Yeah, it's a distro of kubernetes.
Most apps run best as a container, but for appliances and legacy apps they have Openshift virtualization which runs VMs in the cluster by running KVM inside of docker.
The open source tech there is called Kubevirt. All VMs are 1st class citizens in the kubernetes API, so it is actually easier to run than VMware/Proxmox if you already have a Kubernetes cluster and you're not doing complex stuff with qcow images or VM migrations.
I use both containers and VMs a lot with Kubernetes at work.
There’s Openshift Virtualization included, which is based on the upstream kubevirt project. You’re essentially running VMs in containers and managing them (mostly) like the other container workloads in the environment.
Interesting...I'm using proxmox at home but running my containers in a VM. Looks like there's an openshift community edition...I may have to check this out.
I'm not a sys admin by trade (networking), but my opinions at least have some weight where I work.
I imagine being redhat based, I could run FRR at the hypervisor level. For that matter being kubernetes I can use calico. Holy shit this could be awesome. I need to play.
Suse has been trying pretty hard with Harvester. KVM-based, VMs-as-k8s-pods which leverages all existing k8s tooling, as well as the same multi-cluster federation as RKE2.
Seems pretty great from afar, though it's very much under active development.