APP Incentive Pay Model


#1

Continuing the discussion from We’re in the Mastercoin Roadmap:

From what I understood,

Storage is charged by the Network regardless of private or public status. The last I recall, it cost Safecoin to PUT data onto the Network, above the free amount (this amount has not been determined yet.)

Hypothetical Example…
Assume the free amount was 1Gb.
If I upload a 10Gb blockchain file, that has not been uploaded yet, it will cost me Safecoin for the remaining 9Gb of data. When another user tries to upload the same file, the Network will recognize it and just meta tag the file instead of charging that user Safecoin for 9Gb.

The APP incentive pays Safecoin based GET request popularity. So this may be where you’re getting it crossed. The APP usage is not exactly the same as a de-duplicated file. A person does not need an APP to reference/access a de-duplicated file.

However, I can see where one would to create an APP that “middleman’s” the API gateway to access the de-duplicated file. For example, a wallet APP that references the blockchain ledger. That wallet APP would get paid Safecoin based on the APP incentive model. Another example, would be a video APP that points to videos already uploaded to the Network. So instead of hosting the storage for those videos, the video APP collects Safecoin for referring other users to those videos.

Side note…
The APP owner, need not even upload the original blockchain file. They could just wait for one to be uploaded by someone else and then upload their wallet APP afterwards.

This does raise some questions about how the APP Incentive Model will be used. @frabrunelle and myself want to do a MaidSafe Mumble chat to talk about this in more detail.

I won’t get into my personal view as to what is a fair pay model. But I am concerned how this will burden the farmers on the Network. They have to service an overwhelming demand of GET requests if APPs are created to take advantage of de-duplication. If the farmers are not being compensated enough, I fear it would hinder Network growth.


APP Developer Rewards (Discussion)
Streaming large video files with a distributed network
#2

Thanks David, and for creating the topic.

This sounds probable, perhaps rewarding of public content providers for accesses of their content was not brought into the core, though I thought it was being seriously considered.

Content can be rewarded through a type of paywall app of course, but that’s a different scenario.

So my scenarios reduce to the much less severe issue which you outlined. This should I think be covered by the 1% fee on non Safecoin fees, at least in theory… but it begs the question of who and how to charge for people using SAFE to host revenue generating data like a blockchain, where all the fees are handled outside SAFE, on an essentially anonymous decentralised network. I guess we’ll have to pass on that one for now, but it creates a model for anyone who wants a free ride on SAFE, that we may have to tackle at some point.

Maybe this needs flagging as a potential “attack” @dirvine

I guess we could handle it piecemeal, by identifying such freeloaders and offering them the chance to pay a fee or suffer some penalty, but given the philosophy and implementation of Project SAFE (no controlling authority), I’m not sure if there would be a way to impose a penalty if there was no legal avenue (ie no overall identifiable infringing business). So perhaps piecemeal is not viable, and the solution would need to be in code.


#3

My personal opinion,

We cannot verify if the APP owner is also the content owner. Because of that disconnect, it would be unwise to pay based on usage, GET requests. Instead, an APP owner should be paid for the APP functionality that it offers. This is not so easy to measure because it is so subjective.

GET requests should only be tied to the farmers as they are the ones directly burdened with handling this service. So far, this is how the core works. We just added an APP incentive which overlaps it.

One solution could be to distinguish between a meta tag reference and the original uploader. We already know who paid for the storage to upload the file. So maybe that would be a case to protect, even pay creators for their content. We definitely should discuss further.


#4

@frabrunelle and I are talking on MaidSafe Mumble right now. So if you want to join or listen in. That would be cool.


#5

Why do we assume that the second uploader would not also have to exchange safecoin for the 9Gb? Say the first loader then deletes the 10Gb file to make room to store something else within their purchased cap. Then the second storer has a free ride as the sole holder of that data.

I think the safecoin limit should be related to what each individual uploader maintains, regardless of deduplication. I don’t think that it’s planned to be otherwise, but if it is we should look at it closer. Each account that uploads (and thus has access to) a quantity of stored data should be responsible for paying for that data. This eliminates the “freeloader” senario. Deduplication is a network boon, not a right of all users.


#6

I would really appreciate a summary of what the two of you discuss. @frabrunelle and I spent some time talking about it in a hangout last week. Good stuff and different perspectives. The App Incentive Pay Model is a HUGELY important aspect of the SAFE Network that we need to understand and possibly debate the details a lot. It’s a Golden Goose that could be the greatest thing since sliced bread, but could also become the very thing that brings the “internet” aspect of MaidSafe crashing down if people find ways to game the system. And if it can happen it probably will happen. I’d like to see a detailed white paper on this aspect in the near future (perhaps there is one already).


#7

I missed this one, sorry. It may be an idea with such discussions to record them. With mumble its easy you click the red button. Everyone can see you are doing it so its not hidden at all. Then post the link to the chat if it was worthwhile?


#8

Doh, I’m recording our Mumble Chat now.

The Mumble .wav file ended up being pretty large 222Mb so it does not automatically play from the link. I’m trying to make it smaller by converting it to an Mp3 format and see if that helps. Here’s the link to the original wav file.

EDIT: I converted the file to Mp3 and replaced the original link above. It’s now much smaller 22Mb. The conversation is 29 minutes long. We both have accents but I swear we are speaking english.


#9

Thanks for th Mumble invite - only just seen it unfortunately - today is my ĺ don’t touch each tech day" hah hah hah!


