A German court forced the removal of a Wikipedia article’s history | Hacker News Reader

A German court forced the removal of a Wikipedia article’s history

wikimediafoundation.org

140 pointsbonyt73 comments

>This result is disappointing for us, but it is not a notable change in the law

Bullshit. This is a big precedent where Wikimedia will now remove material from Wikipedia if a court orders them to. They even acknowledge that the material in this case was, if not unambiguously true, at least well sourced.

That's a huge precedent. It only remains to be seen whether Wikimedia decide to fight a similar decision made by some other court, perhaps in a less popular jurisdiction like China or North Korea.

Perhaps they mean that this is not a notable change in the law because Germany is not a common law jurisdiction, and changes to the law happen only through the legislature and not through the judicial system. But in that case the claim is trivial, and still equally irrelevant now that Wikimedia have set the precedent that they will obey courts in these cases.

Wikipedia has already been removing material from revision histories for years for many reasons,[1] including legal reasons such as copyright infringement and libel.[2] This is not exactly new terrain for Wikipedia.

I would also say that I think Wikimedia decides strategically which countries to have a physical presence in, and therefore follow the laws of that country, and which ones to stay out of. I don't think Wikimedia will have to worry about a North Korean court decision, because they are not in North Korea.

[1]: https://en.wikipedia.org/wiki/WP:CFRD

[2]: https://en.wikipedia.org/wiki/Wikipedia:Administrators%27_gu...

We should put up a page highlighting controversial, publicized wiki pages as they are historically significant.

Indeed. They even have a verb for it: oversighted.

The WMF has operated in this fashion for years and years; they comply with court orders to remove material.

https://en.wikipedia.org/wiki/Wikipedia:Office_actions describes the English-language policy; German should be effectively the same.

That's what they would like you to think. When Wikipedia gets in trouble with the law it usually complies quietly and under the table. Public "office actions", as they call it, are used when the Wikipedia Foundation wants to shine a spotlight on a perceived injustice.

How does Wikimedia determine which country’s laws it will comply with? Unless they have some physical presence in Germany, why wouldn’t they just tell the Germans to go pound sand instead?

The court's reasoning for why they have jurisdiction was outlined here: https://thewikicabal.com/2019/01/20/waibel-v-wikimedia/

This is troubling because there's nothing preventing a Chinese court from claiming jurisdiction over the page on 1989 Tiananmen Square protests. Of course that request will be thrown out because, well... it's China, but with the direction EU's taking, it's only a matter of time before western European countries start sending more of these takedown requests.

I don't believe they comply with every edict from the likes of Turkey or China. Perhaps it's related to taking donations from Germany, or perhaps they want to be able to lobby its government and the EU without appearing to be a criminal organization.

> perhaps they want to be able to lobby its government and the EU without appearing to be a criminal organization.

It's interesting because they would only appear that way to the EU itself. For ordinary people, especially in the United States (though I suspect elsewhere), Wikipedia is far more of a respected organization than the European Union.

(To be clear, I mean that if Wikipedia says "what we're doing is okay" and the EU says "what you're doing is illegal", I think more people are going to believe Wikipedia. That's an interesting position to be in, whether it's ultimately useful to you or not.)

I think all governments would like to be able to enforce their laws on the Internet. The EU just has a bit more success than most due to the collective size and wealth of its member states. I don't know how to judge what ordinary people want. They seem generally happy to elect politicians who restrict free speech.

Maybe because Wikipedia is a community effort and it would be a loss if Germans could no longer contribute to it without aiding a criminal enterprise. At the very least German administrators would no longer be allowed, because they could be compelled by German courts to remove the entries.

I thought that the history is supposed to be available for legal reasons. It's the only way that contributors to the article are attributed, as required by the Creative Commons licensing.

Edit: quite a bit of the history is unavailable, but the names of the contributors can still be seen. Perhaps that's sufficient. https://de.wikipedia.org/w/index.php?title=Alexander_Waibel&...

If all of somebody's contributions are removed, there's nothing to attribute to them.

They didn't just take away the infringing contribution, but everything prior to 2018-09-28.

Courts typically have power over licenses.

I doubt that it's intending to take away any attribution rights, in this case. Wikipedia could use some other method of attribution, if the history was entirely unpublishable.

If I were running Wikipedia, I would make some effort to programmatically archive online citations, like archive.org does. Imagine some government persona mis-tweets something that leads to article-worthy historical consequences; if twitter deletes the tweet, can that person sue to have it removed from wikipedia because it made her/him look bad, on the basis that the original citation link doesn't work?

There is actually already an ongoing collaboration between Wikipedia and Internet Archive to snapshot almost every single web page to Wayback Machine as soon as they are referenced on any Wikipedia site:

https://blog.archive.org/2018/10/01/more-than-9-million-brok...

The pages are being archived at a rate of 20 million URLs/week.

PS. This is not free. Internet Archive is one of those great online institutions that needs a steady stream of donations to buy the hardware that keeps the project going.

