/
Project Planning Feb 22nd, 2018

Project Planning Feb 22nd, 2018

Owned by Medha Parlikar (Unlicensed)
Last updated: Jul 20, 2018 by Henry Till

Date

Attendees

Goals

Discussion items

TimeItemWhoNotes
10 minInterpreter status - what features will be ready for 3/15
Kyle Butt, Henry Till
  • Open work items, integration of storage layer with Interpreter, namely when do we swap rholangADT.scala for protobuf-generated code.
    • Once he gets enough of the normalizing - will have to happen next sprint.
  • Needs visibility on what is working on the storage layer
    • Henry Till - Implementing the tuplespace in Scala.  Working off Kent's python's example.  
    • Yaraslau and Eitan need to get in sync and in touch with one another.  There is a dependency on the matching algorithm.  
  • Integration testing has not been done yet.
  • On track for 3/15 or ? - Not on track.  
10 min Node Status - what features will be ready for 3/15 Chris Kirkwood-Watts
  • Working through https service with Henry.
  • Dependent on what the other teams can deliver.  Everything else is solved. 
  • Docker is ready.
  • Need rpm and debian packages, code zips as well.  Planned for next sprint.
10 minVM - Status and review Plan 'C'Timm Schäuble
  • Talk through plan to implement multi-threading - Need to talk to Greg about multi-threading.
  • Talk through plan to stop the translation with the transition function + primitives.
  • Outcome - no plan C - we rely on RBL libraries in order to get the VM functioning.
  • Timm is going to send Greg a note to figure out multi-threading.
  • Timm to go learn about the compiler.
10 minLast meeting rolled over itemsTimm Schäuble, Henry Till, Kyle Butt
5 minDepencies Discussion
  • Lay out any dependencies that you may have on other teams work.
  • Storage -Interpreter on matching - Yaraslau & Eitan need to work together.
10 minNode.Void ReleaseEveryone
  • Storage is a library consumed by the interpreter.   No access to the storage tier.
  • Kyle wants to have a well defined api & some matching code. 
    • Dependency graph - demonstrate storage is functioning.  
    • Kyle doesn't feel comfortable that he will be ready by the end of the next sprint with matching. 
    • Wants to see Joe's interpreter churn through and provide a result.   
    • No integration testing done with storage & interpreter. - This piece was not properly budgeted.
    • Storage API has a slightly different shape for produce & consume- proposal is in the specification.

Target for the next release is March 29th.  Cut packages on March 26th  - 

Asking for additional sprint for integration testing, Storage.01 timeline is unchanged.

Interpreter Option

  • Kyle believes he can cut an SDK  - we won't have a process matcher in that time. So some things will be possible in this SDK, others won't.
  • Goal is to offer the same functionality as the SDK 0.1 release.  He doesn't want to release if the interpreter is a step backwards from the SDK 0.1

Action items

From Last meeting

  • Medha Parlikar (Unlicensed): Future feature to support multiple cores - Post Mercury (assumption is that there will be a single storage tier for N VM's / Interpreters)
  • Medha Parlikar (Unlicensed): Schedule a meeting with Chris, Henry and or Nash to discuss how the pieces come together in the node.  Create a spec document that will likely feed requirements into the interpreter.  

New action items

  • Kelly Foster: Follow up with Nash to get tickets updated for build packaging with details.  
  • Henry Till: Follow up with Kyle to on Storage.  Walk through the API.
  • Kyle Butt: Update the Interpreter specification to outline what the interfaces is for Node to call the Interpreter.
  • Medha Parlikar (Unlicensed): Communicate change on the Node.Void release to Nash
  • Medha Parlikar (Unlicensed): Talk with Greg and invite him to this meeting, meeting time may change as a result.
  •  

Related content