Budget Category: Community Service Provider Subsidy

Proposal Kickoff: November 1st, 2020
Budget Amount: 10,000,000 SC per month, from block subsidy
Treasury Kickoff: 10,000,000 SC, from treasury
Timeline: Active until Jan. 2022
Leftovers: To be returned to the foundation treasury

I propose that the foundation set aside up to 10,000,000 SC (10 MS) per month to be distributed to operators of community services. The foundation treasury will contribute 10,000,000 SC to kick-start things and take us until the foundation subsidy begins in January. This budget would be locked in until Jan 2022, meaning that a total of 120,000,000 SC will be set aside over 12 months by this proposal. Any funds that are not used in a given month to subsidize our service providers will be returned to the foundation treasury.

I would highlight the following as qualified community services:

The goal of the subsidy would be to offset the costs of running such services. A requirement of the subsidy would be that all code related to the service is fully open source. Closed source services or infrastructures would not qualify. The subsidy is not intended to be a bounty or a payment system - nobody should be getting wealthy from being a service provider - but instead to ensure that people are not losing money by helping the community.

All subsidies would be paid out in siacoins, and it would be up to the discretion of the service provider to either convert the subsidy to cash or to keep the subsidy as siacoin.

I propose a limit of $1,000/mo in subsidies for any given service. I also propose that services making more than $5,000/mo in revenue (whether through donations or other means) do not qualify for the foundation subsidy.

The foundation would have full discretion over who receives the subsidy and how much subsidy is received.

1 Like

I would add siacentral to that list as well as Nate has some server expenses each month. Regarding Skynet Portals it would be fair to fund only community portals and not Skynet Labs portals as this would be paying infrastructure costs of a for-profit company. Especially since there is a development fee on building Skynet applications that goes directly to Skynet Labs. This would be the same as StoreWise asking the foundation to pay for our server infrastructure which i don’t think is reasonable as a for-profit company.

2 Likes

Would this be limited to services that serve as infrastructure for the Sia userbase, or does this include use cases and products built on Sia as well?

Thanks for formalizing this proposal, @Taek. I expect this will be relatively uncontroversial.

I’m not sure how I feel about this. On one hand, it does seem reasonable that Skynet would cover the cost of running their own portals – it’s their product, after all. On the other hand, why should it matter who runs the portal, as long as it’s public?

This is the distinction that needs to be made, I think. However, I’m having a hard time formalizing that distinction. Requiring the software to be open-source seems like a good first approximation; this is why e.g. StoreWise’s server costs wouldn’t qualify. But things get complicated when the service itself has fees. For example, narwal has a “suggested donation address” – if the Foundation covered the cost of running a narwal server, should donations be subtracted, or not? What if the donations, instead of going to the narwal server operator, all went to me personally?

Perhaps we should require services to disable all fees or other forms of income to be eligible for the subsidy. After all, the purpose of the proposal is to make it break-even to run public services. If all your costs are covered, then any fees or donations are pure profit, which is not what we want to incentivize here.


Another aspect that could use fleshing out is how the services will “bill” the Foundation. Disputes over the accuracy or necessity of budget items could easily turn ugly, so it will be important to have clear terms up-front. Open to suggestions here.

I do think it’s unusual to single out https://siasky.net as the one portal that can’t receive a subsidy. Skynet Labs is unique within the ecosystem as:

  1. everything we do, everything is open source and made accessible so that other people can run their own nodes / versions of our services.

  2. We do all of the heavy lifting for the core protocol development. Skynet Labs has contributed to date more than $6 million dollars in resources directly to the development of the open source Sia core protocols, ecosystem applications, mining resources, and events.

In the particular case of the Skynet portals, all the for-profit elements built into the portals will be open sourced as well. We wish to see the whole ecosystem become sustainable, and we don’t see why other skynet operators should be isolated as able to receive foundation benefits when all skynet portals are on a level playing field.