What's going to happen when someone like this professor decides to swing the GDPR hammer at Internet Archive to get the rotted page purged first?

I think thats why archive.is exists. They do not act on deletion requests iirc.

GDPR applies to EU-citizens, so I think if the Internet Archive contains material referencing an EU citizen they would need to comply.

How does the EU can enforce anything over an organization that does not have a legal presence in the EU? At best they can force their ISPs to block the website in their countries, but I can't see how much more they can do.

I believe archive.org does have a physical presence in Amsterdam, which may not help in this context.

They don't - OP is incorrect in their understanding of GDPR.

I think this is the main concern here. The link for the source isn't available anymore, therefore the claim stands without anyone to be able to honestly argue against it. Should someone with poor moral and enough reach wanted to have something removed legally from wikipedia or any other such website, they would 'just' have to take down any other online source. In the age of the internet and knowledge being available 24h, this is concerning.

They already programatically replace reference URLs with links to archive.org for those URLs.

Background:

https://thewikicabal.com/2019/01/20/waibel-v-wikimedia/

Thank you for this: I was able to learn from this article that a good bit of the now-censored history is, ironically, revealed in public court documents. https://upload.wikimedia.org/wikipedia/foundation/1/14/Urtei...

There's also an archive.org page from late 2015 which contains the word "PRISM" among other stuff. https://web.archive.org/web/20150908002221/https://de.wikipe...

This article from Heise is still up as well, and it makes claims about Dr. Waibel's purported ties to the NSA. https://www.heise.de/newsticker/meldung/NSA-Abhoerskandal-De...

Personally I think attacking Wikipedia ought to have severe Streisand-ing consequences.

I would think the Wikimedia would stand on principle and refuse to remove items that are factually true. This capitulation to the whims of globally driven censorship needs to come to an end, and will only happen with an organization with a strong backbone and a commitment to freedom of speech and/or press.

I would prefer them to keep a separate version for Germany which will have additional indication that this page was censored by German government. People coming from other IP addresses should view original page. Surely Germany can't forbid people from the other countries to get some information from servers located outside of Germany.

I would prefer them to force Germany’s hand and make them build a China-style firewall.

A big black box over one of the paragraphs saying that it was censored would be effective.

Or they could block access from Germany altogether, like the Gutenberg project did after another court ruling.

While a good idea, keep in mind that the article says Wikipedia only had one day to take action.

How unnecessarily expensive would that be to maintain how many versions of facts?

There was a court case that determined what was factually true in this case, and wikipedia lost: turns out that even newspapers can present false information.

What would have happened if the information was in an append only log like git history (where a rebase would be inappropriate) or a blockchain (where it can't occur)?

For efficiency reasons, the append-only part of both git and blockchains does not contain the content, but instead a hash of the content. Furthermore, they always use cryptographic hashes with strong preimage resistance (https://en.wikipedia.org/wiki/Preimage_resistance), so just knowing the hash does not allow one to obtain the content (other than through brute force).

Therefore, they might not be able to remove "746308829575e17c3331bbcb00c0898b" from the append-only log, but they might be able to remove the mapping "746308829575e17c3331bbcb00c0898b -> Hello, world!" from the repository, without breaking the chain of the append-only log.

This doesn't really answer the parent's question though. In a blockchain with a client doing full validation, removing/altering the data would cause the that block to be invalid. You'd need to hardfork the network to accept the altered history. I suspect git also checks the integrity of each commit, at least when pulling/cloning, so that would probably cause git to think the objects/packfiles are corrupt.

At least Wikipedia content is still accessible in Germany.

Because of a German court order to regarding some of their items[0] Project Gutenberg blocks German access to _all_ of their books. Now for more than a year.

I sympathize with them but a total block seems a bit heavy handed, punishing all of Germany for a maybe poor decision. Guess this was the simplest way to respond given the resources they have.

Curious to see how this is going to play out in the end.

Back to Wikepedia

> Because of the very short deadline from the legal proceeding—we were given less than one day to take action

What I don't understand in the Wikipedia case is why the deadline was so tight. Seems unreasonable.

[0]. https://news.ycombinator.com/item?id=16511038

Brexit is seeming smarter every day. What started as knee jerk reaction to immigration seems more sensible than ever, for myrid reasons.

At the current rate, it wouldn't be too implausible to see leave win by more votes in the case of a second referendum.

Still, it also provides an interesting opportunity for political parties if they have the guts to follow through, do something pro consumer/voter and stick to their guns. Think about it. A pro Brexit party who said outright that they were not going to be for internet censorship or privacy violations and pushed their views as a counter to the likes of Article 13 could win a lot of support in the tech industry/with younger demographics, and they'd probably reap a fair few financial benefits for themselves and the country too.

And given its the US that's raking in the money with the likes of Google/Facebook/Amazon/whatever rather than the EU, it could make economic sense to work more closely with them than Europe too.

Yeah, British politicians are so much better, they will produce only fair, sensible policies.

Is there any reason that Wikimedia does not mention the name of the person in question? Historically, Wikimedia has been in favour of free speech to the detriment of all else, so it can't be out of consideration to the article's subject.

