Wikipedia’s shame | Hacker News Reader

Wikipedia’s shame

salon.com

28 pointsedtechdev16 comments

Wikipedia is chock full of human bots that try to turn it into a relational database or some kind of graph of human knowledge. Just look at all the categories Hitler is in:

- Suicides by firearm in Germany - German founders of automobile manufacturers - Austrian anti-communists - ...

Of course Wikipedia (or rather MediaWiki) is completely terrible for this purpose. It's a completely unstructed raw text storage. You can find date A cited on one page for event X, then go to the page on event X and find a completely different date. It's all just text, and the human bots are overwhelmed with keeping it synchronized.

Of course the human bots will be angry if you spit on them in some dead tree op-ed thinking their valiant efforts toward reducing Wikipedias text to ever smaller isolated atoms are hurting some political ideal.

(Of course the people over at Wikipedia have recognized that the problem is insurmountable the way they are approaching it. They have created WikiData in response: https://www.wikidata.org/wiki/Wikidata:Main_Page)

As someone who maintains the largest Futurama wiki (albeit not as active anymore, but I still host it), I can recognise the problems that Wikipedia is faced with. Of course, at a much much smaller scale, as we only have 3,000+ content pages (which is a lot for a Futurama wiki).

And we have had to deal with the usefulness of indexing articles appropriately. That is; giving them categories (and whether new categories is going to help anyone). But since our wiki is so much smaller, we don't use bots but do it by hand (although, we do have one bot to do some basic cleaning up, but nothing truly intelligent).

And the issue isn't just with Wikipedia or our wiki; it's practically with every wiki (even non-MediaWiki wikis). This has led to the creation of Semantic MediaWiki, an extension that is supposed to maintain more rational relationships of data and make it easier indexable.

But most wikis I've seen using Semantic MediaWiki, doesn't really look like a content site, but rather a data site. Which is nice for the purpose, but it makes you wonder if MediaWiki really was the right choice to begin with.

Note this story is from almost a month ago (date stamp on the linked article is April 29th). Previous discussions on Hacker News:

https://news.ycombinator.com/item?id=5624163

https://news.ycombinator.com/item?id=5663072

As the article says, the parts of wikipedia that stay in the history tend to eventually be corrected. The real worst behaviour occurs around articles where even that history is expunged, e.g. those deleted for "non-notability".

The fact that Wikipedia expunges the history of deleted articles always made me a bit uneasy; it seems like the editors want to escape accountability by sending the articles down Orwell's "memory hole".[1]

[1] https://en.wikipedia.org/wiki/Memory_hole

Wikipedia admins (essentially community-elected moderators, there are thousands of them) have the ability to access the history of deleted articles. Articles are often deleted due to copyright violations, etc., so the history is not shown to the public.

It makes sense to make removed copyrighted content inaccessible. However, that doesn't mean you need to delete the entire history: just hide the part that was copyrighted. And in any other case of deletion (e.g., "non-notability"), the entire history could be retained.

Regardless, its not a memory hole.

You say "community-elected" as if users voted for them; wouldn't it be more accurate to say they're appointed by the existing administration?

No, because the users do vote for them.

http://en.wikipedia.org/wiki/Wikipedia:Successful_requests_f... -- successful elections

http://en.wikipedia.org/wiki/Wikipedia:Unsuccessful_adminshi... -- unsuccessful elections

Salon is just part of that cadre of circle-jerking, highly agended websites (along with most of gawker and huffpost) that regularly fire cannonballs at places like Reddit and Wikipedia that allow the kind of free expression and thought they would rather suppress. At least Wikipedia airs its dirty laundry for all to see, and on balance things generally work out OK.

Actually, reading through the article it never seems 'circle-jerking' or a 'cannonball' fired at Wikipedia. The article explains what happened. It then goes on to discuss how Wikipedia airs its dirty laundry and on balance things generally work out OK - pretty much exactly what you said.

Try reading the article aga....oh never mind.

"highly agended"

I don't think there's much of a clearly defined "agenda" other than pagehits. This sort of trolling is no different than their Camille Paglia: "Sarah Palin, Best Feminist?" or "Ask a Wingnut!" vein of trolling.

Qworty's rant is rather entertaining and even a bit more persuasive than the Salon article that quotes it. Oops.

Here comes a flame war. Hopefully some moderator will move drop this entry to the bottom of this list.

Reply on news.ycombinator.com