Grant Opportunity: Wikimedia Community Fund:

The advisory committee, with the help of @Martin, is exploring a potential grant application to the Grants:Programs/Wikimedia Community Fund/General Support Fund - Meta . Application are due September 30, 2024. @Martin already scheduled a meeting on September 3, 2024, with the regional Program Officer for North America (this is a mandatory step for first-time applicants).

As I prepare the application I would like to open the conversation on what activities should be funded. I already identified the following resources as starting points

I am looking forward to reading your suggestions.

Overlap concerns with OpenRefine development and the Wikidata and Commons extensions:

  1. Improve error handling and messaging for Wikidata extension and Commons extension. Specifically, schema alignment issues already open? But unsure if/how some of them impact the Commons extension?
  2. Outside of OpenRefine, I have no input, it's Wikimedia's money, until it internally affects OpenRefine or it's maintainers in some capacity.

Questions:

  1. What expectation does Wikimedia have from OpenRefine contributors? None I hope.
  2. So, will contractors be employed through the grant and managed by CS&S for improvements to Wikidata and Commons extensions?
  3. Will contractors be employed for scope to improve OpenRefine error messaging in general (as previously identified by @tfmorris) or only where it is needed to improve error output for usage with Wikidata and Commons extensions?

For question 2, it might be worth clearing up that managing contractors (or staff) is not a service CS&S provides.

@thadguidry thanks for your questions. Yes, the Wikimedia grant will be run through CS&S (which will manage receiving funds and then redistributing them to contractors or employees). The grant Principal Investigator (PI) will hire and supervise the contractors. As we currently do for other grants, that team must collaborate with other OpenRefine contributors, and only the PI and its team would be responsible for reporting back to the Wikimedia Foundation, not all contributors.

We would need to coordinate with Wikimedia Sweden and @Andre_Costa regarding work on the Commons extension. I will ask if this can fall into the grant scope.

Through this grant, I would like to have part of the fund dedicated to the general maintenance of OpenRefine to help maintain a healthy project and community. That can include work on error reporting and schema alignment as well as more general maintenance tasks as described in this message Grant opportunity: Open Technology Fund - #5 by Martin

Does OpenRefine meet the eligibility requirements? The first two paragraphs say:

The General Support Fund (GSF) is open to Wikimedia Movement organizations and organized groups of Wikimedia volunteers, to support the effective programmatic work of those implementing their work in 100% of the time related to and focusing exclusively on Wikimedia projects.

Organizations outside the Wikimedia Movement that occasionally collaborate with Wikimedia projects, but whose activities are not 100% of the time related to and focusing on Wikimedia projects, are not eligible for General Support Fund.

Tom

I received enthusiast messages from different Wikimedia contributors inviting OpenRefine to apply. Part of the September 3rd meeting would be to confirm our eligibility.

We discussed the grant application at August 22, 2024, Advisory Committee only. During the call, we discussed the opportunity to use the funding to create an OpenRefine Wikimedian liaison role, do active outreach, attend conferences to present OpenRefine, answer Wikimedia-related questions, and escalate bug reports and issues. This proposition also builds on the discussion regarding Support / Helpdesk Role during the barcamp and Supporting wiki* communities.

In a separate conversation, @Andre_Costa suggested that we may extend the scope to related questions and confirmed that this call will not cover the development or maintenance of the Wikimedia Commons extension.

create an OpenRefine Wikimedian liaison role, do active outreach, attend conferences to present OpenRefine, answer Wikimedia-related questions, and escalate bug reports and issues

My impressions after Wikimania are that there are already quite a few of us who do such work in a volunteer capacity. It's not completely clear to me why we would need to have a paid person to take over this work instead. Is it that the volunteers are seen as unprofessional or unreliable on the long term?

Let me try to put myself in the shoes of the users I met at Wikimania, who have asked us for features and fixes they need quite urgently. Let's assume they see us hiring a representative to act as an interface between them and the rest of the project. Is our expectation that they can continue voicing the same needs they have been voicing for some years, but to this person instead? I am worried that Wikimedians feel that we are sending this person to them to "manage" their demands without actually acting on them. But that's just my guess!

For me the priority is to ship the features that have been asked for. We understand what the community needs - none of the points that I summarized in my Wikimania report are surprises. We have heard the same requests from various channels, for quite some time, so I'd rather prioritize getting them done.

It can well be that I feel more confident than I should about our understanding of the needs of the Wikimedia community, especially after attending Wikimania - where I met a biased sample of the community, just like in every in-person event. But still, given the size of the team, I have doubts about hiring someone to do work that is currently already being done, whereas our development capacity in this domain is still pretty limited.

2 Likes

@antonin_d: I think all your concerns could be answered if we in the funding application excluded the last of the four suggested activities (escalate bug reports and issues). Would you oppose such an application too?

1 Like

Can you explain why you think removing the last of the four suggested activities would answer my concerns? Would removing those activities from the role somehow help us ship the improvements? I don't understand how that's related.

But in any case: I don't oppose doing such a funding application at all - I have concerns about it, but it really shouldn't be blocking. If various Wikimedians encouraged you to apply, they surely have reasons for it. Go for it!

Perhaps I misread your concerns because as I read them, you were worried that this liaison would not help ship improvements, which the fourth task could seem to imply. So scrapping that would remove any such expectations. It would also address your concerns about a possible distancing effect by making it clear that anyone is free to engage on GitHub and our forum without an intermediate liaison. (The gain could have been that Wikimedians don't need to go there at all, but just communicate in their own channels.)

