Ultimately, we ask that you look at the number and size of wishes we are able to fulfill in the coming months as evidence of our continued support to community wishes and hold us accountable.

So wrote Suman Cherukuwada, Deputy Chief Product and Technology Officer, while attempting to explain the dissolution of the Community Tech team and the change of the Wishlist from a team into a program on 21 May. It's obviously too early to judge this latest incarnation of the Wishlist. However, there is data to hold the Foundation accountable for the decisions they've already made, which have harmed the community and which suggest the Foundation does not understand how to collaborate with the community when it comes to fulfilling community tech needs.

The Community Wishlist was started in 2015 by Danny Horn, then a program manager at the Wikimedia Foundation for the Community Tech (CommTech) team. The team was formed for a straightforward reason:

The Community Tech team is focused on meeting the needs of active Wikimedia editors for improved, expert-focused curation and moderation tools. The creation of the Community Tech team is a direct outcome of requests from core contributors for improved support for moderation tools, bots, and the other features that help the Wikimedia projects succeed.

The process was simple: once a year people submitted their desires, then they voted. Up to 10 proposals with the highest votes were then worked on by the Community Tech team. This also provided a clear accountability measure for the community: how many of those top ten wishes were completed each year. With some minor tweaks along the way, the Wishlist proceeded along this same path until 2020, when the team started to weigh the proposals and votes to determine the top 10. Then from that weighted top 10, the team would research the various wishes before deciding which ones to implement.

After breaking the idea of "vote and develop the most popular ideas", in 2023 the Community Tech team strayed even farther from the original purpose in a diff post called Shaping the future of the Community Wishlist Survey. The Wishlist would become an ongoing process rather than an annual event, people would no longer vote on individual wishes, but would instead vote on groups of wishes called "Focus Areas". Both the focus areas and the wishes that made up those areas would be selected by the Community Tech team without any community input. Unlike the annual Wishlist, accountability would be much harder.

With the benefit of hindsight, this was effectively the end of having a CommTech team whose purpose was to develop features that help Wikimedia projects succeed. In fairness, the Foundation knew that they were trying to do something that was new and not the annual wishlist. This is why there was an aborted attempt by CommTech to change the name of this new system into something other than "Wishlist". That attempt was abandoned after overwhelmingly negative feedback from the community, without really understanding why the community was giving that feedback.

Unfortunately what wasn't abandoned was this new system. The abandonment of the idea of having a team to solve community needs was justified to help:

1. Improve connection
2. Reach more audiences
3. Collaborate with other wishlists

There were real problems with the old annual Wishlist. If some new format could acheive those three goals, it would have benefitted the community in the way that CommTech predicted:

The fulfillment of the objectives is expected to help make the intake process more efficient, and promote subsequent sharing of the load. Bigger technical issues can be mapped into a process for evaluation for the annual plan. The wider visibility of the intake process and collaboration opportunities will allow volunteer developers to get involved in the important problems sometimes at the most local level and can be offered necessary support as needed.

For those who don't understand WMFese, what this suggested was that CommTech was unable, on its own, to do all the work the community needed, especially around work that was large in nature or would require multiple WMF teams to work on it. This new system was designed to let CommTech serve as a bridge between the community and other development teams, plugging community wishes and desires into the annual plans of those other teams, while CommTech would continue doing some work of its own. In theory this would have been great.

What actually happened is that none of the three objectives were achieved.

The Community Wishlist became something else rather entirely, as instead of reaching more audiences, community participation cratered.

If anything, this chart overstates the number of participants by comparing 22 months compared to single years. In the last 12 months, 339 editors either submitted a wish or voted on a wish. In other words, in the last year we saw nearly half the participation as in the first year of the Wishlist and only about a quarter of the participation as in the last year of the annual Wishlist.

For the wishes marked as done in the Focus Area era, CommTech continued to do more work than every other WMF team combined. There was also no meaningful difference in the number of wishes done by volunteers or affiliates. So the improved connections and collaboration with other wishlists didn't happen either.

