[DAO: QmTkzM2] Curation of Programmatic Collections for Linked Wearables

by 0x87956abc4078a0cc3b89b419928b857b8af826ed (Nacho)

Summary

Define the curation process of Programmatic Collections for Linked Wearables a.k.a Third Party Wearables.

Abstract

This is a follow up proposal to cover one particular use case around Linked Wearables collections that are created in a programmatic way. This means that each item was not crafted individually by hand, but automatically generated with code, many times from traits that were previously designed and modeled. CryptoPunks, CryptoAvatars and BAYC are examples of Programmatic Collections.

Motivation

It is very unpractical, and almost impossible, to curate all the wearables in a large collection of 10k items, or more. It is also prohibitively expensive for third parties. But if that collection was made programmatically, it may not be necessary to curate every item one by one. It would make much more sense to audit the process used to generate that collection and curate a random portion of the items. Like when you are boiling spaghetti, you don’t need to try them all to know when it’s ready :wink:

Specification

This is a proposal to introduce a new and more efficient way to curate this particular case of collections for Linked Wearables. It will make it easier and faster for the curator, shorten the time to put the collection in production, and save costs for the third parties.

The new mechanic consists of 2 parts:

  1. Reviewing the method: the curator will review the programmatic method used to create the items in the collection. If the method proves to be generating items programatically with the same base traits, it will follow up with the second step.
  2. Reviewing items in the collection: as the items were not created individually - but automatically with a previously validated programmatic method - there is no need to review each item in the collection. Reviewing just a random portion of the items will be enough to validate the rest. The curator will randomly choose X% items in the collection and curate only those. X% to be defined in this proposal.

Voting

Vote to define “X”, meaning the portion of the collection to be curated to consider it approved as a whole.

  1. YES: X = 1%
  2. YES: X = 0.5%
  3. YES: X = 0.25%
  4. YES: X = 0.1%
  5. NO: Decline this curation method for programmatic collections in Linked Wearables

Examples

  • “x” = 1%: If 1% of the items in the collection have to be curated and the collection has 10k items, then 100 items need to be curated. Thus the third party needs to pay for 100 wearables*
  • “x” = 0.5%: If 0.5% of the items have to be curated and the collection has 10k items, then 50 items need to be curated. Thus the third party needs to pay for 50 wearables*
  • “x” = 0.25%: If 0.25% of the items have to be curated and the collection has 3k items, then 8 items need to be curated. Thus the third party needs to pay for 8 wearables*
  • “x” = 0.1%: If 0.1% of the items have to be curated and the collection has 10k items, then 10 items need to be curated. Thus the third party needs to pay for 10 wearables*
  • NO: If the collection has 10k items and this curation method is not implemented, then 10k items need to be curated. Thus the third party needs to pay for 10k wearables*

*(USD 500 per wearable as defined in the past proposal).

Note: consider that, if a Third Party has to pay for all the items in a large collection, the price could be extremely restrictive for any Third Party. For example: at current curation prices, if the collection has 10k items, the price for the curation would be 5 million US dollars!

  • YES: 1%
  • YES: 0.5%
  • YES: 0.25%
  • YES: 0.1%
  • NO
  • Invalid question/options

Vote on this proposal on the Decentraland DAO

View this proposal on Snapshot

For reference, on a 10000 items collection:

  • 0.1% is 1000 items to check
  • 0.5% is 500 items to check
  • 0.25% is 250 items to check
  • 0.1% is 100 items to check

I’m voting 0.1% (100 items) , more would be too time consuming

I miscalculated the values. Moving my vote to 1% (100 items, over 10k, to check and pay).

It’s not what was voted for in the previous proposal, the third party needs to pay for all their slots, not just the ones curated.

Option 3 wins - Procedural Collection BAYC collection has 10k NFTs and they want to upload 1,000 cool BAYC sneakers that were created in a procedural way, they will need to buy 1,000 slots.

