Hey Folks - One of the goals for the initial release as captured in https://issues.apache.org/jira/browse/LIVY-898 is to help clarify the contribution process. This is to help folks that would like to contribute to fixing issues and adding features to get started.
One of the items there is to help describe a dev environment setup. I'd love to hear from folks in the community around a couple things: 1. What the minimal dev environment would look like preferably with local deployments 2. Whether we have different types of dev specializations - such as: notebook devs, API devs, server, devs, etc - and we could determine whether we need to represent each or some or some lowest common denominator 3. Do we use docker for things like spark and yarn dependencies for these environments 4. Do we assume all livy contributors have an existing strong background in spark I can say that I am personally used to relying on a remote deployment within my workplace (usually via VPN) for both Knox and Livy testing. I don't know whether this is a common dev env scenario or not but it does seem like it could be a barrier to entry for those that wouldn't have such a deployment available. I think that our recent build improvements with docker have been really good and am wondering whether a similar approach with system dependencies would work for dev environments. What are your thoughts? thanks, --larry
