WMF projects

[WikiWikiWow]

Let's talk about the other projects WMF runs in order to justify their $177M budget.

Commons - Free media repository
Out of 116 million media files, 13.5 million are pictures of cats. Editors uploading pictures of their pets, no doubt. There are 2.4 million pictures of cars, 600k pictures of the sea. 16 million pictures in three categories (I'm sure there's some overlap).

Most pictures aren't even original works, but uploaded images from the public domain or sites with properly licensed media. It is a repository of images available somewhere else, whose primary purpose is to host images for Wikipedia.

Wikisource - the free library thay anyone can improve
I don't understand the point of this website. They only have 638k "texts". They're just transcripts of public domain literary works. archive.org, Gutenberg.org most likely have them already, (annas-archive and libgen definitely have them).

Wiktionary - The free dictionary
Why? When OED, Cambridge, Merriam-Webster, etc. exist. I'd trust the OED before I trust some Wikidiot with an axe to grind against a word they hate.

Wikiquote - Free quote compendium
Useful resource if you're in the motivation poster business, I guess.

Wikibooks - the open-content textbooks collection that anyone can edit.
"97,000+ pages" (sounds better than "3,309 books". A failed project where most articles are over 3-4 years old. Turns out, nobody who has the knowledge to write/contribute to a textbook wants to write for free, and those who wish to write for free usually don't have the knowledge (or ability) to write a textbook.

A featured book, "C programming", has only four chapters/sections marked as complete (two of them introductory).

Wikinews - Free news source
Again, what's the point? WMF made the mistake of thinking that as there were hundreds of thousands of people eager to edit an encyclopedia, there would similarly be thousands of people who want to write news reports. By Wikipedia's own standard, anything not reported on by a newsorg is not notable. If a Wikinews story isn't based on one or several already existing news reports, then it's original reporting nobody cares about. Anyone who is able to do actual original reporting will just send/sell the article to a real news service.

Wikivoyage - Free travel guide
On the face of it, this seems like a good idea, right? Locals can write about their city/region/country, give you insight and information you can't find anywhere else. Wrong!

I just checked the article for my hometown. It's all composed of freely available info about transport to/from, and then at the bottom a list of restaurants and hotels (most of them updated 3 years ago). It's just like any other generic travel book, promotion/ads included. I imagine all the restaurants/hotels were added by the owners or employees themselves.

Wikitravel can't be what it could be because all the info must be "verifiable", which usually means a link to a website or a published work. Then what's the point of wikitravel when I can find the same (and up to date) information with a google search?

Wikidata - the free knowledge base
But actually just a repository of bot/machine-readable data regarding stuff hosted on other WMF proejct sites.

Wikispecies - free species directory
There are actual taxonomy websites and catalogues that are maintained by actual scientists. They boast 904k entries.

Wikiversity - free learning resources
A bit late to the game. Khan academy, EdX, Harvard online courses, MIT open courses, and so on. There are many free learning resources out there.

I guess one would go on Wikiversity to read all the amazing contributions to learning editors have made. Like on the Wikiphilosophers section, where "User:Bobshimdt believes that all was made by mind, and all is illusory." Wow! I can feel the learning, I can feel my knowledge increasing...

Wikifunctions - free library of functions
Made obsolete by github, stackexchange and ChatGPT.

[WikiWikiWow]

Why maintain all these projects? For whom? At the very least Wikiversity and Wikibooks could be combined into a learning platform. Or combine Wikibooks and Wikisource, since they seem to perform the same function with slightly different focus.

Looking at the WMF budget, an answer presents itself. They spent 49% of their $177M budget in 23/24 on "infrastructure". 101M on salaries and benefits in 2023, 106.5M in 2024.

If the WMF only had Wikipedia/Commons, they wouldn't need so many employees. Their internet hosting expenses hold steady at around 3M y/y. In 2022 they had 711 staff. What the hell do they need 711 employees for? I'm guessing there's a lot of programmers/systems engineers. For comparison, Valve corp has 336 employees and they manage to have high-availability DDoS proof servers worldwide with millions of people playing simultaneously, while developing new games and content.

All of Wikipedia can fit on 110GB, you can host that on a single computer and use Cloudflare CDN to deliver that content anywhere on the planet quickly. Yet it is a bloated corporation that would die as soon as the corporate donors stop funding it.

Apparently Wikipedia has meetups, "Wikimania". I never knew this, as I have never seen them advertised anywhere, get newsorg coverage (ironic) or have any sort of lasting impact.

The fact that alternative Wikipedias exist that mirror WP on a tiny (tiny tiny tiny) fraction of the hosting budget, while providing the same editing/community features, speaks volumes about the waste going on at WMF.

[ericbarbour]

We know all this. I wrote it up in-depth for the book and posted it on the book wiki.

And that was more than 10 years ago, yet their practices have changed very little--and I suspect the database's shortcomings have also changed very little. A hundred dollar bill says that this study, if rerun again in 2025 using several thousand random articles, would look very similar.

comparisoncontentsWPvsEB.png (338.88 KiB) Viewed 642 times