Priming: Public Beta “Tech Talk”
This week, we launched Priming’s Public Beta, which you can explore at priming.xyz.
Priming is here!
We’re proud to share that we’ve officially launched Priming’s Public Beta!
medium.com
Priming’s team is led by RarityCapital, who is supported by a talented team of developers. Felix is one of them — he’s Paragons’ Lead: Software Architecture & Smart Contract Engineer.
In this article, Felix shares more about the technical requirements that drove this first milestone, as well as describing the back-end wizardry that empowers the new Priming. Enjoy!
BY PARAGONSDAO FELIX
Overview
ParagonsDAO’s Priming team has been working tirelessly over the past 7 months to deliver a second iteration on RarityCapital and BigInt’s Priming. I’d like to personally thank RarityCapital, Momo, Decoy, and Simon for their good vibes, dedication, and experience. What we have built so far has not been easy.
In this post, I will give some insight into overall system requirements and a critical piece of our system, precomputed metrics.
System Requirements
For our initial public beta, Priming will empower users in the following experiences:
Explorer
- Add arbitrary Explorer columns
- Group the entire Explorer (at the row level)
- Query on a suite of gameplay and financial metrics
- Query for scalar, percentage, or time series display of metrics
- Query over arbitrary time intervals
- Sort and filter on both gameplay and financial metrics
- Grouping and aggregation functions for metrics (at the cell level)
Single Card View
- View asset activity feeds
- View financial metrics
Authentication
- Authenticate and join multiple wallets to a single account
- Control default denominations and other preferences
All of these features are backed by a dataset of 9 million entries that cover many:
- EVM chains x Smart contracts x Transactions
- NFT marketplaces
- Token prices
- Parallel APIs
Precomputed Metrics
Today, most of our metrics come precomputed. That is, we process the data into aggregated tables, before a user asks for it.
Precomputing metrics enables faster responses, custom data transformations, scalability, and prevents your laptop from exploding.
Another point of complexity, precomputed metrics have dependencies. While precomputing a metric, we may reference multiple metrics. This allows us to create more powerful analytics.
Examples of Precomputed Metrics
Denomination Conversions For Sales
By precomputing this metric, we reduce query time (the query doesn’t need to join tables live), and do point-in-time price conversions (rather than converting prices only at the current moment, we convert it at the point when an exchange occurred).
Denomination conversions has two dependencies: sale transactions and token prices. ETH/USD constantly changes, independent to PRIME/USD.
These metrics are an important building block for the future Portfolio feature, which will track your profit and loss based on your historical activity.
Owner Counts of Parallel Assets, At Any Point in Time
By precomputing this metric, we’re able to reduce query time and ascribe cached/staked assets to their owners — an important metric for enabling deep insights into supply ownership.
The owner metric depends on the transaction table. Burns, mints and transfers are taken into account as we build this metric transaction by transaction.
How do you make sure that the rows you depend on are ready for you ?
Conclusion
We’ve built a strong foundation to tackle complex problems and package it up nicely for our users. We’re going to use both our foundation and learned experiences to guide us into whatever Rarity Capital throws at us next!
Pay attention to your SQL classes.
