Andrey Chursin

Hi I came across this issue too. I wanted to compile ring for PowerPC, in order to use it inside HSM custom code, but compilation fails with ``` thread 'main'...

This is not really a breaking change right? Aside from need to update sui code it does not break validator compatibility. E.g. "breaking change" is a marker for PE that...

I see that this is still a draft PR, but have a quick comment. I am a bit worried about adding metadata fields to core data structures that are not...

>In the new world of owned certificate sync via Narwhal, we no longer have a need for execute_cert_to_true_effects(), right? Can you explain a bit why?

Other node will see certificate when narwhal is committed. If everything is normal then it is around ~commit latency. But of course if node have network connection problems or do...

To elaborate a bit more on what I think is the problem with this diff - Current logic in `ConsensusAdapter` is this: - Register oneshot with waiter - Submit transaction...

Yeah, the general reasoning that @huitseeker provides is correct, have we executed everything in non-concurrent environment, but since we do different things in different tasks there are race conditions with...

So the problem does not happen now, as you always wait after submitting txn and because we always register waiter before submitting request there is no room for race conditions...

I don't think single channel fixed the race condition. The race is because we check the database first, then registering consensus waiter, so between those two operations something can get...

For the `le` vs `be` policy - I think it is natural to use `le` everywhere, although it is not super critical for tracking_id, since it is not part of...