Constellation is releasing its V1 base code next week. After a 5 month period of intensive development, we’re excited to share the engineering teams ground-breaking work, and we’re extremely proud of what they’ve achieved. That being said, the real work starts now as we iterate our next release and respond to the community feedback. We’d also like to thank the community for supporting us over these past few months, and in the words of our CTO Wyatt, “everybody can now see what we’ve been cooking up.” We can’t wait to hear your feedback, as we adapt and respond to your suggestions for V2.
This code release begins the process which will lead up to a small network of nodes communicating with each other. The focus will be on extensively testing the core system functionality and handling potential node failure conditions/recovery logic. Basic transaction handling and validation is the primary goal, along with demonstrating some of the fundamental data structures, system architecture, and infrastructure. Core functional testing is paramount, along with stabilizing the release with regression tests. We encourage developers to dive into the code and set up local nodes, however, these initially will not be part of the Constellation network.
“With the launch of our testnet, we’re really excited to provide our community with social proof of what we’ve been talking about for some time. The community has been really enthusiastic and engaged with what we’re building, and this is our chance to reciprocate the trust our community has instilled in our vision.” – Brendan Playford, CEO
Alongside the debut of our code, the second aspect of the release is a basic visualization and UI of the network functioning. The idea is to breakdown the very technical architecture of our protocol in a clear and approachable way. This will consist of a map view of where nodes are functioning, a simple output for a permission node to see how they’re working within the network answering, where were the block/transaction was formed and how it was bundled with other transactions and data. We feel that current DAG projects do a poor job of visualizing the tech and explaining how each piece of data is tied to together. Over time, will be expanding the functionality of the visualization. We want to showcase the underlying democratic principles of the network, self-organizing, and constructed to reveal data which is representative of how nodes are behaving.
Engineering Update – Ryle Goehausen (VP of Engineering)
Below is a bullet point overview the engineerings team main focus for the week:
- Progress on speeding up the downloader.
- Substantial refactoring and code cleanup before release.
- Fixed several crash conditions and node failure edge cases.
- Finalized V2 schema for upcoming transition from directly representing graph sub-partitions to raw graph edges.
- More work on isolating processing logic to more granular actor handlers.
- Fix edge case on ancestry resolver causing stall condition.
- Updated file handling library and pruned unnecessary dependencies.
- Optimized cell handler to remove the redundant group by operations and added more sanity checks around data resolver and cleanup.
- Added docker internal health check to the container.
- Updated API client to scalaj instead of akka-http. More work on internal authentication around endpoints.
- Added memoization of several calls including metrics rendering for dashboard.
- Fixed memory issue around excessive bundle data. Updated bundle scoring functions to discourage stray chain formation.
- Added JMX and debug ports for JVM profiling. Ran numerous performance tests and collected data for debugging purposes.
- Parameter tuning around gossip, mempool handler, and cleanup functions.
Orion Update – Brion Hickey (VP of Product)
Join the Orion Portal
Sign up on Telegram