Probably to discourage a witch hunt. The ire here shouldn't be directed towards the individual, who operated within the German legal system, but the courts themselves. Also I'm not sure what you're basing the latter part of your comment on. It doesn't really seem to match my understanding of the history of the project.

> "the content was in fact defamatory, largely because the source in question had been taken offline"

Is there not an archive.org link for the source? Or some other archive? Or maybe it was taken down from archival sites similarly?

>Three months ago

Why did it take three months to write this blog post?

https://wikimediafoundation.org/2019/03/26/european-parliame...

What does that have to do with anything? OP isn't about copyright.

Looks like Wikipedia only offer dumps for the last 3 months, coincidence? https://dumps.wikimedia.org/enwiki/

The way I read this article in English, were the original article still online, or had it been archived by archive.org and the Wikipedia link changed, then it could stay up. Does that make sense?

The article doesn't link to German coverage so it's hard to learn about the underlying matter.

Still being up does not make it true, particularly if it is in archive.org.

I have no idea what when on, but given the facts as presented in that article, this is possible:

1. Papers publish an article about a professor.

2. Wikipedia repeats was is said in the article, and cites the article as proof.

3. archive.org and others makes copies of the article (perhaps Wikipedia starts taking copied of things it cites - which would be _really_ nice.)

4. Professor notices article and threatens to sue.

5. The papers decides the article can't be defended and so takes it down. If this is what happened copies of the article still appearing on the net doesn't alter the fact the papers acknowledged the article was false and defamatory.

If at this point the paper published an apology acknowledging the article is in error then Wikipedia then IMO Wikipedia doesn't have an ethical leg to stand on.

If they didn't publish an apology didn't this is probably the way it had to happen. They only way I'd change things is I wouldn't not be awarding the professor any damages - if I was the judge I would be saying "it's your own fault for not demanding an apology acknowledging the article was false from the paper".

As it is, Wikipedia left us to drawn our own conclusions from the link disappearing. As I frequent Wikipedia user I find myself in this position all the time and think that's perfectly reasonable, but I can see how the professor and the judge might see things differently.

Wikipedia isn’t supposed to make such judgements. If the article is on archive.org or anywhere it can be cited; articles about challenges to the veracity of such an article can also be cited and discussed.

Well, it depends. The EU doesn't like free speech; archive.org is not exempt (April 10th):

https://www.techdirt.com/articles/20190410/14580641973/eu-te...

Something I've been very unclear about for a long time: to what extent does the EU have jurisdiction over websites hosted and operated in non-EU countries, who don't do any business with the EU beyond serving web pages to EU IP addresses?

Presumably, if China asked Wikipedia to remove something, Wikipedia would say no, regardless fo whether doing so violates Chinese law... right? What makes China different from the EU?

This question applies to GDPR as well.

How is archive.org not exempt? They are outside of EU jurisdiction unless they made the mistake of opening an office over there, which would be trivially fixable by shutting it down.

They are not exempt in the sense that the EU (or individual member states) will start blocking the Internet Archive if they don't comply.

Is there any precedent for EU countries blocking websites due to libel claims? I don't think what you're describing has ever happened.

Cool. Let the EU have their own little mini-Internet in complete denial of the existence of wrongthink.

Wouldn't the same also apply for the Wikimedia Foundation?

I would hope so, but Wikipedia seems large enough to have servers over there.

If the EU can get away with this, what about North Korea and Saudi Arabia? Imagine a Wikipedia that can only say good things about Kim Jong-Un and doesn't contain any pictures of female faces.

Statement from the law firm:

https://raue.com/en/news/industries/media-and-telecommunicat...

What was the supposedly defamatory content?

Remove the server located at EU and denying the access from EU, problem solved.

Why would they need to deny EU access? Let the EU implement its own fascist firewall.

Is it time to decentralise Wikipedia...?

Decentralize means a lot and has a lot of impact. It's not as easy as "installing the blockchain".

Instead, it might be more fruitful to take regular archives of it. Indeed, that is a supported action (wikipedia is intently copyleft) and many people already do it.

Decentralization has been thrown around as a buzzword for the past year.

I can see a strong case for decentralizing Wikipedia. Or a comparable offering.

People say it to seem smart. And then you start asking questions and they look foolish indeed.

For example, decentralize how? At what level? Article? Storage? Editorial? Might as well say all you need to build a car is 4 wheels and some power.

You don't need do decentralize it. You only need to make it easy to mirror content in case it's taken down. Wikipedia has a lot to lose by not complying (site getting blocked, donations cut off, etc.), but unaffiliated third parties don't. Get them to host censored content. Russian users can host German censored content, and German users can host Russian censored content. Combine this with a notice on every page that has content removed so people know when to look for it.

Some like this, with some sort of diff checker, then take the diffs and commit them to decentralised storage + IPFS so they then have no territory.

Germany censoring things makes me nervous. I think they tried something like this before.

Reply on news.ycombinator.com