Standup Notes 2018 03 22
Participants (alphabetical): Conor, Erik, Freddy, Harris, Jen, Kushal, Mike, Mickael. (Loic via Gitter.)
Yesterday: Have been blocking on kernel source offer docs, split and completed one of these tasks.
Today: Coordinating with Loic, continue
Blockers:
Yesterday: Sprint planning ;-), some backlog grooming, non-SD + support work
Today: Support follow-up, digging into SD.org content requirements, build basic burndown chart
Blockers: None
Yesterday: Looked at kernel story.
Today: Support. May help more with Signal CLI feedback
Blockers: None
Yesterday: Securedrop.org: Put in PRs for directory design changes and ForeignKey field restructures
Today: Gotta chat with Conor about the directory design change PR
Blockers: Conor gotta rereview the directory design change PR
Yesterday: Caught up on GSoC student inquiries. Improved docs release process. Started running through signal CLI branch
Today: Continue on signal CLI branch
Blockers:
Yesterday: Started working #3064, attended sprints meeting.
Today: Fixed the test errors for #3064, review found more missing update, will fix those.
Blockers: None
From Gitter: "I will miss the standup/engineering meeting today. I scheduled an i18n AMA, did some UX related work, kept working on i18n_tool and discussed with GSoC applicants."
Yesterday: Writing aobut the Signal CLI branch. Finalized and tested PR for install time removal of vanilla kernels.
Today: Looked at build process. Would really like anyone who reviews it to pay close attention to the way that it is built, especially the sandboxing of Java. A little concerned - not entirely confident that the right move would be to do it right now.
Blockers: None
- Signal CLI discussion - see ticket https://github.com/freedomofpress/securedrop/issues/3182
Package is a fork of libsignal - pretty much a one person team. Simply doing device management, no change or modifications to any code that does transmission of code or encryption. Java dependency might induce operational/security issues.
OpenJDK 7 is EOL in June 2018. Using OpenJDK 8 would introduce a lot more complexity.
Would prefer to have stricted sandboxing than there is now.
Building/packaging: need gradle. Pulls all the dependencies from the net. Can't build gradle from source without using the gradle binary. Five additional dependencies.
Conclusion: defer work other than testing for now until Xenial migration
- Mozilla Sprint discussion
Conclusion: positive opportunity, but date conflict with 0.7 release is problematic for anything more ambitious than i18n sprint. Happy to support i18n as an experiment though