There are many virtues to decentralization, but reducing operating costs and improving performance (assuming equal costs) are not really among them. Some things appear to do so, but only because they are externalizing costs (e.g. shifting costs onto the most expensive last mile ISP paths, which is only 'cheaper' due to unsustainable all you can eat pricing, resulting in blocking/rate limiting, or hosting things on stolen resources).
Well, Terabytes of data are exchanged every day in P2P file exchange systems,
and yet I have never heard of any donation campaign for them.
Wikipedia costs a fair amount to operate, but only because it's so widely used. It comes down to something less than $0.05/yr per monthly visitor (actually more like, 0.02 but wikimedia is spending more than strictly necessary in order to develop new software and increase Wikipedia's usage around the world). The challenge for Wikimedia is that most of their visitors don't donate at all, and payment processing overheads would make operating on 400 million 5 cent payments still insufficient.
First, I want to say that I love Wikipedia. I contribute on the french site, I donated about 10EUR few years ago as I think it is the best thing that came up from Internet.
However, I will NEVER ever donate anything again. Because there are two many things I don't like with this project. For instance, there are several Wikipedia editing rules I just don't agree with. That's one problem with centralization: you must agree with the whole package or go away. If you don't agree with one part, you can't use the whole thing.
I also don't consider storing pictures to be important for an encyclopedy. And yet I assume this is one of the big cost for wikimedia. So I may accept to donate to "wikipedia", but certainly not to "wikimedia". Unfortunately, the wikipedia project is not separable from the wikimedia project.
I use to argue that decentralizing Wikipedia (while retaining the property of having a single coherent website people could browse, rather than e.g. a forest of never updated forks, preserving user privacy, etc)
What is wrong with diversity??? There is no reason why there should be a *single* encyclopedic wiki. An encyclopedy could have a "editorial line". There would be nothing wrong with that. On the contrary.
In Eric Raymond's metaphor, Wikipedia moves much more toward a "Cathedral" than a "Bazaar". And that's the main thing I don't like with this project.
... was actually impossible, but the bitcoin distributed algorithm disproves that. That said, many people complain about the time it takes to sync a full bitcoin node, it would be far worse with 400 gigabytes of article history and 12 TB of image data. So even ignoring the fact that it would cost more in total to operate it's not easy, and the required technology simply hasn't been built.
Yeah that why I think we should not fund old technologies with charity, so that people can have an incentive to imagine better technology.
Of course, you can go to download.wikimedia.org, and pull a static dump and serve it up via a CGI and people do, but that isn't the same as decentralizing the sites* it's not getting the hundreds of updates per second, it's not providing a single, coherent, usable, and reliable view of the site. The prior attempts (for which there have been many) have all been laughable.
reliable: ok
usable: ok
coherent: ok
single: WTF?
I'm sure if some bitcoiner wanted to build the technology to enable this they could find many patrons in the bitcoin and Wikipedia communities to sponsor their work. Sadly, it seems that on this subject most people are all talk.
Wait a bit. Some people will create a decentralized collaborative encyclopedy. But it sure won't help if people continue to support and fund wikimedia.