Cycle plan: October 1st - October 31st

[rudolfs]

• Experiment with org-node to get replication to work. Can we make it work reasonably well in a reasonable timeframe? (@geigerzaehler)

UI improvements:

• Expand Search design #1031 (@rudolfs)
  • Fix hints in onboarding input fields #987
  • Instant navigation to replicated project #1356
  • Update error state when user has entered a Device ID into the search box #1456
• Separate locally stored transactions by network #2230 (@rudolfs)
• Profile hover cards #2235 (@rudolfs)
• Add What's New popup #1865 (support from @geigerzaehler)
• Add org dropdown to anchoring modal #2419
• Add loading indicators #1596 (@rudolfs)
• 💪 Separate any data stored in Electron localStorage by profile ID #2263
• 💪 Org Profile Hover Card #2245

Tech debt:

• Use rad-identities to deal with session information #2459 (@geigerzaehler). We prioritize making replication work with the old setup
• Make the design system exportable #1877 (@rudolfs)

Design

• User badges #2234
• Explore a shadow system for hover/active states #2441

[rudolfs]

Ideas for what to focus on this cycle:

• now that we have an org node, automate pushing latest changes from the radicle-upstream repository's master branch to the org node
• since replication is sketchy and we're waiting for v3 of the protocol to land, a potential workaround is to build project import/export functionality in Upstream which would allow us to transport projects via centralised means, this would also solve a use-case outlined here: Could not create project: the URN already exists #2343 (comment)
• spend some time to familiarise ourselves with the multi-sig nature of radicle-link and come up with a way to integrate it into Upstream, because this is a requirement for any further collaborative objects work as outlined in Dogfooding radicle #2379 (comment) which, in turn we need to get to "dogfooding"

[juliendonck]

Things that are missing here from the top of my head:

• Ship org exploration
• Align org project cards with the new square project cards
• Test out replication v3 (which would mean we could skip working on org node replication hacks)

[rudolfs]

Test out replication v3 (which would mean we could skip working on org node replication hacks)

Used these branches for testing:

• https://github.com/radicle-dev/radicle-link/tree/rudolfs/rep-v3-fixup
• https://github.com/radicle-dev/radicle-upstream/tree/rudolfs/rep-v3-testdrive
• https://github.com/radicle-dev/radicle-client-services/tree/rudolfs/rep-v3-testdrive
• fixed a missing dependency in the org node setup Fix org-node replication radicle-client-services#40

Overall felt like networking connectivity was improved.

• I could replicate:
  • maintainer Upstream -> org node
  • org node -> contributor Upstream
• contributor changes couldn't be replicated (not supported by org nodes)
• peers could not be found when their device ids are added in the peer selector (not supported by org nodes)
• still had to restart Upstream before new seeds would be picked up
• replicated projects sometimes couldn't be opened, a sign of corrupted state in the monorepo
• waiting room was occasionally spinning in a loop generating high cpu load, but would recover from this state by itself after a while:

Oct 14 14:06:10.140  WARN radicle_daemon::peer::run_state::running_waiting_room: waiting room error error=StateMismatch(Cloned)
Oct 14 14:06:10.141  WARN radicle_daemon::peer::run_state::running_waiting_room: waiting room error error=StateMismatch(Cloned)
Oct 14 14:06:10.142  WARN radicle_daemon::peer::run_state::running_waiting_room: waiting room error error=StateMismatch(Cloned)

Screen.Recording.2021-10-14.at.13.55.05.mov

[rudolfs]

Next one: #2555.