Thread by @hus_qy

I think it's time for a short update around our progress on coordicide:

A few weeks ago we merged the refactored consensus code base and we have been running it in an internal testnet since then.

After fixing a lot of bugs, the node looks increasingly stable (we also found ...

... the memory leak that we were fighting with for almost 2 weeks - people who closely follow the development process on github will know what I mean).

The only remaining thing for the prototype to be feature complete in a first MVP version (apart from getting rid of ...

... possible remaining bugs) is the chain switching, which allows nodes to automatically recover after i.e. having being eclipsed / in a minority partition.

Me and Andrea started working on this 2 weeks ago but we had to pause and first change the way we manage state to ...

... simplify the necessary logic for the switch. Everything else is already working - snapshots, commitments, pruning, and so on.

We will most probably finish the state refactor in the coming days and then start to work on this very last building block but since christmas is ...

... approaching fast, and people have to take their remaining vacations before the end of the year, we will most probably start to run thin, soon.

I am officially already on vacation as well but since there isn't much sunlight and it's pretty cold - I am coding anyway.

There is still some old code that needs to be cleaned up even after finishing the MVP, but we are mostly talking about performance optimizations like adding a caching layer in the ledger and so on (which currently operates 100% on disk) but this can be done after everything ...

... is stable.

So TL;DR: We are getting really close now but since bugs take some time to find / fix and christmas is close, it's a bit hard to predict how much we will be able to progress in the coming days and weeks.