Learning from WMF


I have been part of OSM for many years now. I wonder if there is anything OSMF can learn from WMF when it comes to governing the project?

It seems to me that there are quite a big difference in how these two organizations are run. Why is that?

What can we learn from WMF?

Maybe it’s because I’m German-speaking, but you don’t mean the cooking pot company WMF, do you? :wink:


At the risk of spoiling a joke, WMF means the Wikimedia Foundation behind Wikipedia, Commons, Wiktionary, Wikibooks and a multitude of other projects.


Surely it’s the people behind https://geopandas.org/ :slight_smile:

1 Like

@pangoSE I supposed we’ve pulled your leg enough now, but you were asking for it :grinning:

Obviously the WMF and the OSMF are really the only two larger global organisations in the space of crowd source data and there are lots of similarities between the organisations, but many differences too. In any case your question is quite justified.

But to understand the joking you need to know that the WMF is widely considered to have completely lost its way and purpose. You might want to start off reading User:Guy Macon/Wikipedia has Cancer - Wikipedia (5 years old now), and then go on to more recent material on running ads in developing countries, claiming that there is a danger of WP being shut down even though it could easily be run for dozens of years with existing funds. Or the millions that were spent over the last couple of years devising a strategy of which the most notable outcome was creating a paid API service with the sole paying customer google already being the the largest donor to the organization.

So while most people would agree that there is a lot that can be learnt from the WMF, it is mainly what -not- to do.

PS: to be fair the WMF does do lots of good things, it is just buried in a pile of …


I think, their practice of trying (and paid for) multiple different ways of improving and innovating the service, and then discarding some, might work for OSMF. Currently OSMF operates from a point of necessity: not enough options, so we have to settle for what we have or can afford.

I thought that (per " Wikipedia has Cancer" linked already above) their need to try and throw away multiple different approaches in parallel was simply due to having too much money, and keeping fundraising rolling to support the size of the organisation for its own sake rather than any technical need?

Well, spending money is a skill OSMF has to learn as well : )

One aspect of the WMF which we could copy at least to a small extent is how they support additional projects besides Wikipedia. Wikimedia Commons and Wikidata are quite large and significant initiatives in their own right (even though they have obvious synergies with Wikipedia), and I find some of the smaller projects such as Wikivoyage useful as well.

For the OSMF, this could mean integrating things like OpenHistoricalMap. There’s a recurring pattern of conflict where people attempt to add data to OSM that don’t really belong (say, former railway lines). These conversations would be a lot less heated if there was some other place for them to put it. And it would be a lot easier for an organization that already knows how to to run the infrastructure for a crowdsourced map database to host another instance than it is for someone to start an independent project + organization from scratch.


Hmm I not quite sure why you get the impression that that is part of the way the WMF works, could you give us some examples?

I think that it assumes too much malicious competency behind this. And WMF is anyway fundraising despite having a massive reserve anyway and making random grants (some really dubious).

So I would not treat it as a a part of some deliberate plan, just regular mismanagement.

Also, why unfinished technical projects would be really needed to justify fundraising? Finished and deployed ones would work even better.

Unless it is “lets make project, ignore outcome” not “lets make project and deliberately fail it”?

Not a comment on WMF itself (I simply don’t know enough) but a reminder that the easiest way to not have projects fail, is to never attempt any projects! That’s not too justify failure but a view that we need to be comfortable with some risk of failure if we want to progress more projects. The key thing is learning and not repeating mistakes; much easier said than done!

While I would agree with the sentiment, this is actually part of what I suggested in the Overture thread, the WMF has been massively criticized by the Wikimedia commons community (and other secondary projects) for barely keeping them on life support.

When evaluating the WMFs efficiency (or rather lack thereof) I think it is important to concentrate on what the involved and affected communities say (and not what right wing conspiracy theorists conjure up)…

IMHO good sources for that are Wikimedia-l - lists.wikimedia.org (naturally there is now and then a derailed mailing list thread that shouldn’t be given too much weight) and the Signpost Wikipedia:Wikipedia Signpost - Wikipedia

In the specific context of this discussion the following may be of interest: this article Wikipedia:Wikipedia Signpost/2022-10-31/News and notes - Wikipedia in the October edition of the Signpost, and this open letter Commons:Think big - open letter about Wikimedia Commons - Wikimedia Commons

1 Like

I think we miss the opportunity of a Toolforge-like (see https://www.toolforge.org/) equivalent for hosting for related projects. This really, really would be worth putting in discussion.

By the way, I would even volunteer to help with the infrastructure as code, setup FaaS, etc. Or at least help people dockerize (prepare) the software for the hosting.

1 Like

The OSMF dev server has historically been a/the place where smaller projects have been hosted. Though in practical terms it is more part of the problem than an answer.

Maybe I misunderstood you, so… at the risk of taking seriously a reply that was meant as a joke:

This triggers me a lot. If the WMF (or OSMF) has tons of money they don’t need then they should either do something useful with it, or save it for future needs and stop asking for donations as if they’re desperately dying.

If I’m understanding this correctly, they received 160M$ in cash donations/contributions between June 30 of 2021 and 2022?

From the “Wikipedia has Cancer” page I got the same impression as @SomeoneElse:

Yes that is correct.

While I don’t have an account there, I’m using it for some API testing, which is fine. I don’t know exactly what problem you’re alluding to here.

That it has been used for some (long term) widely used services instead of those being either run by their devs on their own HW or being migrated to a properly supported (by the OSMF) production service.

Well, maybe that was an issue affecting the old errol server which had been decommissioned recently. The new development server faffy appears to be mostly idle these days. Grafana