The Graph’s Community Talk #21 was held on Tuesday, March 21st @8AM PST.
The Graph’s Community Talk covers a wide range of subjects from all corners of our network and provides attendees with a comprehensive recap of the latest happenings within The Graph ecosystem. The meetings are held in The Graph Stage channel of the Discord Server.
Tip: Go to the Events channel at the top of the Discord server and set yourself a reminder for The Graph’s Community Talk.
Community Talk #21 Topics Include:
- Community Updates
- File Data Sources
- EthDenver
- Chain Brain
- The Graph x Base Partnership
- Subgrounds
- Layer 2 Scaling Process
- AI & The Graph
Detailed notes regarding each topic with timestamps can be found below.
The Graph Community Talk is geared towards a broad audience regardless of background. We are looking for active participation from the community and are looking forward to your engagement!
You can find this and all other community events on the ecosystem calendar. Subscribe to the calendar to stay up to date on future events!
Community Updates (00:10)
In Community Talk 21, a variety of updates were shared with attendees. The event started with a welcome message and introduced a lineup of speakers who were going to discuss file data sources, Base’s test net launch, Eth Denver, Subgrounds, scaling on Arbitrum, and AI and crypto. The talk then moved on to community updates, where 24 new graph advocates were welcomed from Nigeria, the United States, and the Middle East to make great contributions to The Graph ecosystem. The Graph will also sponsor $7.5 thousand dollars in prizes at Eth Tokyo and $10,000 in prizes at Eth Lisbon.
The Graph Foundation released their grants report for 2022. Also, check the link below for the multitude of positions available within The Graph ecosystem.
The Graph Foundation - 2022 Annual Grants Report
The GRT IQ podcast released four new episodes, discussing topics such as The Graph’s move to L2 and AI and crypto.
The text translators in the advocates program completed the translations of The Graph’s documentation to Spanish, Japanese, Portuguese, and simplified Chinese, which still need proofreading. Community members who want to help with proofreading can reach out to members of The Graph AdvocatesDAO.
File Data Sources (04:00)
The Graph is great for on-chain data, but Dapps also need to query offchain data. File Data Sources – shipping to Indexers in the next release of graph-node – allows Indexers to sync data from files stored on decentralized storage networks and will initially support IPFS.
During the first segment of Community Talk 21, Simon from Edge & Node discussed the new file data sources feature on the decentralized network. This feature enables developers to seamlessly fetch and filter NFT’s by metadata – and other files – and is one of the most exciting features recently launched with graph-node version 0.30.
Previously, subgraph developers were able to download data from IPFS and parse their subgraphs, but the process was blocking, which caused problems with indexing for blocks. The new file data sources feature allows the process to happen in parallel. The main shift is that when a subgraph attempts to download a file from IPFS, default processes continue indexing while a background process retrieves files. This new feature also opens a wide array of new use cases for subgraphs and The Graph.
For more technical details on the implementation, interested individuals can read here: Creating a Subgraph - The Graph Docs
Eth Denver (8:15)
In the second segment, The Graph held an extensive workshop at the University of Colorado at Boulder during Eth Denver. Numerous speakers discussed The Graph and its uses for students. A small hackathon was also hosted for hands-on learning. The Graph team had a booth at Eth Denver, attracting 40 visitors and engaging with thousands of attendees and hundreds of protocols and projects. They sponsored hackathon bounties for the best new subgraph and best use of an existing subgraph. Chain Brain was one team that participated in the hackathons at both the University of Colorado Boulder and Eth Denver.
Chain Brain (10:03)
Chain Brain, a no-code analytics platform, was developed during the UC Boulder hackathon and Eth Denver. It enables users to search and generate visualizations for blockchain data by asking Chat GPT questions. Built on The Graph Protocol subgraphs, the platform uses an AI model to interpret questions and create Graph QL queries. Results are displayed visually and can be saved to a dashboard. Chain Brain can be used for various purposes, such as identifying daily NFT trade volumes or finding the top USDC pools by volume on Uniswap. The team is currently seeking a machine learning engineer with natural language processing expertise to help further develop the platform.
Check out the Demo!
The Graph x Base Partnership (14:58)
At Eth Denver, a partnership between Base (Coinbase’s new Layer 2) and The Graph was announced. This allows developers to deploy subgraphs on The Graph Studio for dApps on Base’s testnet. A workshop was held demonstrating how to deploy a subgraph on Base using The Graph Studio and CLI. The Graph CLI version was updated to support Base, so users need to ensure they have the latest version to deploy subgraphs on Base’s testnet. The partnership is significant as Coinbase is a prominent exchange globally, and the OP stack chain is likely to attract interest from various institutions and dApps that will utilize The Graph for indexing needs.
Subgrounds (18:45)
The fourth segment introduces Subgrounds, a tool designed to access subgraphs using Python instead of Graph QL. Tachi, the CEO of playgrounds analytics, explains that Subgrounds aims to make it easier to use subgraphs for data analytics. Subgraphs have several advantages, including standardization, scalability, and flexibility. However, the complexity of Graph QL hinders their use for data analytics. Subgrounds addresses this by offering a simple, open-source, and familiar solution. It allows users to load subgraphs, run queries, and process data in a Python environment, generating tables, data frames, and charts. Subgrounds is easy to use, customizable, time-saving, and compatible with popular Python tools like Plotly and Pandas, making it suitable for data analysts and data scientists.
Subgrounds | Join Their Discord
Check out the Demo!
Layer 2 Scaling Progress (31:05)
The L2 bridge and the The Graph’s network have already been deployed to Arbitrum. There is an audited PR to enable L2 Indexer rewards which is going through formal testing and then will be put forward for a council vote (hopefully in the next couple of weeks).
A layer 2 migration for The Graph aims to lower gas fees, make indexing cheaper, and enable more complex on-chain mechanisms. The transition plan involves multiple stages, starting with deploying L2 protocol without indexing rewards and a bridge for sending GRT from the main net to L2. Indexing rewards will eventually be enabled on L2 (upon the council vote), incentivizing Indexers to serve subgraphs on Arbitrum. Migration helpers will assist in moving tokens, signal, and delegation from layer 1 (L1) to L2. The process will affect all network participants, who will have to make decisions about migrating their activities to L2. There will be an option available for Delegators to skip the unbonding period if an Indexer fully migrated to L2 and the Delegator does not want to migrate their delegation to L2. Follow the progress on the relevant governance proposals and reach out with questions on Discord or the forums!
AI & The Graph (43:25)
Sam Green, of Semiotic AI, discussed the deployment of AI tools in The Graph protocol and the relationship between artificial intelligence and The Graph in the future. Two products currently use AI methods in-protocol: Auto Agora and Allocation Optimizer. Auto Agora is a tool for Indexers that automates the measurement of resource costs and uses reinforcement learning to adjust query prices, maximizing revenue. Allocation Optimizer, a collaboration between Semiotic Labs, Edge & Node, and Graph Ops, helps Indexers solve the problem of selecting which subgraphs to Index. Both tools aim to save time and increase the quality of service in the network.
Into the future! There have been many advances in large language models, such as ChatGPT, and their potential to expand the use of The Graph. In the future, AI models could use data verified by The Graph. Currently, large language models scrape data from sources, which sometimes have contradictions and can lead to inaccurate information. The Graph could provide these models with the information they need to provide accurate information to its users. The Graph could also be used as a source for training machine learning models and neural networks in the future.
Stay Tuned!
Join us next month for Community Talk #21!
Keep up to date by joining discussions in the forum, following The Graph on Twitter or joining the Discord server.