Idea, still not sure.... it may stink badly

Ideas & Suggestions
1

so the idea has beens doing circles and donuts in my head for months now

but in simple terms a community satellite with a twist,

where the twist is the storage nodes are paid out via paypal in either Euro € or USD $ depending on where the storage node is located

i would have put this in the project and stories category but of course that one is read only so that was a no go… are there a better fit?

not sure i would change much on the customer side other than maybe set up pricing packages

for both sides no traffic metering to keep things simple

how bad does the idea stink ?

2

Paying SNOs has never been a problem. There are dozens of ways of doing it. And we’ve had the ability to run community satellites for years.

The problem is attracting customers. With so many other S3 providers out there… who would pay to use this new community version?

If you built all the other pieces of a community-run Storj competitor… at the very end of the process… finding SNOs to hold the data would be the easiest part: they’ll work for pennies, no matter how they get paid :money_mouth_face:

1 Like
3

yep, but that is where my version of good moral says that they should be paid fair and in a way they can go buy bread without to much loss

but yeah, the customer would not care how the data is stored, just that its safe etc

so yeah, why should a customer pick me over storj or any other competitor for that matter… will make a hard think on that at least.

4

Where do you get the code for satellite? Don’t think it open source (even if it is, don’t think a community version can afford google spanner anw).

If you can inferred back the source code of satellite from storj client, what if storj client slightly change the api, would you able to keep up or this will be a hardfork?

5

So, in nutshell, you want to compete directly with storj, and your competitive advantage is… slightly different payout method to operators, who are not even the most important entity in the whole process? What will you put into your marketing materials? Here, service like storj but more expensive and less trustworthy but we pay operators in cash!

Why not focus on that then? Create a service to automatically convert tokens to dollars and send to people’s bank accounts. Oh wait…

2 Likes
6

It has been available for years now. But yeah nobody really wants to donate the resources needed to run it in their homelab (or pay to run it in a Cloud provider). There’s not really a point: lots of work for little gain…

2 Likes
7

I see.. it was right on storj/storj. Honestly, I can see this might work, I’ve only use about 1/3 of my storage, the rest is wasted (about 80TB).

You are right, for me it can go down further, storage not in use is a waste anyway, so create a satellite with $1/TB or less and I still even consider.

8

well i need to create an admin interface to control all the settings, that does not come default and will take a good amount of time, also i need to learn about the settings as there are very little documentaion on them

but hey, i’m on a pension so time is not an issue

9

So, about two evenings worth of prompting LLM, I’d estimate?

10

i’m a bit of tired so sorry for asking stupid, but what is LLM ?

11

Large Language Models, it is the Artificial Intelligence stuff, you ask and it answer - eg: ChatGPT, Claude, you could give it the link to storj github and ask it how to set thing up.

12

yes but AI still needs to find the information somewhere, or else it would just be guessing

but you are right, it could be used as a kind of “wingman” for some of the work

13

I meant this: https://cloud.google.com/discover/what-is-agentic-coding

Yes, the key is to prompt it correctly, to go find information, brainstorm the idea, bounce ideas off of another agent, etc.

Human is no longer needed in the loop for small-to-mediup projects. Welcome to the future.

Yes. it started like that. Now it’s more of a contractor, third party software vendor. In teh future it will be even more self-encompassing.

I like this quote form Alexandrescu’s recent talk:

image
image1920×1087 141 KB

14

While you are at it, you may try multiple servers with different pricing range, eg: $1.5, $1.2, $1, 80¢, etc.. For example: I might set 80TB for $1.5, but after sometime without customer, I’ll split 30TB to $1.2 server, of course if it too low, I won’t border. Let see where demand meet supply. May also get rid of /24.

15

getting rid of the /24 requirements is also in my plan.. its simply to strict, but i plan to add an option so that nodes inside the same ip(address) dont get the same data

ie the same idea as /24 but not as strict