One of two major outputs of the new search experience project is developer productivity and happiness. For this reason, it is important to request direct developer feedback after completion of the project. This task encompasses the work to build a developer survey, circulate it, and present the results.
The purpose of the survey is:
- Understand the overall workflow happiness and status.
- Identify changes to make.
The following should be avoided:
- Questions that dwell on things that cannot be changed such as the unalterable past.
- Questions can be framed with respect to the project just completed but the answers should generally be framed forward-facing.
== Survey questions
- How did the project go?
- What was worse about this the new workflow compared to the previous?
- Was your developer workflow improved? How did it compare to previous projects?
- During the project, what are some component patterns that are working well?
- What are some component anti-patterns to be avoided?
- How would you rate the overall documentation now available?
- How would you rate the overall internationalization, accessibility, and styling support now available?
- How would you rate the overall ability to develop and use components outside of the MediaWiki stack?
- How would you rate the overall ability to develop and use components outside of the MediaWiki stack?
- How would you rate the overall usefulness of components outside of the MediaWiki stack? I.e., was the code written useful beyond its immediate use case? Is it possible for third-parties using Vue.js to easily consume and use our components? Is the code shareable?
- How would you rate the overall test workflow?
- What is your confidence in the results?
== Acceptance criteria
- [] All developers directly involved in the project are surveyed.
- [] All participants have the option to submit feedback anonymously.
- [] Actionable tasks are created as part of the aggregation and presentation process.
- [] The results are presented in written form to all sponsors.
- [] There are opportunities to provide feedback on low-level technical details as well as high level project concerns.
- [] There are both free-form and rating inputs for every survey query possible.
== Open questions
- The results are published on wiki?
- How are we gonna do this? Google Forms? The biggest QuickSurvey ever? Something else?
- Should we ask someone who's done this before for their recommendations? I think RelEng recently did a pretty neat one. Max probably also has some ideas on running successful retrospectives and how a survey should differ.
- About how many people is this going to encompass? It could be quite a bit of work to aggregate the results and that might shape the survey.
- We should probably ask the sponsors what they'd like to ask us too so we don't miss an important feedback loop they were counting on. Maybe we can just add them to this task and let them edit it.