Standup Notes 2019 02 08

Participants (alphabetical): Allie, Conor, Emmanuel, Erik, Heartsucker, Jen, Kevin, Kushal, Mike, Mickael, Nina

Allie

Yesterday:

• https://github.com/freedomofpress/securedrop-client/pull/242

Today:

• Debug why on_source_changed is called for each previously-selected source when refresh occurs (this ends up calling show_conversation for the same source more times than I believe is necessary). If we can simplify refresh logic then it'll potentially improve performance when there are many sources as well as make debugging gui issues much easier.
• Start a conversation around what behavior we should expect on refresh (or find existing conversation somewhere).
• Read up on Qt development, especially regarding classes we use for the client
• Start on https://github.com/freedomofpress/securedrop-client/issues/238

Blockers:

• None

Emmanuel

Yesterday: Some support, non-SD work

Today: More support, but also #learning if time permists

Blockers: None

Erik

Yesterday:

• Spent a fair bit of time working through https://github.com/freedomofpress/securedrop-client/issues/187 with Nina and Jen to figure out the correct behavior. tl;dr: Client will not use existing read/unread and will instead require its own per-user tables on the server.
• Made a video for Nina to show how Nautilus behaves on Qubes as part of ongoing prototyping of export workflow
• Non-SD stuff

Today:

• More on client issue #187, start draft of server-side issue
• A bit of Xenial related prep for QA, docs
• May continue my own physical install adventures as time permits
• Non-SD stuff

Blockers:

None

Jen

Yesterday:

• Review/merged heartsucker reply PR
• Wrote a diff for the base OS metadata endpoint to get tests passing
• Investigated the haveged apparmor issue, looks like the upstream profile needs read/write capabilities for one file adding (there is an upstream bug for this), and then there is some other reboot time race condition where the haveged process is starting prior to apparmor (so will not be protected via apparmor)
• Allie and I chatted a bit about how to best test her fix for the message bubbles appearing on top of each other upon update_conversation getting called again.

Today:

• Review of https://github.com/freedomofpress/securedrop/pull/4101 is stalled, will address the final testing point there so we can merge
• Pick another xenial ticket back up as time permits

Blockers:

• None