  • Total cost: USD 500,000
  • Curators will receive USD 2,500 assuming a portion of 50 out of 1,000 wearables were curated and the procedural process is validated by the curator

I’m voting no because all the slots need to be paid for.
Allowing someone to publish 10k different wearables for the price of 10 or 100 is not fair for all the other creators, and it will lead to abuse.
DG could create their own NFT collection, get approved by the DAO as third party and then publish 1000 different wearables for the price of 1 or 10? It doesn’t sound right.

4 Likes

More i read the shit Ebriel says and proposes he seems to want to fuck up the system and gives no shits to actually support dcl or those who invest in it. Kinda funny considering his vote power. ill vote no with you.

totally agree with you
:+1:

Usually, I totally agree with @Eibriel suggestions but this time I partially support he/she.

I support the idea of programmatic generations, being able to create wearables in such a way will increase diversity and simplify the wearable creation process for large projects, such as games. However, I disagree with the payment of the proposal. and it’s even funny that the most generous option is %1. :sweat_smile: It is like Pawn Stars bargain style. start with the bare minimum then if increase a bit, the customer will feel that he won the bargain. lol

There should not be any discount for programmatic collections at all. Only discounts can be for the whole community. that’s it.

2 Likes

I think this is sub-optimal from a technical / cost perspective, and too focused on the burden on the wearables committee. I vote no.

I think we should split the discussion in two/three:

  • Burden on the curator, their reward, and number of items to view
  • Cost for the submitter

Burden on curator, their reward, and number of items to review
It’s harder to verify that the collection is generated programmatically. This might require some coding skills on top of 3D CAD knowledge. The cost of reviewing items, if they are actually programmatically generated, has diminishing costs to scale. The curator will likely catch any major issues on the first ones: any major problems or incongruousness on the meshes will be spotted right away. The second source of errors will likely be on different “traits” (like CryptoKitties’ color, expressions, etc). After checking most of the different traits, the rest of the items reviewed will probably be just fine, potentially just having issues with a particular combination of two traits that might not work well together. Due to this, I expect diminishing efforts to review multiple items. That is to say, reviewing 2 items will be just a little less effort than 2x the effort of reviewing 1 item, but reviewing 100 items or 101 items will roughly be the same effort.

I propose we ask curators to review 1% of the items, or a number proportional to the amount of traits the programmatic collection has. For example, CryptoPunks have 5 base attributes (ape, male, female, zombie, alien, and 87 attributes (beanie, beard, etc). Curators should review at least one of each of these, and then some combinations. Maybe twice or three times as much, for example: (5+87) x 3 = 276 cryptopunks. The first reviews should be the most time-consuming, but after the first hundred procedural generated models, it’s unlikely the curator will find issues. I agree they should receive compensation related to the number of items they review; the higher cost of the first few items reviewed eventually gets compensated by the ease of review of the later items.

Cost for the submitter
The submitter’s cost doesn’t need to be tied to the reward or effort of the curator.

If we analyze the burden on clients and catalysts to support the feature, whether something is programmatic or not makes no difference. It’s still 10,000 new items, a lot of megabytes on disc, on the network, and on the CPU that everybody will have to download and process in order to see the item.

The fee on wearables is not there to “make creator’s life impossible”, “prevent the platform from being successful”, “go against building a sense of community”, on the contrary. It’s our best effort at striking the right balance between liberty of expression versus scalability, server costs and performance burden on other users.

Please see my previous posts on this subject:

3 Likes

Consider this information before you vote:

  • Linked Wearables cannot be sold in the Marketplace (see original proposal). So as opposite to regular wearables, the fees payed to upload them are not being recovered from the sales of the wearables.

  • Forcing third parties to pay for all the items in the collection when collections are too big makes it completely restrictive, as it would cost 5,000,000 USD (at the rates defined in this proposal) to upload a collection of 10k NFTs and 2,500,000 USD for a collection of only 5k NFTs.

I’m sure there may be a better way to choose the portion of the collection to curate. That could be a new proposal in the future to change it for the best.

