This month in curation #6 | August

What a month!

  • On July 8th, Curation was launched alongside two new products: Graph Explorer & Subgraph Studio! Now anyone can publish subgraphs to the decentralized network in a permissionless manner, allowing for truly decentralized applications.

  • Figment is awarded an 80M GRT grant to join The Graph as a Core Developer Team.



Curation, Graph Explorer & Subgraph Studio Launched on July 8th 2021

  • The graph launched the Decentralized Data Economy with Curation, Graph Explorer & Subgraph Studio. Read the Blogpost.

    • Developers can now deploy and publish their subgraphs to The Decentralized Network in a permissionless manner, allowing for truly decentralized applications

    • With the Graph Explorer, anyone can use existing subgraphs for their data needs.

    • Curators assess the subgraphs that are being deployed, and signal to the network whether the data is valuable. Curation is key to scaling the network as thousands of subgraphs are migrating to the decentralized network. Curation markets represent the cutting edge of Cryptoeconomic Primitives, and has never before been done at this scale.



Figment is joining The Graph as a Core Developer

  • The Graph Foundation and The Graph Council have awarded Figment an 80M GRT grant to join The Graph as a Core Developer Team. The grant will be distributed over 5 years.

  • This news is a major milestone in the shift to Web3 and decentralization. It marks one of the largest shifts of a centralized company into a decentralized protocol, as part of a larger mission to make open data a public good

  • Figment has technical expertise across 35+ proof-of-stake networks. Their broad knowledge and experience and their passion for open, borderless systems, makes Figment a perfect technical- and mission-driven team for The Graph ecosystem. They will work on tasks like:

    • Improving indexing performance
    • Adding support for native network subgraphs
    • Expanding multiblockchain support
    • Contributing to the evolution of the core protocol
  • Read the announcement.

  • Visit figment.io. They are hiring. If you have the right skills, make sure to get in touch with them.



Grants & Requests for Proposals

  • Wave 2 grants have been announced. The Graph Foundation have provided over $2.8 Million in grants and ecosystem support in Wave 2. Grants have been awarded to a wide range of projects , many of which you might interact with in the future. I highly recommend taking a look at the announcement and the detailed descriptions for Wave 2 grants.

  • Curator Rewards Dashboard There is an open RFP for a dapp or UI that can help Curators make well-informed decisions. The tool should help with subgraph discovery. It should also give Curators more information about their positions, past rewards and taxes, and estimations of potential future rewards. More details in the page. If you know anyone that have the necessary skillset to make a great dashboard for Curators, make sure to direct them to the RFP.

  • There are still 16 open RFPs for subgraphs, tooling, community building and more. Make sure to check them out and appy. The Graph Foundation will be in touch with the best candidates.



Community Talk

  • The monthly Town Hall meeting has now transitioned into The Graph’s Core Devs Meeting.

  • The Graph’s Core Devs Meeting will focus on all things protocol and GIPs

  • The Graph Community Talk is a new initiative, that will focus on subgraph development, Curation, Delegation, Foundation updates, grantees and other community initiatives. The Graph Community Talk is geared towards a broad audience regardless of background, whether technical or non-technical. The Community Talk will look for active participation from the community.



Curation

As with the launch of any new product, there are of course feedback and improvements to be done. The community has rallied in several forum posts, please join us collecting feedback and discussing improvements:



Curator Feedback
I encourage everyone to actively engage in protocol discussions.



Graph Improvement Proposals
The feedback from the community has kickstarted the following Graph Improvement Proposals:

  • Batch GNS Transactions - This proposal would allow batching transactions on the GNS. With this feature in place, subgraph developers may publish and signal on a subgraph in the same transaction. I believe this goes a long way to address the Curator challenges outlined below.

  • Subgraph Change of Ownership - This feature would allow subgraph developers to transfer the ownership of individual subgraphs.

  • Proposal to Reduce Curation Tax - A proposal has been made to reduce curation tax. A copy of the full proposal can be found here.



Curator challenges

In all of this, it’s worth highlighting the following challenges and potential solutions that have been proposed:

  • Frontrunning bots - Due to the nature of the bonding curve, being the first to curate on a subgraph carries the least amount of risk, while also having a large upside potential. The first Curator takes the following risks:

    • 2.5% curation tax
    • gas cost
    • opportunity cost

    While the risk of subsequent Curators are:

    • All of the above
    • The GRT Valuation of their curation shares might decrease
  • High volatility - With the bonding curve starting at “0”, there is a great deal of volatility on subgraph launch. This encourages Curator behavior that might be less than ideal:

    • Low quality signal: As noted in the point above: earlier Curators face significantly lower risk. Curators might signal on a subgraph without first assessing whether the subgraph contains valuable, complete and accurate data, and whether a dApp intends to use the subgraph in production.

