Todd Benzies <tbenzies@...>
Technical Steering Committee (TSC) Meeting
February 18, 2016 (7:00am - 8:30am PT)
Deutsche Boerse Group
Richard G. Brown
5-10 mins: Statement from each TSC Chair nominee (3-5 mins per nominee)
10 mins: Code of Conduct discussion (Chris Ferris)
10 mins: Intel contribution overview (Mic Bowman)
5 mins: Draft project proposal template (Stefan Buhrmester and Vipin Bharathan)
30 mins: Discussion on basic use cases to address
20 mins: Initial discussion on how proposed contributions could be converged
TSC Chair Elections
Begin nomination period: Today, Feb 11th, nominations includes a short bio/pitch less than 1 page
End nomination period: 9pm Pacific Time on Wednesday, Feb 17th
Send slate of respective nominees to TSC Members: Thursday, Feb 18th
Send TSC members voting instructions: Thursday, Feb 18th
Voting period ends: Tuesday, Feb 23rd at 9pm Pacific Time
Election winners announced: Wednesday, Feb 24th
Code of Conduct Discussion
Reference point to start: https://www.cloudfoundry.org/foundation/code-of-conduct
Question: Materials looks great, but how has it been applied?
Question: How do you determine if complaint is warranted? Also, some language could be overreaching -- i.e. silence is taken as assent (if you are silent, you are somehow agreeing with poor behavior).
ACTION: Mic Bowman, Chris Ferris -- create a draft and then send to the TSC mailing list for review and approval.
Intel Contribution Overview
Presentation deck is attached.
This project is a complete ledger, being used for real applications.
Principally architected as a research platform.
Absolutely everything is plug and play.
In process of hardening code.
3 basic layers: communication, journal, & ledger
Intel focuses on open marketplaces and consumer marketplaces (both extrinsic and intrinsic)
Built in a complete way with purpose to ensure we get right interfaces for APIs but also to ensure consensus algorithms were robust enough
Testing via internal trial (Football Exchange) -- exchange shares of teams
Implemented in Python
Unique aspect: looking for ways in which we can take advantage of the secureguard extensions
Intel interested in evaluation platforms based on workloads so that requirements can get discussed
Have a variety of suites of tests
Finding ways to use crypto-acceleration that Intel uses
Q: Testing -- way to simulate network partitions?
Partition tests are done by hand right now. So, yes, but not automated.
Largest test is ~2,000 validators running in a cloud.
Q: Control over latency? Force some participants to be high and some low?
Q: How long until code will be ready?
DRAFT Project Proposal Template
The seeds of a new project proposal has to be vetted in a public forum like hyperledger-technical before creating a project proposal.
It needs a technical champion who believes in the project and who should be the technical lead on the project. The project champion can change in the middle of the project.
The proposal template for HyperLedger Improvement Proposal (HIP):
Author(s) name and contact details including email address, a short name for the project (less than 10 words).
Abstract (less than 50 word) description of the project.
The description and need for this project, substantiated by:
Technical details Effects on
Transactions- including confidentiality, signing, traceability, identity of participants, contracts (scripts)
Effects on User facing Clients that help with transaction formation (similar to Wallets in BTC)
Effects on the network, throughput, visibility to other participants, change in protocol if any, criteria for network participation
Block formation and ledger formation viz. Consensus algorithm, size overhead, effects on the throughput and rate
Cryptographic implications if any.
Backward compatibility (hard fork or updates by all network participants needed?)
Rough design and thought experiment (Gedanken Experiment) on the probable effects, if any
Address any possible objections and also support that came up during seed proposal from technical community on the lists.
Traceablity, testing criteria to gauge effects on installed base.
Any other technical details, including languages used, other technology needed and (preferably) public sources for additional tools and binaries
References and defense of why this is needed, especially if the same ideas have been already addressed in any other project.
Rough gauge of effort and resources committed (coders and any other resources that are needed) and timeline if available.
Readability which means following certain conventions. The style itself could be controlled by Chicago Manual Of Style say, explanation of abbreviations, references, diagrams. A proposal editor appointed from by the tsc could help with this. As technical people we often do not think about this aspect. This is extremely important as the clear statement of the problem and its technical details are helpful to elicit a solution in the community and prompt volunteers. This can be helped by having a proposal builder software or a template itself.
ACTION: TSC to weigh in and provide guidance on what they expect to see. This is a project proposal template that will be used by the TSC to evaluate. Most projects have top-level projects, as well as sub-level projects. Note: may also want to add a project lifecycle. (i.e. incubation, core, mature, etc.)
Points of reference
Initial Discussion on How Proposed Contributions Could Be Converged
Chris Ferris: Yes, this would be good. Would like to get to consensus on initial starting point and a near term of proposed activities. Could then actually get everyone together from a dev perspective to get things started. Potential to host space in Raleigh for short period of time, to contribute to initial sprint.
Suggestion: Would like to see a sufficiently diverse set of use cases covered to meet the needs of all participants.
Agenda Draft for 2/25 TSC Meeting
- DAH/IBM to discuss proposal which will be sent in advance of the call.
Senior Program Manager
The Linux Foundation
+1 (415) 412-0310 (m)