1 Like

I don’t like this. It is my goal that all portals eventually become financially sustainable absent help from the foundation, but I don’t think that a portal should have to give up foundation help as soon as it wants to achieve independence for itself.

Instead, I think we should just put some reasonable cap on the total price the foundation is willing to contribute to a community project, and put a cap on the total amount of income that project is allowed to have before it no longer qualifies for foundation help. I might say example limits of:

$1,000 per month per service as the cap for foundation contributions
foundation contributions to be withdrawn once the service is drawing more than $5,000 per month in revenue.

I would have the foundation give out a flat grant based on the approximate cost of the service. If we feel that proof is necessary, the foundation can request proof. The dollar amounts are small enough here that I don’t think it’s a big deal.

If the foundation is giving out $200/mo to a service that costs $170/mo to run, that’s not a big deal. If the foundation is giving out $500/mo to a service that costs $560/mo to run, that’s also not a big deal. Especially because the subsidy will be in siacoins, so the exact delta may fluctuate.

I feel the same. The foundation should group the grants.

<1000/mo --> give flat grant based on budget proposal
1000…5000/mo --> Monthly proof of spent budget to be submitted
Above 5000/mo --> setup a detailed contract. Check regularly the spent budget. Keep the possibility to readjust the monthly grant.

Instead of $5000, I think services making above the maximum should no longer qualify for the grant. I’d like to see this grant be as needs based as possible without being too restrictive. The goal being to get lots of developers building and actively maintaining projects built on Sia.

I agree with David, a flat grant would be easiest, justifying and billing line items every month will be a mess. In addition, working towards being profitable by adding revenue streams should not disqualify.

Justification for costs is going to be important, but any guidelines can be exploited. A lot of the particulars are probably going to differ per-service and need to be up to the judgement of the Foundation. I think a defined timeline at which point the grant needs to be reevaluated and approved would be a good addition. Quarterly maybe?

This language technically disqualifies Sia Central and SiaStats. My explorer, troubleshoot, and benchmark are closed source, but Host Manager, Lite Wallet, and things like Sia-WASM are under open source licenses. IIRC, SiaStats host monitor is closed source.

I like requiring open source, fully closed source projects should definitely not be able to benefit from an open community grant. However, the amount of open source required should be relaxed from “all open source or nothing.”

I have no intention of applying for a grant for Sia Central at this time, I’d like to see grants go to projects that wouldn’t be built without it

2 Likes

I’m not opposed to siasky qualifying for a grant, it’s a valuable community service just like the other portals that will qualify.

I think the hangup on siasky stems from Skynet Inc’s alternative funding. As a company you’ve secured 3M in funding in addition to ~10M prior. I believe, going forward Sia development by Skynet is going to be at least partially subsidized by the Foundation, which I also support. While Skynet Inc itself isn’t profitable, it has proven to be capable of sustaining itself.

So the question becomes should a company with alternative funding qualify for a relatively small community grant because it operates an unprofitable community service? It’s a unique situation from the other services in the list.

$10m includes the new $3m (just to clarify).

Perhaps we should require services to disable all fees or other forms of income to be eligible for the subsidy. After all, the purpose of the proposal is to make it break-even to run public services. If all your costs are covered, then any fees or donations are pure profit, which is not what we want to incentivize here.

But then once paid portals come out, will that disqualify them from a subsidy even if they are losing money? I get not funding somebody who is already making a profit, but I definitely think it shouldn’t be all or nothing. Same with open source.

1 Like

The idea behind the all-or-nothing element of open source is that we want the community to have a strong chance of being able to pick up the service if the original provider moves on. For the vast majority of these services, being open source while you are still dependent on the foundation subsidy (which is capped to $1,000/mo) is not going to hamper any long term business potential.

Especially for software, I don’t think that requesting fully open source is a big ask. The reality is that the original developer tends to retain all of the market share even if there are clones. And, once there is significant revenue, there’s no harm to close-sourcing new features.

