The DialNet testnet, launched this morning, contains some improvements to the issue of dialling unroutable peers - the one that was causing some nodes to remain chunkless. We have what we believe to be a fix in place now which we want to test out. It will still need some tweaking for performance, but we want to try it in the real world.
This week we’ve also been looking at the API and developer experience. We’re getting close now and we really want things to ‘just work’ for devs. This will be our primary focus over the coming weeks.
We’re also at the point where we can start seriously examining data replication, incentives, updates and resilience at scale. It goes without saying that the core functionality needs to be robust. Our approach is simple: simplification, simplification, simplification.
Looks like there are some improvements in libp2p coming in, including UPnP support which should allow some nodes from home. We try not to get excited by these developments, but we are quite excited!
General progress
@JimCollinson and @Andrew have been hard at work on product positioning. What should our initial approach be? Who are the main user groups and how should we speak to them? Should be something to share shortly.
A big PR has gone in with changes and tweaks to allow pay-per chunk. This means we now send to each node the payment it has asked for instead of trying/failing to randomly sample the network. This should make uploads more reliable.
@Bochaco is back (yay) and working on storing node rewards in local wallets, plus a CLI command to query node balances.
@Qi_ma continued investigating file upload issues and test failures with benchmarks. These are mostly fixed now. He also resolved a few CI failures on the pay-per-chunk PR.
@Bzee has been looking at the dialling unroutable peers issue and has implemented a fix. This is the main issue we want to observe in the testnet.
@Roland worked on client/node record stores and fixed a bug in the workspace.
@Chriso has been using GHA workflows to create and destroy testnets automatically. He’s also looking into alternatives to logstash, which we use to forward logs to S3 but which is turning out to be a bit of a resource hog.
And @joshuef has refactored the wallet code to improve performance and also worked on fixing the slow uploads issue.
Useful Links
Feel free to reply below with links to translations of this dev update and moderators will add them here:
Russian ; 
German ; 
Spanish ; 
French; 
Bulgarian
As an open source project, we’re always looking for feedback, comments and community contributions - so don’t be shy, join in and let’s create the Safe Network together!
