All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Decent Cloud - A Chance to Shape the Future of Open Source Cloud
Hello everyone using or providing cloud services,
I’m building a peer-to-peer platform for renting cloud resources. It’s something I’ve been tinkering with for a while, and I’d love your thoughts and early feedback. Before you ask or kill the post -- I'm not earning ANY money out of this, this is a call for action and for participation, it's not a business advertisement and not a "shilling" post.
At this stage I mostly need some interested providers who would be willing to join me in the effort of building out such a platform, e.g., by listing their offerings and helping iron out the kinks in the process.
Here’s the gist of the platform:
Unified API
Through a single interface, it makes it possible to access both brand-new, lesser-known providers (selling extra capacity/GPUs from on-premise or home servers and workstations) and established providers like AWS, Azure, GCP, or Vultr. Ambitious providers (like you) can either broker existing offerings or list their own—and earn money either way. End users win by getting a straightforward, programmable interface to switch providers or use multiple providers at once, boosting fault tolerance in the process.Unified Payment
Crypto payments (available today) enable anyone to rent or provide services without needing a credit card. This keeps the platform open and avoids reliance on any single payment monopoly. Credit card payments are surely possible in the future.Reputation & Reviews
Reputation is the cornerstone of Decent Cloud. Reputation grows gradually when renting resources but can be degraded quickly for bad actors. The community can collectively downgrade a provider’s standing when issues arise, effectively pushing them out of the marketplace. Because data doesn’t lie, it’s easy for users to inspect historical records and provider satisfaction before making a choice.Community-Driven
Everything—from development to governance—is open source, decentralized, and self-sustaining. This is a small attempt to reduce monopolistic control by big-name providers. Nothing more, nothing less. No big marketing budgets here—if you can, please join and help drive the project forward. Your help and contributions matter!
Again, there are no VCs in this project, and no big money yet. You decide the project future. There is no Initial Coin Offering (ICO) either. Tokens are generated fairly every 10 minutes (similar to Bitcoin) and can be claimed by all registered providers, me and you included. If you’re a potential cloud service or VPS provider—or a sub-leaser—you’re invited to jump in, list your offerings, and regularly check in to earn these 10-minute rewards and start building your reputation.
Please share your experiences—either in the project discussions or by replying to this post. I’m happy to answer questions and would love your feedback and ideas. It’s still early, so don’t expect a grand enterprise product just yet—but if you want to help shape how this project grows, hop aboard and let’s see where it goes!
—Yan Liu (https://decent-cloud.org/)
P.S. Check out the whitepaper for more details.
Comments
It's difficult.
Sooo basically you got tired of waiting for wquil to go somewhere and doing a spinoff?
This sounds like caasify's pitch but with less chatgpt and more scamcoin.
Good luck with this project.
As a valued customar I would not use your decent cloud to rent any services.
Fighting vendor lock-in would probably be against providers' interest because they only make money as long as you rent services from them specifically.
So instead of a single payment monopoly, it relies on you and decent cloud and you whatever crypto coin. Interesting thought process.
Based on this Hosthatch would be the number 1 best rated provider by their cult members even though they're kinda ehhhh.
You kind of missed the time of the "blockchain" and "crypto" buzzword. How about you build an AI that finds you the best deal based on user input?
just my $3.50
reguards
One thing I've noticed is that it's almost always near impossible to actually 'use' these types of decentralized cloud projects. Quilibrium for example, the groundbreaking 'decentralized supercomputer'. How do I actually pay money to use this decentralized compute cluster? It's impossible to figure out. Is it even usable at all in its current state?
FileCoin is another one. Decentralized storage, sounds like a great idea right? Good luck trying to figure out how much it costs to store files on the network, or how to use it.
StorJ did pretty well at it honestly
Make it no code for us and you have our business. We idlers pay to use and it be sad day for providers when they see us subletting idlers.
DO IT.
Cassify with blackjack and hookers
Wow, so many thoughtful replies! Thank you all—I’m flattered by the engagement. Let me try to address the questions, and who knows, maybe together we can turn this into something great.
@suut
Do you mean it’s difficult for you to use something like this, or for builders (like me) to create it? The latter part is not a concern for me, I already completed what I thought was the most difficult part. If you have specific concerns or see areas that seem overly challenging, I’d love to hear your thoughts so we can tackle them head-on.
@jfreak53
Interesting observation! I see Decent Cloud as a much simpler approach. This simplicity lets us deliver something useful quickly, focusing on what matters most rather than trying to perfect privacy, E2EE, and other features that, frankly, very few people in the world actually care about today. Decent Cloud focuses on finding the right resources easily, handing reputation, payments, and that's it. Like Airbnb, just for cloud services.
