Pay the Developer (PtD) Reward Amount discussion

Let’s discuss the total allocation of SN Token across the network, how much apps should be rewarded, how rewards curve down once extremely popular, and how accessible PtD should be (for example, strictly for ‘applications’ or also for publishing, eg blogs, curated lists, etc).

@neo @frabrunelle you are probably the most informed on the topic of SN Token allocation, could either of you provide a graphic or breakdown to give participants of this thread a foundation to build from?

5 Likes

To discover what value in dollars should be rewarded in SN Token to an app, should we be creating an app for the network that is a comp (comparable) to a real world application and that applications annual revenue?

Granted there is no time element in Safe Network so how often PtD is rewarded is also a factor.

Something else to consider also is there is no cost to security or scaling on Safe Network to the developer so that cost should be subtracted from the reward amounts.

What other considerations should be included in the amount an app should or should not be rewarded?

1 Like

Very quick thoughts:

  • hard to peg to old web/desktop revenue, also dubious approach?
  • amount available can be related to network (size, revenue etc)
  • amount to app can be related to network contribution metrics (eg access/storage)
  • share per app can favour early apps to drive developer adoption, reward risk
3 Likes

No. There should be no need to link to explicit data in the real. Network conditions alone are all that are needed ( they convey real world data implicitly).

It can be rewarded for each block/chunk of data in the same way the farming rate is applied. No need for “time”.

No, that would be redundant. The network can just set a reward rate, end of story.

KISS. A fixed rate relative to the farming rate is a good place to start. PtP and PtD require Pay on GET though unless there is some other clever accounting method applied.

I’m not so sure that their needs to be a distinction between PtP vs PtD. It may be beneficial for the network to be content agnostic. A chunk is a chunk. Thus a type agnostic reward for data provision should therefore be the same whether the chunk relates to an app or content.

3 Likes

This is where my train of thought has been heading too and I like the way you put it. Makes a lot of sense, to me at least.

My intention in pondering this is to try to pin down what is profitable or how that would compare to a comparable app or industry in the real world. We live in the real world and we will compete. We will also have certain cost advantages so I thought it fair to consider.

Again the point is, if an app on SN is to be competitive it needs to receive fair or competitive compensation. But how is that number or floating value decided or denominated? I like the idea of saying we don’t need to value in fiat but that doesn’t seem to be going away anytime soon. I’d rather value in SN Token cost for SN storage but that has the same question as here, does it not?

This might be useful in the beginning to help get an app ecosystem going but it should die not long there after. Network shouldn’t be a welfare for programmers effort. It amounts to the imposition of a private tax. Instead of a commons belong to end users who pay to keep the network going it is right away reversing that like a government that taxes people to give to private businesses like the public private partnership form of privatization. But it could be good if it would support ad free media. Still like the idea better of letting end users without pressure submit after the fact exactly what they think a services is worth.

But if you were to take it seriously shouldn’t the amount of tokens the producer gets be exactly proportional to the generic value of a second of human attention at the global rate? SAFE can’t afford so it would have to be a set fraction of a second of human attention- and that isn’t bad anyway because these systems aren’t matching human attention for human attention only code to human attention. .

2 Likes

“Reward” follows, rather than precedes.

I like that idea of assisting in the promotion of apps after they are put forward.

I don’t know what measure there should be on what has effect but would that matter. Also unclear if any reward would go directly to a developer themselves, before delivering a clear sense of the small next step, as money is a corrupting motivation and paying ahead of development might see too many unfulfilled promises… just like so much of crypto. The market is weak enough, I think we need to pay and hold to our strengths.

Reward to devs can also come afterwards as users say a thank you for the service they prefer. Early BTC days, people not shy of throwing their thankyou.

The measure of it is a balance of what is affordable to see real world advertising on a fair scale… and anything else that is a force multiplier… so, costs for devs doing presentations of their work, etc. Meaty not fluffy spend.

The other thought is that good apps are not hard to develop, especially if they are on the back of an existing type that is known to be successful. There is good potential for fast app development on the back of Tauri, using Rust and React.

1 Like

I disagree with your first point, and also disagree about the welfare comparison. The network is an entity with a set of internal economics. The goal/object of the network is to survive and grow in perpetuity. There is nothing wrong with the concept of an autonomous network that provides rewards to those who help bootstrap and accelerate its growth indefinitely.

It’s not welfare to offer compensation to developers in perpetuity. It’s more of an employer/employee relationship (you’ll probably hate that analogy too, no? Other analogies exist, ex. patron/artisan). The developer is not salaried or given a stipend but rather receives a partial commission based on work already performed. The developer likely had to invest significant time and energy into their work without any support nor guarantee of return. The commission is proportional to the merit of the work and how much use it receives, not how hard it was to produce. This is a huge benefit to the network. For pay on GET, the merit of the work is determined implicitly by the network on a chunk by chunk basis based on how much the chunk is used/accessed. Chunks that are popular and used frequently cause the network to become more popular and to be used more frequently. This increases growth and demand simultaneously, keeping everyone’s data safe.

