Comment on Archive.today CAPTCHA page executes DDoS; Wikipedia considers banning site
TheTechnician27@lemmy.world 23 hours agoI don’t really see it as a complicated issue.
That makes sense from (what I think is) an outsider’s perspective. From an insider’s perspective, here’s the problem:
- Wikipedia has a strict verifiability policy.
- This policy states that “Each fact or claim in an article must [correspond to reliable sources]”.
- This policy is the bedrock of Wikipedia. The project is fundamentally unsustainable without it, and we’re still undoing damage from decades ago when the policy either didn’t exist or was too loosely enforced.
- I’m making a third bullet point because I cannot emphasize enough how much “just ignore it lol” cannot work.
- Hundreds of thousands of articles have citations sourced to archive.today.
- This is despite the fact that the Internet Archive is prioritized whenever possible. We even have a prolific Internet Archive bot that (when possible) automatically recovers citations.
- The Interrnet Archive complies with blanket takedown requests of a domain very easily. Even if we ignore the ones going forward because now both resources are unreliable, archive.today would have untold millions of webpages archived which the IA does not – many of which are used on Wikipedia.
- Archive.today will archive material that the Internet Archive will simply fail to archive because, on a technical level, it’s just better at capturing a static snap of an article (which is what we want). It’s especially true for paywalled articles, which the Wayback Machine is often stymied by.
- This would also make the Internet Archive the only remaining avenue for archiving URLs, meaning Wikipedia effectively collapses if something happens to the IA (granted that’d already be catastrophic with archive.today, much moreso than archive.today’s hypothetical removal).
- Archiving URLs isn’t just some incidental thing.
- Citations are the backbone of Wikipedia. Casual readers might find them comforting to have. Researchers will rely on them. But editors cannot operate without them. We might actually use them more than readers do, because they help us a) check what’s already there, b) better understand the subject ourselves, and c) expand out the article.
- Link rot is so much more pervasive than I think people fully grasp. When I’m writing an article, if possible, I archive every single source I use at the Wayback Machine and archive.today, because relying on the link staying up is objectively a mistake.
- The security archives offer generally just incalculably reduces the workload and mental load for editors.
If you’ve ever tried to add a citation on Wikipedia to a sentence that says “citation needed”, you’ve rubbed up against Brandolini’s law. A corollary is that it’s much, much harder to cite an uncited statement than it is to create one. If you remove archive.today, you flood Wikipedia with hundreds of thousands of these. It’s dampened a bit by the fact that the citation metadata is still there and that some URLs will still be live, but I cannot emphasize – as an editor of nearly 10 years, with over 25,000 contributions, and who’s authored two featured articles – that you’d introduce a workload that could never be done, whose repurcussions would be felt for decades.
Even if you somehow poofed away all that work, there are bound to be tens of thousands of statements in articles you have to get rid of because they simply cannot be reasonably sourced anywhere else. For many, many statements, this is not incidental information independent from the rest of the article; many of these removals would require you to fundamentally restructure the surrounding prose or even the entire article.
It’s hard for me to explain that you just have to “trust me bro” that those people voting “Option C” take what archive.today did very seriously and recognize that either option is going to mean major, irreparable damage to the project. Wikipedia is a lot different from the editing side than it is on the reading one; sometimes it’s liberating, sometimes it’s horrifying, and this case it’s “I could use a hug”.
VonReposti@feddit.dk 22 hours ago
“As quickly as possible” pulls a lot of weight in my statement. Just like when the EU is trying to cut our dependence with US payment providers, Wikipedia can’t do it overnight. The best time to plant a tree was 10 years ago, the next best time is right now.
Cutting ties with archive[.]today takes a long time, but the longer the decision to cut it takes, the longer to the ties are actually cut. It’s all about “make haste slowly”, ie. do a lot of planning on how to actually cut the ties with minimal impact so you can do it when forced to (for example if FBI were to take the servers one day) or when you decide that the independence from archive[.]today is more valuable than the remaining impact of cutting dependence. This could take half a year, a year, or more.
But indecision will at some point put you in a worse position: You are funneling your traffic to a malicious website that actively participates in DDoS attacks by using users’ traffic (including those coming from Wikipedia) to carry out the attack. Indecision can open you up to serious litigation and reputational damage by proximity. Given that archive[.]today crossed the line to malicious activity by misusing their traffic, what’s to stop them from malicious activity by misusing their content? IMO even if you think the integrity of your content and its sources are too valuable (and trust me, I think it’s very valuable) you need to consider this as a warning sign and realise that nothing’s stopping archive[.]today from losing the editorial integrity that you rely on.
So my suggestion, brainstorm ideas that would make you independent: Make agreements with IA to improve retention, roll your own archiver, make a deal with news orgs to show their articles as citations (this last one I actually like most the more I think about it. A good negotiator can call it advertising for the news org and you’ll at the same time not infringe on copyright like archive[.]today is). If you wait until point of no return, the choice has already been made for you whether you like it or not. And worst part is that you’d scramble to find a solution instead of the best solution.
TheTechnician27@lemmy.world 20 hours ago
Editors have been doing this for years.
The IA already lives on a razor’s edge in terms of copyright and is doing everything it thinks it can to push that. Many websites leave the IA be because having free, independent archives can benefit them, but it doesn’t take a lot for a copyright holder to say: “Hey, you’re hosting my IP verbatim, I sent you a takedown request, you didn’t comply, and I’m taking you to court.”
You can’t just “make agreements” for the IA to violate copyright law (more than it arguably already is). They’re already doing the best they can, and pushing them to do more would endanger Wikipedia even worse. It’s not an exaggeration to say that the IA dying would be a project-wide apocalypse.
I’d bet it could be done if the IA went down, triggering a project-wide crisis, but among other things, I’m sure the Wikimedia Foundation doesn’t want to paint a target on its backs. We’re very cautious when it comes to copyrighted material hosted on Wikimedia projects, and this would be dropping a fork into a blender for us.
I don’t think I understand one. The Wikimedia project gets to host verbatim third-party news articles? This is creative but completely unrealistic; you’d be asking news organizations to place their work under a copyleft license for citing on Wikipedia (that’s what we host except for minimal, explicitly labeled fair use material that has robust justification). It’d be a technical nightmare any way you slice it, and logistically it’d be a clusterfuck.
Even if you magically overcame those problems, Wikipedia exists to be neutral and independent, and this “*wink wink nudge nudge * ;)” quasi-advertising deal would look corrupt as fuck – us showing preferential treatment for certain sources not based on their quality but on their willingness to do us favors.
Here’s the thing: we know. This RfC is full of highly experienced editors deciding if Wikipedia is going to amputate. Option A means immediate, catastrophic, irreversible, mostly unfixable damage to Wikipedia. That is something that needs to be thought through, and your suggestions – which are appreciated for showing you’re giving it real thought – reflect that people who don’t regularly edit can’t really, viscerally understand how completely screwed Wikipedia is by this.
Aatube@thriv.social 16 hours ago
It would be just like the extant https://en.wikipedia.org/wiki/Wikipedia:The_Wikipedia_Library.
In the worst case we could just run Megalodon on all the archive.today URLs
TheTechnician27@lemmy.world 12 hours ago
I think you have a very severe misunderstanding of the Wikipedia Library, which I have access to and frequently use. The WML allows active editors in good standing to access paywalled sources.
I can’t emphasize enough how absurd this comparison is. “Solar farms exist; building a Dyson sphere would be basically the same thing. Let’s get to work.”