2018-10-31 Engineering Meeting Notes
Community engagement plan and metrics
Background: There’s been a lot of research into Open Source projects about what leads to contributions, and an active community, and one of the biggest factors is response time on repository communication. Because of this, we’ll be trying to reduce our response time (and collect open source metrics in the first place) this quarter.
Desired outcome: A rough overview of how we’ll be reducing response times, and how we’ll be collecting important metrics.
- Aaron: Mary will be first responder and tag people that can help
- Ken: Does Mary have enough bandwidth?
- Mary: She’ll only respond to external people
- We can collect engagement data
- Zenhub is a tool that provide some metric
- OSStracker by Netflix that gives you a dashboard
- Ken: It’s important to track traffic and source of visitors & churn
User-owned Gaia Configuration Next Steps
Background: A major goal for this quarter is getting some mechanism in place for users to specify their Gaia hubs. Do we know our rough plan of attack for this?
Desired outcome: Figuring out if there are any remaining cross-product requirements, or if this is solely a browser/authenticator project at this stage.
Jeff, Jude, Mark, Hank and Ken are formulating a plan
Support user selectable storage on a Blockstack ID basis this quarter. Migration in the next quarter.
Continuous Integration / Deployment for Web Site
Background: We run a lot of web properties as an engineering team. Pretty much every single one of them has a different production and testing deployment strategy. We should have a unified system. We’d like to start collecting input from people who frequently deploy those web properties for decisions around this.
Desired Outcome: Basic requirements from (1) the perspective of the web properties producers and (2) the perspective of the devops team.
- pull-request/branch-based deployments are very helpful
- nice to have: ability to assign domain aliases
@Aaron B will start a private forum post on this
Background: Proposal: Gaia Hub Inboxes — this has come up in a number of conversations with developers, and we can actually implement this pretty easily. Are people roughly okay with that spec? Do we need to revisit?
Desired outcome: People should either +1 that spec, or we should work on hammering it out.
Jude: wants to do this now
Hank: proposal looks good, but not sure it solves the problem of push notifications
Jeff: Ranked very high in developer survey
Larry: what were the other top choices?
Jeff: very few people want to build their own on-boarding, some want white-label
Discussion on how an attacker could abuse this.
Conclusion: use an adaptive proof of work puzzle
Background : Larry is transitioning to working on a standalone browser. Is he going to continue as maintainer of blockstack.js or is that responsibility falling on someone else?
Larry will continue as maintainer for now and will have more to announce re-transition in the future