Benjamin Diamond

Ah, so training will happen on the server, and *only* self-play will occur client-side. Nice. @Zeta36, do you think this would be easy to implement? @brianprichardson I think the key...

@lundmikkel thanks for raising this, it's in important issue. You approach is interesting. Here are a few comments: * It could be a bit of a drag to have to...

The other option would be to simply use seconds as standard ethereum and IBFT already do, and then disable the requirement that new blocks' timestamps be strictly greater than their...

hey @lundmikkel, just following up on this. I've written a new branch of quorum, called [`seconds`](https://github.com/jpmorganchase/quorum/tree/seconds), which solves this issue. Just add `"raftSeconds": true` within the `"config": { ... }`...

I have succeeded with regular datagrams---just not raw. Here is the simplest scenario I can find---adapted from Kalle Vedin's [sam3](https://bitbucket.org/kallevedin/sam3/src/master/). First, clone his repo into `"$GOROOT /src/bitbucket.org/kallevedin/sam3"`. Once this is...

Excellent. Let me know if I can help in any possible way.

still same problem :( **EDIT** maybe not. give me a minute more to test.

@orignal no more segfault! But now, still not successfully receiving the raw datagram in the server session. To test, please clone from my fork instead: [benediamond / sam3](https://bitbucket.org/benediamond/sam3/commits/). In particular,...

ok. Actually, the segfault appears to be build-dependent... On Ubuntu 16.04.6, when I build from source using ```bash apt install -y cmake libssl-dev libboost-all-dev cd i2pd/build cmake . make ```...

works!!!! in certain situations. i am still doing some testing, one minute.