That makes sense to me, but how do paid portals fit into this? Will subsidized portals not be allowed to have paid options? So then will the foundation be paying for people to view expensive content? I’m a bit confused on this.

This could ofc be worked around by not allowing content over xSC be accessed on subsidized portals, but I feel like that is kinda a band-aid solution to something that doesn’t have to be an problem.

Summarizing some of the feedback here: do people generally agree that the Foundation should:

  • award grants in fixed, bracketed amounts, rather than trying to make each service exactly break-even;
  • set a fairly low maximum grant amount ($1k-$5k/mo?);
  • and allow collection of fees/donations?

Another consideration is whether we should “price” the grants in terms of SC or USD. The former risks overpaying or underpaying recipients, but the latter risks exceeding the proposed budget of 10MS (if e.g. the exchange rate falls through the floor).

btw, a potential compromise re: open-source would be that the service must provide their full source code to the Foundation. That way, the Foundation could hand it off to a new maintainer if the old maintainer leaves. We would still push for 100%-open-source wherever possible, but I think there are some cases that aren’t unreasonable, e.g. I know the SiaStats host benchmark code is closed-source to make it harder for hosts to cheat.

As long as a not-insignificant portion of the project is open source I don’t think the Foundation should discriminate projects that are open source in principal but have closed source features.

Personally, I think service loss is an issue that will rarely occur and be relatively simple to work around while alternatives fill the void. I’m not sure the Foundation should worry about securely storing third-party source code.

1 Like

I like the idea of having a open source mandate/bias, but agreed on points that specific in some cases, closed source projects make sense. And definitely agree with Nate that front-loading the code-janitor work doesn’t make sense near-time.

We should follow a model like the CNCF (Cloud Native Computing Foundation) for how they take-over + incubate + graduate projects.

From my limited understanding of how it works, there are three categories: sandbox, incubating, and graduated. It gives quite a bit of flexibility in how code bases can be transferred into the Foundation, as well as how it is maintained, managed, and so forth. Developers can set expectations when working with these codebases as well, as they will be given an implicit-grade on production-readiness based on the project status.

Overall, I’m in favor of the broader idea of providing community service provider subsidies, and hope that it will bring in alot more interesting innovation around the Sia ecosystem, now that server costs are covered :wink:

This will really depend on what the funds are being used on.

If I am requesting funding because I need to pay an engineer for something, I am going to want USD-denominated funds. But if I am trying to buy lots of storage on the Sia network, then I will want SiaCoin.

1 Like

(post withdrawn by author, will be automatically deleted in 24 hours unless flagged)

$5,000 a month in revenue - I wish! I think SiaSetup has collected something like 50,000 SC in donations over it’s lifetime of the last three years.

Will the community subsidy apply to development hours and similar soft costs? If so, all of the community sites you listed could pretty easily cap out the $1k monthly limit.

If one of the listed services also runs a Skynet portal, do those count as two separate services? Granted, you could run a hell of a portal setup with a $1k/month budget, and I don’t see a situation currently where someone couldn’t cover their service costs and a portal with the $1k subsidy.

Everything on SiaSetup is licensed under GNU free documentation, and the majority of calculators and the like are client-side in Javascript - most of the site is static HTML. Is that sufficient for being open-source, or do I need to clean things up and throw it onto GitHub as well to qualify? I don’t mind doing so, I’ve just been lazy and it’s more work. I’ve also got some proprietary dev-to-prod upgrade/management code I’d probably want to pull out and separate first as well.

Somewhat of a side question but related to revenue calculation: Will the premium Skynet payment model which is planned for the future be portal-dependent and result in income for individual portal operators, or will that ultimately go to Skynet Labs through licensing for the use of any portal? Even if portal-dependent, I don’t expect that many people are going to be lining up to sign up to SiaSetup or SiaStats or any other portal specifically when they don’t offer anything special compared to SiaSky.

1 Like