I’m also sure that if the result for this proposal is “NO”, then it might kill the feature, as most of the communities interested in having Linked Wearables have collections of at least 5k NFTs and will be blocked because the price to enter would be extremely high.

First a disclaimer. I am one of the top 3 holders of wearables so I have quite a bit of skin in the game here. I am also the founder of a P2E game and avatar collection and would love to see people be able to be able to play in DCL as some representation of their avatar using linked wearables. We have already created the collection which simply varies 71 textures on 1 model, but publishing would be prohibitively expensive at current prices.

I am voting in support of this proposal at the .1% option because I think the economics voted on in the previous proposal are completely unworkable when applied to linked wearables. The previous proposal places a prohibitive financial burden on the creator and does not properly reflect the cost of curation. Additionally, these linked wearables are not like other wearables in that they are not independent NFTs that are part of the DCL wearables collection. They are simply in world representations of items tied to ownership of a third party NFT. They cannot be bought or sold individually.

This was simply an example of an instance where all the contingencies were not thought of in the original proposal and now that it is being put into practice and projects are giving feedback, an obvious issue with the original proposal has been reveals and a solution needs to be developed or we risk undermining the original intent of the proposal.

The one caveat is that the curators should ensure that this exception for linked wearables is not used to circumvent the prior proposal as it pertains to normal wearables. For example, a project should not be able to create wearables as third party NFTs specifically for the purpose of selling and using those wearables in Decentraland. But with an ever growing desire to “play” as your ape or punk or legionnaire, not adopting this proposal will put DCL at a disadvantage.

First, I must say: very nice proposal. Great formatting and wording.

However, I’m voting no. I understood something different from the original proposal. It sounds weird that committe members need to evaluate a % of a collection of 10k wearables. I also don’t like that each of the 10k CryptoPunks, BAYC, etc have 10k different wearables. They will flood the marketplace.

Also,

+1 to Esteban.

Can’t this be implemented by a single “dynamic” NFT which changes attributes according to an NFT inserted in a “slot”, instead of having 10k different (but very similar) wearables, that need curation?

For a simple example, imagine the wearable BAYC creates is a white t-shirt. When you “assign” your BAYC NFT to the wearable (UI/UX to be defined, but basically the avatar needs to own both NFTs), the front of the white t-shirt displays the BAYC NFT.

More complex dynamic wearables can use different attributes of the NFT to change colors, shapes, and other specifics. This, to me, is what “linked wearables” sounds like. Not a collection of 10k independent wearables.

Again, great proposal, but I don’t support it.

However, I’m voting no . I understood something different from the original proposal. It sounds weird that committee members need to evaluate a % of a collection of 10k wearables. I also don’t like that each of the 10k CryptoPunks, BAYC, etc have 10k different wearables. They will flood the marketplace.

Just to clarify something that could be misunderstood. Linked wearables are not going to be traded in the Decentraland Marketplace, there is no flood. They are not new NFTs. They are just 3d representations of already created NFTs. The linked wearables proposal may give you more context.

This proposal is to discuss how many items the creators will need to pay (and therefore how many items the curators will need to review) in order to submit their 3d representations if they create those representations by using an algorithm.

In addition to this, asking for 10k items is going to be prohibitive. The idea of the whole feature is to take one step closer and taste one bite of interoperability. For well-known collections, Decentraland will also earn new communities. It is not only about money, it is also about building a strong and higher community.

2 Likes

Curation of Programmatic Collections for Linked Wearables

This proposal is now in status: FINISHED.

Voting Results:

  • Yes: 1% 7% 419,547 VP (9 votes)
  • Yes: 0.5% 1% 1,617 VP (1 votes)
  • Yes: 0.25% 1% 66,795 VP (5 votes)
  • Yes: 0.1% 7% 425,952 VP (9 votes)
  • No 84% 4,848,655 VP (7 votes)
  • Invalid question/options 0% 0 VP (0 votes)

Curation of Programmatic Collections for Linked Wearables

This proposal has been REJECTED by a DAO Committee Member (0xfe91c0c482e09600f2d1dbca10fd705bc6de60bc)