Also, I’m a doer, not a talker. My focus is on creating value for users and providers, not convincing VCs to fund features they think will maximize profits. I care about building something that matters to people, not chasing VC money.
@ralf
Thanks for your candor. I completely understand concerns about potential scams, but rest assured, that’s not my intent. Are you worried about being paid in tokens that might lose value overnight? If so, what if payments were in stablecoins like USDC or USDT? Would credit card payments feel more reliable? I’d love to hear what payment methods you consider trustworthy.
@dedipromo
Thank you for the kind words! If you’re open to collaborating, I’d love to explore how you could contribute or list your offerings on Decent Cloud. Your expertise could help shape the project.
@Rubben
That’s a valid point, but I believe good providers will retain customers as long as they’re fair and reliable. No one switches providers just for fun. Decent Cloud doesn’t make it easier to migrate data or services, so there’s still a level of vendor lock-in. For smaller providers, the real advantage is exposure—they get a fair chance to reach users who might never have found them otherwise.
The platform minimizes reliance on any single entity. Payments use decentralized mechanisms, and we’re planning to integrate multiple options, including fiat, to keep things inclusive. Additionally, the project is fully open-source with a liberal license, so there’s no lock-in—you can fork it and make your own changes if needed.
Great idea! An AI-driven recommendation system is already on our roadmap (check out the whitepaper). It aligns perfectly with the goal of making the platform no-code and user-friendly.
@DataWagon and @Advin
You’re absolutely right—usability has been a pain point in similar projects. That’s why we’re completely focused on building a no-code platform with a seamless onboarding process for both providers and users. Clear pricing and easy service discovery are top priorities to avoid these common pitfalls. Your feedback reinforces this focus, so thank you.
What I'm trying to get from this forum is: what kind of automation do providers here currently use, and how can we best integrate with it to minimize the cost of "trying it out"?
@DrNutella
No-code usage and a great UX are central to our approach. The goal is to empower users to deploy and manage resources without needing technical expertise. If there are specific features or workflows you’d like to see, now is the perfect time to share them. We’re building this for you, and your input will help shape it.
What’s the best way to integrate with your resource allocation backend? I’d love to understand your needs better.
@Void
I wasn’t aware of Cassify—it’s a cool project! That said, I’m confident Decent Cloud will stand out with greater transparency, community input, and usability. As for blackjack and hookers, those would likely be offered separately upon user request.
Thank you all again for your feedback—these discussions are incredibly valuable to me. Let’s keep the conversation going!
>
Disagree. Only scummy providers should fear easy migration. Provide a solid product at a fair price, and people aren't likely to leave. A business that provides good services/support will always have a solid customer base. Clients typically only leave if the service sucks, or they can obtain a higher quality product at a lower enough price the time and expense of migration becomes worth it....
Disagree. LET is probably the user base that would choose to migrate over $1 cost per month difference, should a cheaper option arise. You don't have to be scummy to try to maximize income / customer, look at AWS, Oracle Cloud etc.
Now we could argue whether you think they are scummy or not, but the same thing can be argued about literally every single provider on LET.
Not Caasify being flustered by the mentions haha
This is very complex indeed.
Question: How do you handle cases where the request is to buy cpu from one resources, and gpu from another.
Of course, the gpu (say, running on PCI) is a slave device, and then you will implicitely have to buy CPUfrom the GPU provider as well (at least to drive the GPU). Indeed I have given a rather oversimplified example
But i am curious how you are going to handle challenges like this?
Probably a free unlimited VPN that you can pay for with crypto after you use up the first 10GB or something.
Once critical mass for the project is reached, idle VPS can just be sublet repeatedly until the unused capacity disappears into the nether realm of nested virtualization overhead. VPS will be idling while making full use of the resources, it's the best of all possible worlds!
Then how will he set priorities?
If my app is sleeping and i sublet my cpu and it is doing something and a new request comes in - who gets priority? setting all these up correctly will be very difficult.
Good luck for the project, hope will release soon.
Of course, the gpu (say, running on PCI) is a slave device, and then you will implicitely have to buy CPUfrom the GPU provider as well (at least to drive the GPU). Indeed I have given a rather oversimplified example
I'm not sure I understand the question completely, but I do have an example of a GPU resource renting/subletting here
I assumed that renting the GPU implies that the CPU is rented from the same provider. Was my assumption wrong? And if yes, how would the other setup look like?
