Community Insights/Community Insights 2020 Report/Platform Evolution

Improving and modernizing Wikimedia’s technical ecosystem to respond to a landscape where artificial intelligence is creating content, rich media dominates learning, and the tools people use to collaborate work easily across a multitude of devices and require minimal technical capabilities.

As the Foundation works to improve the performance and accessibility of Wikimedia’s technical ecosystem, the Community Insights survey tracks changes in developer and contributor satisfaction with that ecosystem, as well as changes in the demographic diversity of volunteer developers.

Volunteer Developers

Volunteer developers are younger and less likely to live in Europe than other Wikimedians, and this trend appears to be increasing. The gender, age, and geographic diversities of volunteer developers are similar to those of all contributors, with several notable exceptions. The developers who responded were more likely to live in Asia and less likely to live in Europe than contributors generally (see Figure 5). They were also more likely to be under the age of 25 (24%).¹ While year-over-year data is needed to determine whether developers outside of Northern America and Europe are being retained, it is notable that volunteer developers who started participating since 2018 (and account for 25% of developers surveyed) were more likely to live in Asia and were half as likely to live in Europe as more tenured developers. Even so, developers disproportionately live in Northern America, Europe, and Oceania.²

Developers were only moderately satisfied with the health of the MediaWiki codebase. A small majority indicated that they find the codebase safe, healthy, and sustainable (see Figure 6). Most also agreed they were confident in the code’s applicability and appropriateness to Wikimedia projects. However, not as many found MediaWiki deployment tooling effective or easy to use. Few demographic differences were found among these measures, though developers in Europe were, on average, less confident in the overall safety and health of the codebase (54%), and newer developers were more positive about MediaWiki code’s appropriateness (84%).³

When asked about the barriers to contributing code, volunteer developers indicated that they most often face a slow code review process (50% experiencing this most or all of the time), though those in Asia were less likely to experience this barrier (26%). Another common barrier, unclear or missing documentation about how to contribute, was cited slightly less often (42%). The least frequently cited barrier was an unhelpful or unfriendly technical community (14%); still, developers in Asia were more likely to experience this (only 31% “never” experiencing this, compared with 62% of volunteer developers elsewhere).⁴

Contributors’ Technical Satisfaction

Contributors are moderately satisfied with the Wikimedia platforms' ease of use. While they were less likely to know how or where to report technical problems that affect their editing (42% agreed, see Figure 7), they reported moderately high ease of use of Wikimedia software. 71% agreed that they have the right buttons or tools to do what they need, 71% agreed that the software works the way they expect, and 69% that the software is easy to use.⁵

We observed many differences in response to these questions by role, region, or identity. Contributors in South America (73%) were more likely to report that the software is easy to use, and newcomers were more likely to agree that buttons and tools served their needs (76%) as well as that the software works the way they expect (80%). Contributors living in Western Asia (78%) or Southern Asia (63%), as well as on-wiki administrators (66%), movement organizers (57%), and the most active editors (54%) were more likely to know how to report a technical problem.⁶

Others were less positive about Wikimedia software. Contributors in Eastern Asia were less likely to agree that the software is easy to use (56%), that they had the right buttons or tools (58%), and that they knew where to report technical problems (35%). Both contributors identifying as women (63%) and those not fluent in English (66%) were less likely to indicate that the software worked as they expected, and contributors without English fluency also agreed with less frequency that they knew where to report a technical bug (45%).⁷

Contributors are less satisfied with WMF’s process of contributing to the technical ecosystem, though they were moderately positive about the quality of that product. Respondents were particularly negative about WMF’s collaboration with communities in the creation of software (see Figure 8). The most highly active editors, and those living in Western and Northern Europe were especially dissatisfied with the Foundation’s collaboration, whereas newer contributors, youth under 25, and those in Southern Asia, South-eastern Asia, and South America were more likely to indicate that the Foundation collaborates well.⁸ Contributors were more positive overall about WMF prioritizing the software that is most needed, though the most active editors and those living in Northern Europe were less likely to agree. Youth were more positive about prioritization, as well as those living in Southern Asia, Sub-Saharan Africa, and Latin America or the Caribbean.⁹

Contributors were more positive about the quality of the software the Foundation produces, especially those living in Western Europe. The most active editors were less likely to agree that WMF-produced software is of high quality.¹⁰