Discussion about the technical spec of the EVM

evm

#1

@CJentzsch tweeted about this important issue, and several people commented.

Initially the discussion focused on barriers to participation due to lack of formal background required to work on the Yellow Paper. @Arachnid commented:

At one point @cdetrio mentioned the Jello Paper., “Human Readable Semantics of EVM in K”:

Do I hear the sound of a a ring forming?


#2

Here is a related comment on reddit. I am interested in why @maciej believes the spec is “from EF”.

This is not a quality issue, it’s a communications issue. Parity is a 3rd party following technical specification from EF, we are less involved in writing it than Geth team is. The end result is that there might be hidden assumptions in the spec that not everyone outside of people involved in writing it are aware of. That’s just the nature of things, writing complex software is hard, writing technical specification for complex software is hard. That’s why we have a test net.


#3

I honestly haven’t looked into the Ropsten issue much, and based my response there over a casual conversation I had over lunch. Thanks for clarifying things on Reddit!