#10

The idea is sound, but as mentioned in previous posts, it is very easy to cheat the system. For video, audio – you simply compress it at a slightly different setting… for images, you can crop it 1 pixel shorter… etc… There are many ways where you can alter a file where it appears to be identical, yet recognized as completely different. The network will be flooded with variations of original content in an attempt to cheat the system.

Being in the creative/marketing field, I’d like to think presentation is what really defines success. So going back to video as our original content viewed on the network… Do we reward a user safecoin on an original video if he were to tag it “some stupid video”… Or do we reward the user who tags the video “Boy Chops Tree Down with Bare Hands” and ultimately drives 95% of the viewings.

The idea of reduplication should occur as a benefit to the system itself… not to the user. So if 2 users so happens to upload the same file – both users will be charged the same amount (yet the network only stores one).

Also, just a thought… an idea might be how to find an innovative way to copyright content… like the DMR embeded in itune songs.

Note: Disregard if I have a misunderstanding of the Incentive Pay Model plan.


#11

If this is true, then I’ve been misleading how storage is charged and no one has corrected me until now.


#12

I hadn’t understood before that you thought this, though thinking back on some earlier exchanges, I can see it now.

I don’t see how it can operate without storage caps being calculate without considering deduplication. If the original storter deletes their copy, what later storer then gets charged with the amount towards their cap? You could never really know where you were at or predict how storing something would affect your own cap, or what effect your deleting something from your account would have on someone else’s cap.

Also it seems that deduplication is a natural network function, but then back-referencing all accounts affected would add tremendous unnecessary traffic to the network. When I PUT or DELETE data, I and my close nodes know it’s put or deleted. My percent of storage cap can then be easily kept track of “locally” (among my concensus nodes). If you try to factor in deduplication, the traffic necessary to calculate and confirm goes way up, and adds other complications.

I could have it wrong, but I sure don’t think so. Recommend any big guns to call in to school the both of us?


#13

My opinion on this is that it has to be in symbiosis. Developers must get paid in my opinion partially in response to the usage of their app but not exclusively by this.

But users also in my opinion should be given incentive to use the apps. I don’t know that Safecoin has to be used for this purpose though.

I hope we can find a symbiosis.


#14

I’ve been thinking about it since my last reply and I believe you are correct. It would cause a complication, trying to determine who is charged for the storage after the original uploader deletes their reference.

For the example below, I refer to bloat as hard drive space required by the client and how the SAFE Network is supposed alleviate it.

Non-Shared Private Files
The topic (Blockchain Bloat) stated each person will upload their own ledger. I understand de-duplication reduces the storage space used by the Network. What about the cost to the consumer? If blockchain users have to pay for their own copy of the ledger, would it not be cheaper to store it on their hard drive instead?

These issues keep leading me back to a consumption model as a solution. I’ve talked about it on this forum. But I don’t want to step on any toes, so I’ll just leave this alone.


#15

You know, there’s really nothing wrong with people storing stuff on their own hard drives. I’m sure I’ll continue to do it in great quantities. I don’t think the object of the SAFE Network is to do away with all local storage.

Will it be cheaper to store your own copy of a blockchain on the your own hard drive or on the SAFE network? I don’t know, and not sure it makes a big difference. Also I’m really not looking for SAFE Network storage to be all that expensive. Deduplication will tend to drive the cost of SAFE storage down in general, so even storing a lot of unique data will probably tend to be fairly cheap.


#16

Thanks David. quick question.

What if a person created a bot that initiates GET requests to fake ‘popularity’ from numerous nodes, can the SAFE network detect this…


#17

We should be able to detect many such patterns, the better answer will be the usual MaidSafe way, improve the core algorithm so this is impossible. Otherwise we get into the virus / anti virus way of working. We always just try to make teh problem impossible.


#18

I remember this discussion and suggested that not only do we need to charge every uploader, but that we need to continue to charge for storage, not just an initial upload fee. I suggested some variations on how that might work, but there seems to be a gap on many issues between community discussion and what the team are actually working to. I saw this as @dirvine having to give the community space to develop and thrash out idea, and not wanting to jump in and say “we spent x time discussing this, and explored all those ideas, and this is what we came up with and here’s why”. That would have been my tendency in his position, but he’s much better at this than me. I don’t know how many areas we discuss that this applies to, if any. I’m sure David reads just about everything and takes it into consideration, but he’s spent years thinking through so much of this, and seems fairly good at it ;-).

So, where are we? What is being implemented and why. I think the community is strong enough to see how it is, and let’s face it. 99% of what is being coded can’t be tinkered with because we are weeks from launch. Of course there may be other reasons for not laying out all the specifics. Maybe it would just take too much time from the development and we’re still waiting for documentation?

I trust David and the team, but if there are areas where the community’s input could be usefully focussed, we’re all eager to help where we can.


#19

Copyright requires surveylance and leads to censorship. I for one would not be in favor of copyright on the SAFE network. I want my anonymity.


#20

In the real world, governments are getting very serious about copyright via a bunch of secret ‘free trade’ negotiations, which include making corporations immune from states. They also want to end anonymity for the people whist staying secret themselves.

If governments are not willing to put themselves on a public blockchain, then the people will need to do it from within by utilizing SAFE…this network is infinitely more important than the uninformed masses realise.

It just needs to launch for the ability to whistleblow, the rest will get sorted along the way I’m sure…I have confidence in the inbuilt voting system that allows for consensus on core network changes via the Pods.

So called investors, can sleep well in the knowledge they are investing in humanity and our way of life.