3 Likes

Just a thought would it be by access to the whole file or on chunk by chunk basis?

Just thinking about someone who develops a light weight app or say writes a novel that takes up very few chunks as opposed to video streaming that uses lots of chunks?

3 Likes

Surely should be about user engagement and utility, rather than just volume of data.

Just thinking, for marketing purposes, you could play into the pyramid motive that cascades the effort… get a 1000 users and get a reward. It needs to be meaningful and worth the investment or reward… the measure of that is complex… unless made simple by the money available being limited and then assessed on merit. People will spawn entire coins and deliver nothing, so a small up front is ok but there’s work assessing pitches.

3 Likes

Very good point but a “chunk is a chunk” from the network’s perspective, no?

Imo what you describe was a reason that people thought the network should give different reward rates for content vs. apps, or rely on tipping and human judgement. How much do you think an app chunk is worth relative to a video chunk? From the networks perspective, data is data and they are worth the same on a per chunk present worth basis. Popular chunks send more rewards to the human over time. Obviously, from the human perspective they may be valued differently based on the required human effort to create, but this causes individuals to try and create things that have merit. Any human subjective values can be handled via tips or direct producer/consumer relationships like patreon, and that is not the point of PtD/PtP imo. The point of PtD/PtP is to give the network a say in what is important (data, growth, resource availability, user interaction), and send its value signals to human counterparties in the economic system to ensure its own survival.

One way around the content vs. app value debate is to reward the app developer for content chunks accessed via apps. For example, if you develop a stock analysis app or word processor like that available in google docs, you get a reward for all content that is read or written using your app.

Example:

  1. @neik writes an image processing app that detects whether an image shows either a) “not a hotdog”, or b) “a hotdog”. (sometimes real life mimics television)

  2. The network instantaneous farming rate is determined based on available resources and total PtD/PtP rewards are set at a perpetual/constant fixed ratio of the farming rate. Let’s say 5%.

  3. For each chunk of image data accessed using the app, @neik gets a 2.5% reward as app developer, and the individual who provided the chunks for the image that was analyzed also gets a 2.5% reward. For each chunk of new data written using the app, @neik gets the full 5% reward.

These mechanisms rely on a hybrid “Pay on Put” (PoP) and “Pay on Get” (PoG) economics, ie. “Pay on Put or Get” (PoPoG).

2 Likes

The application itself is stored as chunks on the network. So I thought GETs of the applications chunks by an address would count as one reward event. Also as an application gets popular, as with any popular data, the chunks will be cached and therefore not count as a GET request or be rewarded.

Not chunks you can access through the application. Maybe this changes if the application stored the chunks itself on behalf of the user? Maybe that shouldn’t be. I think applications should let the user pay for storing their data and have full control of it, not the other way around.

Another address GETs the applications chunks, another reward, etc etc until (at some point) the reward curves down.

Does anyone remember what that curve was or at what point it kicked in? This info was from sooo long ago. @dirvine? @neo?

IIRC it was a sigmoid curve aimed at a slow start, rapid increase then slow smoothing off to a defined limit.

4 Likes

Nice, thank you! I like the sounds of that :smiley:

1 Like

That’s one option, but I don’t think its as good as rewarding based on content accessed/processed/PUT/GET via the app.

1 Like

I wouldn’t be opposed if the chunks reflected the use of an app but not if the user has to store the data through the app and the app then controls the data.

1 Like

$1 per 1000 gets isn’t too far from what could be expected from the clearweb.

How would this be paid for though? From PUTs just like the farming reward? If you PUT a popular chunk you’ll make money,if you PUT an average chunk you’ll break even and if you PUT a chunk that nobody GETs (e.g. a backup of your private files) you’ll end up paying (and subsidizing those more popular chunks).

This would incentivize making data public, so maybe that’s a good thing. If you to keep your data private you pay a premium.

The issue I see though is how do you prevent someone from GETing their own files over and over to reap a reward?

3 Likes

As far as what, PtD? I don’t think private data would even be rewarded.

Obviously farming is rewarded either way and the private chunks are split up and randomly distributed so nothing to game there for the one who uploaded the data.

How many people do you think would sell goods and services on the original Internet if the prices and processes were dictated by the engineers who invented the Internet?

2 Likes

I was not suggesting that.

That is not what we are describing. The network rewards serve to promote onboarding of users, apps, content and computational resources for the primary benefit of the network. Any free market of products and services you can dream up can then ride on top of that, just like e-commerce does on the current web.

No private data rewards for content, but an app that is used to generate private data could earn PUT rewards using the method I described above. Only public immutable content earns rewards when the GET is made on their chunks if PoG is implemened.

1 Like