Nor did the other benefit of allowing larger projects to get completed happen. This can be hard to measure, but the data, when analyzed form several different perspectives, tell a similar story. This chart shows one such way:

The idea of more smaller wishes being worked on seems to be in keeping with some changes where CommTech was more aggressive with marking wishes as completed starting in 2022 where the team began noting the completion of wishes not just through its own work but the work of other teams. What didn't happen then, and what got worse in the Focus Area era, is CommTech taking on bigger problems of the kind that individual volunteer developers would be unable to resolve on their own.

The Foundation does seem to have realized the absolute failure of the Focus Area era. Having realized this failure, a logical option would have been to listen to the consistent community voices suggesting that a formerly productive team should be allowed to be productive again and the community should be invited back into the process through an annual event. Of course, that is not the option the Foundation chose, instead deciding the best way to get more done was to have no team responsible for doing the work the Community has identified to help projects succeed.

Both Suman and his boss, Chief Product and Technology Officer Selena Deckelmann, have insisted publicly and in private discussions with volunteers (including me) that this new format will be superior to both the annual Wishlist and the disaster of the Focus Area era. It seems like they genuinely believe this so deeply they are willing to ignore the community outcry that has resulted from the decision to disband CommTech. It would be easier to give credence to their analysis, as people who are experts at development and who have the time and resources afforded by their position, were it not for the incomplete plan that has accompanied this shift. For instance, both Selena and Suman have promised that there won't be less time spent doing Community wishes. However, there has also been no explanation for how the CommTech worker time will suddenly happen in existing teams and staffing; Tamzin has suggested this could be 10,000 hours worth of lost work. There also was, and as of this writing remains, no plan made about how to continue supporting the tools maintained by CommTech. The best that has been offered is an unpublished list compiled by Selena for wishes that will be incorporated into the annual plan.

Some of the biggest successes from the annual Wishlist, like the fixing of the New Page Patrol tool bar and creation of a dark mode option, were projects that the Foundation had adamantly and repeatedly chosen not to include in their annual plan. Refused, that is, until the community came together with a strong enough voice that the Foundation decided to do the hard and necessary work involved.

The Foundation ignoring the community carries its own risks of course, as has been documented repeatedly throughout our history. Ignoring the community in a process explicitly about partnership between the community and the Foundation really is a new kind of failure. It suggests that there is no actual commitment to meeting the needs of active Wikimedia editors for improved, expert-focused curation and moderation tools... for improved support for moderation tools, bots, and the other features that help the Wikimedia projects succeed. It presents the worrying and distinctly possible future that, when this new system fails, the Community will hear something along the lines of "We have tried everything to make a Community Wishlist work (except doing what was demonstrated to work in the past) and so we are not going to even try any more. Community members who wish to give suggestions can file a phabricator task or learn terminology like OKR so they can productively contribute to the Annual Plan."

This is a particularly unfortunate time for the Foundation to decide that the community should be ignored. Given the alarming drop-off in readers, this is a time the Foundation and the Community should be drawing closer together. Jointly we need to figure out how we continue to promote free high quality knowledge in an era of AI. Under the leadership of previous CEO, Maryana Iskander (and until this debacle Selena) that is exactly what had been happening. Trust was high enough, and the Foundation skilled enough, that IP Masking was rolled out in a way without the community deciding to turn off IP editing. Other teams had found ways to forge productive ongoing relationships with volunteers, including two teams, Moderator Tools and Product Safety and Integrity, that have worked on tools requested by and useful for editors.

That goodwill has been squandered. Actions like this send a signal that the Foundation does not care about volunteers. At a time when decreased readership may threaten our future pipeline of editors, demoralizing current editors is not a risk the movement can afford. Demoralized current editors can easily turn into former volunteers.

The Foundation has asked us to hold it accountable in the future. But we also should be holding it responsible for their past decisions. By that standard, the focus area era does not inspire confidence: fallen participation, lack of meaningfully increased work across the Foundation, and completed wishes are smaller in scope than during the annual Wishlist era. By dissolving Community Tech, not only is a productive team gone, but the Foundation has removed the clearest line of responsibility for community technical needs. So far all that has been offered is promises that the new system will work better. If the new model is real, the Foundation should publish the list of wishes assigned to teams, the team responsible for each wish, the expected timeline, the maintenance owners for former CommTech tools, and develop in partnership with the Community the metrics by which success will be judged. Because "hold us accountable" cannot mean "wait and trust us".