The other three has nothing to do with shipping improvements, but help both projects in other ways. As far as I can see, these also do not take over any tasks that volunteers are doing, but could be done as an add-on, or where there are no volunteers available.

Thanks, that's clearer. Let me clarify also on my side. I did not think that any of the tasks that @Martin originally mentioned implied that the grant would help ship improvements. My concern was precisely that the grant application was unambiguously not focused on software development. I think we need to create more capacity (be it via funding or other means) for software development in this area. So, making the grant application be more distant from software development does not make it more appealing to me.

Currently we have @Sebastian working on improvements to the Wikibase integration, with a narrow focus on Wiki Commons. That's fantastic, but the grant that funds this work is rather small and will expire soon. I have also made some improvements in that area recently (with the error reporting changes, a byproduct from the reproducibility project), but I'm sun-setting my activity in the project as a whole. That only leaves @tfmorris, who considers himself unfamiliar with the Wikibase integration and reluctantly reviews PRs on this part of the project. So, as things stand, in 6 months we may well not have anyone actively working on this integration as a developer.

So, assuming this grant gets awarded, my worry is that will have a great community manager talking to the Wikimedia community, attending conferences, hearing all the needs and wishes people have for this integration, but no-one to actually make changes to the tool. I am worried that this is going to be a frustrating job for this new hire, and that the community will get impatient. Given the public feedback we got at the OpenRefine meetup at Wikimania from Asaf (that funding to improve the OpenRefine-Wikimedia integration can be found), it's likely that the community will find it difficult to understand why the project does not just do that.

1 Like

Here is my update following my conversation with the Program Officer for North America.

She suggested that our application be between USD 50,000 and USD 70,000. The funds are from the North American fund, but we have no geographical restriction to use them (the hire can be outside of the North American region, for example). If our first application is accepted, we can request a renewal based on the same grant application and budget. After one successful renewal, we will be eligible for the multi-year funding.

To @tfmorris's question regarding eligibility, the criteria are not strict. We need to demonstrate how the Wikimedia ecosystem relies on OpenRefine. Then, the fund must be used to support work on the integration with Wikidata or Wikimedia Commons.

@antonin_d I agree with your reading on the situation. We initially scoped our approach this way because @Ainali @ej2432 and I thought that any technical development was out of the scope of this call (as per this point on the grant page)

  • Proposals that are primarily focused on large-scale software development are not eligible for review. Smaller-scale software development projects may be eligible for review in the Rapid Fund pending review from Community Resources and Technology department staff.

After discussing with the PO, we can allocate funds for a technical position, specifically for maintenance and not new software development. This means we can use the funds for bug fixes, general maintenance, and minor updates. A liaison role will also be in the scope of this call. However, with this new information and the available funds, I believe it would be more appropriate to tailor the proposal for a technical position. We will need to coordinate the application for the Commons extensions with @Andre_Costa from WMSE.

I will start drafting the application in this Google Document

4 Likes

I submitted the grant on Monday, and it is now available on meta: Grants:Programs/Wikimedia Community Fund/General Support Fund/Maintenance of OpenRefine and its Wikimedia-related extensions.

I also open this PR to update the website

2 Likes

Ping @Martin we got a response on the talk page now, and it seems they read the proposal as a development proposal even though we tried to make clear that was not the purpose. We should think of a response that clears up the confusion.

@Ainali, I saw the answer and was thinking about how we could respond. I think the key point is (emphasis are mine)

This work involves significant developer time and makes up the majority of the proposed budget. My conclusion is that in spite of some non-technology activities that have been proposed here, the success of this proposal depends on supporting ongoing development work on OpenRefine.

Two things come to my mind

  • From my call with the program officer, I understood that supporting the software could be part of the budget. As discussed in this thread, since OpenRefine is a software project, it needs developer time the most. Can we better explain that our request is for routine maintenance and not new "large-scale software development"?
  • The proposed budget allocates 60% of the grant for developers (38.57% for the new hire and 21.42% for contractor). Should we rebalance the budget to reduce the software development part? Can we adjust the contracting part and allocate more to the trainer stipends or reduce the total amount requested?

Yes, I think we could do both, but especially the first part. Perhaps there is some confusion of the reviewer thinking that all the work a software developer is doing is development? One thing we could do is to rename the role to software maintainer to make that clearer, besides also in more words show the distinction between development and maintenance.

I don't want to hold you back from responding, but to me the answer is pretty univocal and well motivated, so I wouldn't expect them to change their mind.

I would be tempted to just accept that there is a mismatch between the scope of this particular funding opportunity and the current needs of OpenRefine (which is something we have also identified on our side early on). That doesn't preclude us from applying again to another funding round from WMF in the future, when the scopes align better.

One could try to overhaul the application entirely, to completely change its goals and make it distinct from software maintenance, but I think it would be a bit bold given that they have already reviewed the proposal in the current form. I would also be worried that the proposed activities wouldn't be very well aligned with the needs of the project.

I just answered on Meta.

I also had a chance to discuss our application with User:LWyatt (WMF) - Meta during the WikiConvention francophone/2024 - Meta this weekend. WMF is seeking a more effective way to fund technological projects and is currently restructuring its funding process. There is still hope for securing funds in this funding round or another one in 2025.

1 Like