Topic on Project:Village Pump

@Emu @Gymnicus @Wd-Ryan @Rdrg109

Recently at Wikidata we created items for all of the Wikimedia developer teams under Wikimedia Product (Q86030252) and Wikimedia Technology (Q27983251) and sitelinked them to their pages on mediawiki.org. We also started creating items the team members of these teams and linking them to their teams. The intention was that they could be used to display the team members in the Template:Wikimedia Team Info infobox template without requiring the text to be manually added and would also then be queryable. I particularly wanted to generate a tree of all the Wikimedia teams and their developers with this data documented.

However, documenting data about Wikimedia is currently in a grey area regarding its notability on Wikidata. This is demonstrated by the recent deletion request to delete some of the Wikimedia developer items.

What would probably make more sense is for mediawiki.org to have its own Wikibase instance so that these development teams and their developer teams could be properly documented where they most-closely belong. That way we don't have to worry about notability issues on Wikidata.

Could this be done?

I don’t have any experience in Meta stuff but as a layman, this would probably solve this conflict pretty well. Links to some more context here: :wikidata:User:Emu/Notability#Wikimedia-related_stuff

Any other ideas of what we could add to the Wikibase instance? I like the idea of it.

We could document anything related to the Mediawiki tech stack! Mediawiki documentation and data from Template:Extension could use the Wikibase!

This sounds great, count me in for extensively modeling everything about Mediawiki.

Tbh, im not really sure how mediawiki.org would benefit from this.

I am in favor of putting the data in Wikidata. If for some reason not all data can be in Wikidata, then much of the data already is in Wikidata and people will spontaneously add more anyway. I set up https://www.wikidata.org/wiki/Wikidata:WikiProject_Wikimedia_Foundation to advance the discussion about how the Wikidata community can manage this content.

Just as a reminder, while WMF is very prominent in mediawiki, mediawiki has many different groups involved (as well as unaffiliated people) not just foundation employees.

Using my volunteer account here, because all I'm saying is what I've learned as a volunteer, or it's public already, and I'm only sharing my private view on this.

I think I understand why you're doing this. I respect that. Several years ago, I'd do something similar. There's a chance that the past me shares the motivation. Now, however, I'd strongly advise NOT to try to document all this. Let's not go too much down the org chart.

• Individual people come and go. They change teams and their roles, sometimes they work temporarily in different teams and roles, they join and leave the Foundation. Even if they contribute to public projects, not all of them are public figures.
• Similarly, teams are created and disbanded, but also over time, they evolve. Their scopes and names are changed. This is due to plans, strategies, leaders' will, team members' will. This isn't consistently documented on wiki.
  • Examples: Growth in 2014 and Growth in 2022 are two totally different groups with different stories. The same naming is like a coincidence. Mobile Web and Web - that's more like an evolution. Contributors in 2017 and Contributor Tools in 2022 are two different concepts, not even different teams (because the current Contributor Tools is a group of teams, not one team).
  • Example of d:Q27983256: I'm not sure if the former Contributors team at the WMF might have used the name Wikimedia Contributors. Did it ever, btw? It never was a "department" equal to the Product department. It never was the same as the Technology department either.

In practice, documenting the WMF departments and C-levels is doable. (Although don't ask me if VPs/directors reporting to the CEO when there's no C-level in their department count as fractional C-levels themselves). But documenting all teams and team members correctly and consistently - nah, that's unrealistic.

@tar We are talking about an organization that spends ~US$1 billion every five years. It is entirely reasonable for our community to ask basic questions about where this money goes, what teams we have, what products they develop.

We have this superpowered tool, Wikidata, which sorts knowledge with clarity and intelligence beyond human capacity. Data entry into it is routinely achievable with minimum wage untrained labor. It seems like Wikidata would solve so many problems where currently we are storing the information in the brains of some highly paid humans who have a conflict of interest in identifying, discussing, or addressing some of the challenges we face.

The problems you identify with Wikidata are valid, but for those very issues that you identify, the current path to accessing that information is that a human has to draw you into a conversation where you communicate it with your fingers typing or tongue speaking. Are you arguing that human to human communication is a sustainable, affordable, and reasonable alternative to this amazing database infrastructure which we designed seemingly for cases just like this? Or is your position that there generally is not knowledge here worth sharing in new ways?

@Bluerasberry I take your point but I don’t think it’s a good idea to use Wikidata to solve all of the world’s problems. From what I hear on DE.WP, the WMF is essentially accountable to nobody and about as transparent as a 10 inch concrete wall (slightly exaggerated for dramatic effect) but doesn’t that mean that there not only is no independent outside coverage but we also have problems with the “inside coverage”, too? And our data collection would mostly rely on sparse documentation and a lot of guesswork, combined with WD:BLP issues raised on every corner? As much as I value transparency, Wikidata is maybe not the right venue for this.

@emu you say Wikidata is not the venue @bawolff you say that mediawiki is not the venue. Fine, that hardly matters. The data could be in any Wikibase instance - here is the registry. https://wikibase-registry.wmflabs.org/wiki/Main_Page

Data sharing of the Wikimedia Foundation merits more conversation and a serious response. If someone wants to argue that the data is too dangerous or misleading to share then that is a interesting claim worth exploring. This data does not seem expensive or challenging to share. There is also Wikimedia community demand for it, and even if not shared officially, the Wikimedia community is already haphazardly developing it on-Wikidata now. It would be nice for the WMF to share a complete dataset and give some cooperation about hosting it somewhere, because significantly more WMF transparency in data seems inevitable. Does anyone know of a senior person in WMF who has been or would be brave enough to publicly state opposition to converting the team, staff, and product information which is already public from prose into structured data?

No one is asking for new information; we all are here just asking for plain text posted into Wikidata or a Wikibase instance.

To clarify, I weakly think mediawiki is not the right place (By which i mean, I probably wouldn't be actively editing it, but don't let me be the one to stop you). My main objection is that keeping all this info up to date seems like a lot of work, and normal pages on mediawiki don't really benefit from this as far as i can see. I think ideally this would be kept in wikidata, although i guess that depends on what they think of that.

I do think we should keep in mind that these are real people, who do have a certain amount of rights to privacy. However, they are contributing to a public project, so i think keeping track of org charts is super reasonable and not a privacy violation.

I'd worry a bit you are over-indexing a tad on the "formal" org chart. Formal org charts often don't reflect the reality on the ground, in any organization. I think this applies to WMF more than most. The power structure of WMF is complicated with both formal position as well as informal influences. I guess there really isn't an alternative if you want to track things, but you can run into the trap of collecting the data that is (relatively) easy vs the things you actually want to know.

I suppose my hesitancy comes down to - I'm not sure what you aim to achieve with this project, and i don't think it will bring the transparency you desire.

> Does anyone know of a senior person in WMF who has been or would be brave enough to publicly state opposition to converting the team, staff, and product information which is already public from prose into structured data?

I highly doubt anyone at WMF really cares. If anything they would probably be happy about this so they could make sense of their own org chart.

Here's another page that could be easily generated with Wikibase data: Developers/Maintainers

Well, the Wikidata items all got deleted, so I think we should definitely consider this.

Well, the Wikidata items all got deleted, so I think we should definitely consider this.

Presumably there was a discussion that led to them being deleted? Can you link to that discussion so the wider community can participate in a single conversation about whether to track this in structured data, and if so, where?

@Jdforrester (WMF) it was linked in the original post, but here's the archive now: https://www.wikidata.org/wiki/Wikidata:Requests_for_deletions/Archive/2022/11/20#Wikimedia_developers