Wikimedia Discovery/Meetings/Maps retrospective 2017-03-22

"Retrospective prime directive": Regardless of what we discover, we understand and truly believe that everyone did the best job they could, given what they knew at the time, their skills and abilities, the resources available, and the situation at hand.

Previous action items

(None, I think)

What is going well?

Much better understanding of map server setup thanks to Paul and Guillaume’s work
Much better understanding of deployment process thanks to Max’s work
Maps were deployed to the Swedish Wikipedia yesterday by community members
(Dan) Two of the “must haves” were resolved today. Yay!
We are addressing a lot of small pain points (osm replication being stuck, cleanup of logs, notifications, …)
Working with Paul is great! +1
(Dan) The team rallied around the “Stabilising Interactive Product” task and really collaborated to fill it out. That went significantly better than I expected it would.
Technical debt is being addressed.

What is not going so well?

[1] Interruptions in our rollout schedule, being contingent on complete plans for future support in the middle of annual planning DJG
[3] Maps were deployed to the Swedish Wikipedia yesterday by community members DJG KH EB
[3] (Dan) Progress on the “must haves” is very slow. DJG EB JG
[1] Not really knowing where we are going makes it hard to address longer term issues GL
- Makes it hard to have a product vision/direction, think in terms of user needs
We should re-start to collaborate with analytics/analysis team
The future of maps is still unclear
[1] The future of previous Interactive team members is still unclear JG
[3] (Dan) I am frustrated by the continued lack of a decision by the senior management on the future of maps. DJG MS JG
[1] (Dan) I am frustrated that the senior management has not asked me for my opinion on the future of maps, either as its current product owner or as product lead for Discovery. GL
(Dan) I am frustrated that the senior management let the situation get to breaking point and, well, then break in January.
[2] (Julien) I am frustrated by pretty much everything that has happened since January… GL MS
- Has gotten more frustrating since the disbanding decision was made
[2] (Max) I think maps aren’t really needed across the foundation’s teams: services’ block on deployment and iOS decision to use Apple maps being two latest examples. MS KH
The Visual Editor team has not been as responsive as we would like. In one case, they didn't reply to a task for a month, despite repeated requests, leaving us blocked.
- Has this been individuals, or a larger effort?
- See phab https://phabricator.wikimedia.org/T157996
[4] Getting PHP code reviews from outside our team is a pain GL MS EB JG

What else is on your mind?

(Chris) I think I’m attending the right interactive meetings, but don’t feel like I have much to contribute at the moment.
[2] Intersection of maps team work with iOS apps work, take a more holistic approach? Probably hinges on decisions to be made CK EB
(Chris) I’m really looking forward to the off-site and am hopeful that we’ll be in a place to have some useful conversations - assuming interactive is still a ‘thing’ in our department then. :)
(Julien) On a personal level I don’t know what to expect from the off-site because I don’t know what I may be working on in the future
(Dan) I unsubscribed from wikimedia-l after trying to answer questions about maps on iOS. I’m happier now.
- This does not mean I should not be pinged if something relevant comes up that needs my attention. I do acknowledge that the list continues to exist. ;-)

Deeper conversations

[4] Getting PHP code reviews from outside our team is a pain GL MS EB JG

EB: Talked w/Erik about this, and he agreed to allocate more of his time for CR. Will talk to Stas as well.

GL: I’m also going to look at the work Max is doing

JG: I have a similar issue with the front-end side. I haven’t been reviewing Jan’s search code, and he hasn’t been reviewing maps code. Yuri helps with smaller patches, and Max also. For bigger patches, CR process feels a bit weak within the department, partly because we don’t have much front-end dev capacity.

[3] Maps were deployed to the Swedish Wikipedia yesterday by community members DJG KH EB

CK: Community discussion started on the 8th (community-driven). Last comment was the 13th. Yesterday someone created the task, and 3 hours later it was done.

EB: I learned this could be deployed without our help/consent/etc. I’m concerned that other wikis might look at that and decide to do the same. Is there concern on the team about maintenance load?

MS: Zero additional maintenance load. But it doesn’t feel great that they did this without our participation.

GL: The team was ready to deploy in January, after we took time to stop and reflect. So I don’t think as a team we have concerns about the technical issues.

CK: This may have been good faith where they were unaware of us pausing things.

EB: Should we have stepped into the conversation in that moment? If more tasks like this are filed, it may get difficult.

DG: I share Max’s feelings to a degree. I’m concerned about ongoing support. Might be a pattern where a small discussion happens in a community, and then the foundation ends up responsible for supporting it. I think this was in good faith, although I question the judgment to deploy it within hours, without consultation.

KH: Seems like a clear vote that this work is important and needs to move forward. Good timing to influence the annual plan. It was not ideal, but I’m kind of happy that it happened.

MS: I saw in email that it was moving toward deployment, but then saw it had already been deployed.

JG: Same here. No time to act before it was deployed. I agree with Katie that this shows that communities still want maps. It’s not always the case that communities really want a new feature.

KH: Has something like this happened before, where communities have enabled something before the foundation is ready?

DG: Yes. Some communities really wanted VE and Flow. But it seems stronger/larger with maps than with previous examples.

[3] (Dan) Progress on the “must haves” is very slow. DJG EB JG

DG: The team was going too fast before Yuri left. Hastily, shifting priorities. Now, things have slowed a lot. There are fewer devs, and the visionary is not in the team which is demoralizing. I don’t know if there is anything we can do, but I have been frustrated by how slow it has gone.

GL: Several of the issues have ended up being much larger than we expected. We have been dealing with really big/deep issues. Not small issues. There was a lot of lost knowledge with Yuri, so we have been learning.

KH: Often when things appear to be moving quickly, things tend to focus on the glamorous/fun parts. Since January, we have been working on un-glamorous stuff, making sure everything is solid. Not as obvious, but very necessary. It’s the work part of the work. Tech debt is chronically undervalued, but very important. You should be proud of the progress.

EB: Agree. Lots of tech debt has been addressed, and a lot of learning.

GL: I also lost a lot of motivation, and am not as efficient as I could be/should be.

JG: Yes, we lost a lot of motivation. And Yuri wasn’t just working 8 hours/day, so we lost more than 1 person. Most of the must-have are back-end heavy. That’s where Yuri had a lot of knowledge, so that explains the learning curve.

[3] (Dan) I am frustrated by the continued lack of a decision by the senior management on the future of maps. DJG MS JG

KH: Lack of decisions by mgmt are completely tied to being in the budget process.

JG: I hope they are aware of the consequences for people on those teams.

DG: I agree. It can’t be resolved in the current situation, but leaving the team in limbo for 4 months doesn’t seem like the only option. The answer is understandable, true, and unsatisfying.

KH: I hear you.

Action items

Kevin will help with VE issues
??? Try to figure out how to improve front-end code review
Kevin will look at [2] and [1] voted items to see if they should be brought forward in some form
- [2] (Julien) I am frustrated by pretty much everything that has happened since January… GL MS
  - Has gotten more frustrated since the disbanding decision was made
- [2] (Max) I think maps aren’t really needed across the foundation’s teams: services’ block on deployment and iOS decision to use Apple maps being two latest examples. MS KH
- [2] Intersection of maps team work with iOS apps work, take a more holistic approach? Probably hinges on decisions to be made CK EB