Wikimedia Discovery/Meetings/Maps retrospective 2016-06-30

Format

This retrospective was conducted using the "Five finger retrospective" format.

Yuri: create a longer-term roadmap/vision
- Discussions happened; needs to be written up. New action item created below.
Max: investigate making puppet manifests more self-sufficient
- Mostly done; waiting for Guillaume to wrap it up
Guillaume: upload (documentation) diagrams to commons
- Not done, will do. Action item added below.
Guillaume: restart conversations about opening up rights on production boxes
- I'm not actually ready for this conversation yet... It will come, but later
Kevin: Create phab task to improve Elastic Search GeoData tests
- DONE: https://phabricator.wikimedia.org/T137257
Kevin: Send email reminder about meeting more frequently but shorter
- DONE (sent 2016-06-07)
Chris: write a draft of "what we learned" from interactive maps conversations
- DONE - https://meta.wikimedia.org/wiki/User:CKoerner_(WMF)/Work/Lessons_learned_from_maps_conversation

Code is improving rapidly!
Launched ENwv, RUwv, community is experimenting at DEwv, FRwv
Got GeoShapes semi-production
Team is collaborating rather than fighting :)
Yuri handling the conversation on Wikivoyage regarding Maps (50+ replies!)
Feeling to be working FOR and WITH the community +++
Announcing on the community traveller's pub was very efficient to get feedback, bugs, feature requests (< hard to track all the bullets and sync the phab tickets, onwiki discussions and actual live deployments all together).
Front end code got a large refactor, WV implementation code also got (is getting) a large refactor. Cleaning it was exhausting but now puts us in a much better state for adding and maintaining features. ++
Good rework on the productionizing! (Жульен & Max)

Every little change with ops requires a week-long discussion
- Concerns about balance between autonomy and top-down control
- Would prefer isolated maps servers to avoid concerns about impacting other services
- We are in a confrontational relation with Ops. This needs to change, but I'm not really sure how. I'm sure there is work to do on both side though...
Checking participants' calendars before scheduling meetings
- (This was in reference to a recent case of double-booking)
hard to track all the bullets and sync the phab tickets, onwiki discussions and actual live deployments all together (copied from "thumb" above)
- CL could track and manage the wiki conversation, and make sure actions become phab tickets
- Chris did great managing the dialog with community earlier...the work continues, and Julien is feeling pressure to do (some/all) of it
- Would be great to have Chris officially assigned to geohack work

You all might not see it much, but having a team of liaisons I (Chris) can talk to is like having a (small) army behind our work :) ++
It is sometimes hard to collaborate with other teams (e.g. outside Discovery) simply because everyone is busy with their own thing
- We should be slightly more proactive working with other teams
- Re-reading SoS notes helps. Should doing so be mandatory?
Do we have any clue of the content (and purpose) of Katherine's keynote at State of the Maps US?
- Not to my knowledge, Yuri will try to talk to Juliet
- Yuri need to prepare his own presentation

I really appreciate Alex Stinson's input on how the GLAM folks could use Maps. We should keep examples of uses cases in mind (always).

Kevin: Invite Guillaume to next maps retro
Yuri: Publish long-term maps roadmap
Julien: Make sure work progresses on the State of the Maps presentation
Julien: Follow up on documentation improvements
Guillaume: upload (documentation) diagrams to commons (carried over)
Kevin will schedule a meeting to discuss ops/maps relationship
Kevin will schedule a meeting with Yuri and Chris to figure out wiki conversation <-> phab tasks
Kevin: Propose making reading SoS notes "strongly encouraged"
Yuri might email Discovery asking for examples of maps uses that he could cite in the presentation