    • Pump-and-dump: Early Curators on a subgraph stand to profit from convincing fellow Curators to curate on the same subgraph.

    • Migration help: Curators are incentivized to actively help developers migrate from the hosted service to the decentralized network. This is great. However, some developers might feel pressured into migrating without having critical knowledge about The Decentralized Network, such as;

      • Developers should be the first ones to curate on their own subgraphs.
        • Indexers allocate resources according to the signal on a subgraph. Developers that hold a significant self-signal would receive a more predictable service. I believe this will be healthy for the network, as I explained here.
        • Being the first on the bonding curve, allows developers a sizeable rebate on their query fees.
      • Some developers are not aware of how versioning works on the decentralized network, and the cost of auto-migrating curation shares to a new version.
      • Several features are currently only available on the Hosted Service, and have not yet been added to the decentralised network.
  • Developer “Bait and Switch” - This is not a huge issue as of now. However, I think this should be addressed;

    • Subgraph developers that wish to upgrade their subgraph pay a 1.25% migration tax on all auto-migrating curation shares. With the current system, a subgraph developer can, instead of upgrading, choose to “bait and switch” their Curators, and publish an entirely new subgraph. They are economically incentivized to do so:

      • They won’t have to pay the curation tax of migrating to a new version
      • Assuming the subgraph developer is the first to signal on their own subgraph, their shares will have appreciated in value. They would retrieve more GRT than they put into the bonding curve.

      A counter argument would be that repeat offenders would “get a bad reputation”. However;

      • A decentralized network should try to minimize “human trust”
      • Word of mouth reputation systems do not scale well. I expect a large number of subgraphs will be deployed to The Decentralised Network.
      • It would be easy for developers to deploy the same subgraph under a different name.



Solutions
Several solutions to these challenges have been proposed.

  • Batch GNS Transactions This solution have been in the works since before curation launch, and has already been made into a GIP. This feature would allow subgraph developers to publish and signal on a subgraph in a single transaction.

    • This would limit the impact of frontrunning bots (but would not eliminate them).

    • We might also see reduced signal volatility, as subsequent Curators would be pushed into an area of the bonding curve with less relative steepness.

    • For some more thoughts on the benefits of Batch GNS Transactions, I recommend these two forum posts: Post1 | Post2

Other solutions have been discussed as an addition to the Batch GNS Transactions

  • Subgraph Showroom - Shortly after curation launch, the idea of a “collective buy-in” for early Curators was discussed. The suggested implementation would expose the protocol to economic exploitation. Community members are looking into improvements and alternatives.

  • Dynamic Curation Tax - A solution whereby newly published subgraphs have a very high tax, discouraging Curators from buying at “Day 1”. This tax would then decrease over time. This solution would give Curators some time to assess expected query volume and find an “ideal entry”.



Curator resources



This month we speak with David from Synthetix.

Synthetix are using a microservice approach to building subgraphs, and are managing an impressive 46 subgraphs across 4 networks. We do a deep dive into the javascript code they use to manage their subgraphs.

We also talk about indexing Proxy Contracts. Synthetix have found a very smart way to track updates to a proxy contract that doesn’t emit events.

Watch the full episode, or jump straight to each topic using the timestamps below the video.





If you have any suggestions on topics we should cover or initiatives you’d like to see, we would love to hear it. Comment below, or contact me directly. You can also find me at discord Slimchance#1699

8 Likes

Whole lot of great information here! Much appreciated! :clap:t2::clap:t2:

3 Likes

It’s been a busy month! I just wanted to say I love being a part of this community and we are all taking strides towards building a beautiful thing. Keep up the great work everyone!

4 Likes

Hi, Slimchance.
I believe the information on Curator | Graph Docs (second graph/image) is wrong. 120,000 GRT does not mint 2000 shares, but around 346 (see Enzyme subgraph for a practical example). Likewise for the explanation below the image.
Afaik, the graph/image should reflect the following equation:
shares = signal^(1/2)

1 Like

Hi @wfqfw
Correct - the numbers for the example in the docs is not correct. I am not sure where the numbers are taken from. It is possible they come from an earlier iteration of the bonding curve.

Curator A would receive
SQRT(120000) = 346.41 Curation Shares

Curator B would receive
SQRT(480000)-SQRT(120000) = 346.41 Curation Shares

The Price as a function of Shares is
f(x) = 2x

The Total GRT Signaled as a function of Shares is
g(x) = x^2

1 Like

Perfect. Would be nice to see those equations in the docs too. Also, for anyone wondering about the relation between g(x) and f(x): f(x) = g’(x).

1 Like