If my app is sleeping and i sublet my cpu and it is doing something and a new request comes in - who gets priority? setting all these up correctly will be very difficult.
That would be the standard Linux scheduler deciding. Not much that the user (or me) need to do.
@yanliu8 you should read all the feedback threads Caasify has posted here as your ideas seem similar.
However, the first and most important question is if a user buys a service from a provider through your platform, who's account will be the service allocated to?
As in, do you have accounts with the providers, buy the service and then loan it to the user? Or do you create a new account with each provider for each user and then use that account to buy the service?
There's a lot of reasons why this matters, but I'm interested to know how well you've thought this through
No, your assumption is right. The idea here is that you need to communicate that to the user - unless of course the implicit CPU costs are taken care of by the provider. Then you have to check how the overheads look like - or whether you should just give it to the user to play around.
Cool project with Rust and all. OTOH, the whole depin concept is flawed, as it doesn't solve any real pain points, which are addressed elsewhere by infra brokers. The current default tie to blockchain for payment is a friction for adoption. Is the crypto stuff the reason that you got banned on reddit?
Like I have a public API currently, is the idea to sell my resources for the projects coinage to build an additional translation layer to my API so that the decentralized network can query my backend? The network would have to have access to my API key to be able to do that; so would I have to deliberately build in support for this into my backend where it checks the blockchain for cloud resource requests?
How would you plan it work work with AWS?
How would it be better than someone who builds a centralized platform as a translation layer between multiple cloud providers and you pay the centralized platform for access to provision resources for various providers and manage them in one place?
That’s an excellent question, and I’ve been pondering about it also. Both approaches have drawbacks:
a) If I buy the service and then sublet, I’m responsible for any illegal activities the user might engage in.
b) If I create a new account for the user, why should the user go through me in the first place, if they don't get better price this way and they can actually reach out to the provider directly?
Still, from a legal standpoint the second approach seems more sensible. That said, there’s an obvious question: why would anyone use such a platform in the first place?
From my perspective, Decent Cloud value lies in:
a) Transparency of reputation, ratings, reviews, etc.
b) Streamlined payments, accepting both crypto and fiat/cards (although fiat isn’t supported yet).
c) Ease of becoming both a provider and a user.
d) A unified API for renting resources from multiple providers.
e) As an open community project, Decent Cloud won't disappear tomorrow.
Regarding this last point, I’d like to add Decent Cloud to Terraform and Pulumi so one could programmatically (with infra-as-code) allocate and pay for new resources without creating a new account or dealing with extra hoops for each provider. That flexibility and simplicity could be impactful. In every company I’ve worked at, getting approvals and payment details once is tough enough—needing to do it for five different providers is painful.
Caasify already addresses several of these needs. It’s a pretty cool project indeed. But, I prefer more openness, transparency, and flexibility, with less dependency on any single entity—Caasify or Decent Cloud included. I believe an open community project like Decent Cloud might have a strong advantage on those fronts. For my company it would be quite important that I don't introduce such a single point of failure.
I would like to keep the "infra brokers" concept in Decent Cloud. It seems to work, why ask for something else...
After reading through the comments here, I do agree that the dependency on the crypto payments is a friction point. I'll look through alternatives.
And no, crypto stuff is not the reason the posts got banned on reddit. In the meantime I learned a bit more how reddit works, and it seems like you need to have an account with good karma to post. And me beeing a total noob there had a karma of zero
Actually the idea was to have it the other way around. Your backend would have a public+private key (generated locally), and would periodically check the blockchain for any renting request for the given public key -- i.e. your provider. If there are requests for renting, then you can accept/reject the request based either on the reputation of the user, or anything else, and respond with some explanation or instructions. There is already CLI tooling for this, so it would only require a bit of scripting to integrate into pretty much any backend. I don't have any integrations done yet, since I have no idea what do people actually use for their backends -- please let me know, and I can do it! Of course, it would be even better if someone from the community would do it but we have to start somewhere
I was hoping through their reseller/partner program. But I would need to find someone first who already has such an agreement with AWS and is technically savvy to avoid some silly mistakes and work with me as we polish Decent Cloud.
Fair question. The best answer I can think of is here: https://lowendtalk.com/discussion/comment/4270174/#Comment_4270174
Frankly, do you think such answer would be enough?
The project has a new website. Check it out!
https://decent-cloud.org/
We're also working on many other things including the dashboard, and help is needed in many areas, listed at https://github.com/orgs/decent-stuff/discussions