The state of Gossamer July - Mid August

[P1sar]

Highlights

JAM protocol implementation

Gossamer team is targeting to implement the new JAM protocol!

During this perioud we completed our design phase of JAM first milestone implementation.
With help of ChainSafe RnD team we already finished some of the modules that are passing JAM test vectors as well as new JAM encoding scheme (replacement of SCALE) and Safrole.
We've also made significant progress implementing a PVM interpreter as well as researching and a developing compiled PVM runtime for future milestones.

Big shoutout to Jimmy and Eric for making this happen!

Eclesio is doing research on using JIT compilation for PVM which would allow for much faster execution times than using an interpreter. This might make block executing time way faster than interpreting the bytes

Hiring

Gossamer team is super close to finish this round of hiring. We have two great candidates almost finished with their last technical assignment. We might welcome a new Protocol Developer to our Gossamer team soon.

Infrastructure

Together with ChainSafe infastructure team and with Eclesio and Jimmy we have overhauled our automated CI/CD and regression testing platforms.
As well as we are wrapping up Gossamer infrastructure migration from AWS to bare-metal servers that will dramatically reduce our costs of infrastructure usage and eliminate AWS vendor lock.

Web3Summit participation

We are actively preparing to participate in web3summit as a nodespace where we are going to talk about client diversity, our JAM implementation and other aspects of protocol development.

Articles

Big shotout to Diego for writing amazing technical article about Polkadot Storage.
More articles are coming soon, stay tuned...

Development

Syncing with Paseo and Westend

Gossamer Host has now reached Paseo tip of the chain and keeping up with the chain though the Uptime SLI is still not production ready (< 99.99%).

Eclesio is doing outstanding work in fixing slowdown problems during our sync with Westend.
Currently we are synced with 20.4M blocks and are very close to be done. Though we are still facing a new Storage root must match that calculated panic that we are debugging, as well as an unexpected slowdowns in syncing at certain block heights.

Warp sync

Eclesio finished design phase for WARP sync implementation and we are getting ready to start implementing it in next cycle.

Polkadot Storage

Diego wrapped up our implementation of lazy loading trie implementation for Gossamer storage, this will dramatically decrease node start up time and reduce memory usage.
Tim Wu is working on optimizing underlying gossamer storage code and particulary already have improved interfaces around storage Database primitives. This integrates Diego's lazy loading trie, manages child trie support, as well as caching and state proof generation.

Polkadot Parachain protocol implementation

Axay has finished implementing Candidate Backing subsystem that is responsible for verifying candidate blocks submitted by parachain Collators. Currently he is wrapping up with integration tests that will ensure subsystems works properly together with other parts of the code.

Ed is doing exciting work on implementing Candidate Validation subsystem that is crucial to most of the Parachain block proposing phases by validating parachain block against runtime exection.

Kishan is wrapping up implementation of Network bridge subsystem that will coordinate all networking around subsystems and a crucial part of our Parachain protocol. This is a huge part of our work that we are intend to finish during next cycle.

Kishan, Axay and Ed have spent an enourmous amount of time educating our team about Parachains protocol by conducting workshops and writing internal documentation and articles. All this will be used in nearest future to release public articles around this topic.

Product

We are currently conducting product research that is focused on identifying the top challenges faced by node operators in the Web3 space. Through targeted surveys and interviews, we’re focusing on key pain points and exploring the critical areas where developers encounter the most difficulties.

Our objective is to determine how Gossamer can effectively address these issues through potential features, services, or tools. The feedback gathered will be instrumental in shaping future developments of the Gossamer as the Polkadot ecosystem evolves.

We invite all node operators and Web3 developers to participate in our survey ! We would greatly appreciate it if you could take a moment to share your insights with us. For any additional inquiries, please reach out to us at gossamer@chainsafe.io.