ouroboros-network VS cardano-sl

I'm one of the consensus team members at IOG. We maintain and extend the consensus layer of the Cardano blockchain, which provides a rich set of challenges related to distributed computing, type level programming, networking, concurrency, among others.

According to a document on pipelining I saw on Github they might be building a whole new Ouroboros protocol called Leios (https://github.com/input-output-hk/ouroboros-network/pull/3579/files) so there might be something we don't know yet

• cardano-node - top level integration code • ouroboros-network & ouroboros-consensus - network code & consensus (which integrates network, consensus and the ledger) • ledger - ledger(s) specifications and implementation • plutus - the smart contract language and related infrastructure

Check out cardano-ping.

This is non trivial, you can check an example how we generate valid subset of Haskell to check that our implementation of orElse is corect: https://github.com/input-output-hk/ouroboros-network/blob/master/io-sim/test/Test/STM.hs

So from what I gather the transaction fee is a fixed protocol parameter. I like how fees are distributed amongst the epoch, If I had to design a blockchain system I'd also have chosen this approach TBH. It addresses some of the issues and misaligned incentives if you only have to rely on fees for block rewards. However if the demand for transactions constantly outweighs the available space you get into other issues. I tried to find details on how Cardano handles transactions and manages its mempool. I believe somewhere here is the code? (https://github.com/input-output-hk/ouroboros-network/blob/master/ouroboros-consensus/src/Ouroboros/Consensus/Mempool/Impl.hs)

Ouroboros(the consensus algorithm's formal specification): https://github.com/input-output-hk/ouroboros-network

The daily developments were reduced by August but resumed vigorously in September. The whole of September was a busy month for developers as the daily submission rates for most coins were very high. At this time, ETH, DOT, and ADA were the leader as SOL followed closely behind. It was not till towards Mid-October that the daily developments reduced.

So, uh, the two things the blog post covers are actively awful practices. Let's look at Logging. Logging boils down to ServantLogConfig, which carries a function Text -> IO (). This is bad. Text is not a great type for logging, since it's relatively slow and encodings aren't guaranteed to be sensible (cf this bug where logging to a file was broken based on environment encoding). You want something like LogStr from fast-logger, or even a ByteString.Builder or Text.Builder even.

oops I want my Daedalus wallet (actually very surprising it's not there as Cardano uses Nix). Meanwhile this packaging request has been Open since 2017.

Otherwise you can use the following instructions for adding the caches to your local Nix install: https://github.com/input-output-hk/cardano-sl/blob/master/docs/nix.md#binary-cache

For what it's worth, coincheckup is tracking https://github.com/input-output-hk/cardano-sl which is the old repo from before they split up a lot of the infrastructure into their own repos.