Thomas Eizinger

Yes, thank you for laying it out clearly @stormshield-frb ! As @PanGan21 said, I think it makes sense to focus this PR on (1) & (2). Using an exponential backoff...

> 1. we bootstrap on new connections until one succeeds then go to `2` What I wanted to add to this is, we should only start the bootstrap iff the...

> 2\. On start-up of a node, some users call `Behaviour::add_address` without actually connecting to it (i.e. without `Swarm::dial`). Reacting on new-node-inkademlia-routing-table instead of new-connection would cover this use-case. I...

> From what I understand the condition `new-node-inkademlia-routing-table` is actually true when there is an `Event::RoutingUpdated` with `is_new_peer: true`. Correct me if I am wrong or missing something That sounds...

> > LGTM, thank you very much! :) > > Will give @mxinden a few days to review this too. > > Still need to fix the failing tests 😅,...

I am running into a similar issue. For me, Firefox replaces the `search.json.mozlz4` file upon start up. I did a little bit of digging and the file generated by home-manager...

> The main functionality of this action is to publish AUR package, not to build it. But it does offer `test`! Unfortunately, I cannot use `test` for my package because...

Yep that is currently not really well supported. We need to rework the API for this to work well: https://github.com/testcontainers/testcontainers-rs/issues/386. Contributions welcome!

Whilst that may end up being better for now, non-async state machines would compose even better than that and would likely be more performant and more idiomatic Rust! Channels are...

> Here's a starting point for a Sans I/O implementation. https://github.com/algesten/str0m Amazing! Thank you for starting this! Let me know if you need financial support for pushing this forward. I...