There is an article called The Cathedral and the Bazaar that has been circulating since the beginning of the Linux world . Of course, it is much more about software development than cryptocurrencies and web 3.0, but it is interesting to analyze further.
The Cathedral and the Bazaar
To understand this text, we must first understand what a cathedral is and what a bazaar is. Let's dig a little deeper...
Cathedrals are solid, robust and resistant. A cathedral can take decades to build, and a few more years to make absolutely any changes. Cathedrals also face the time factor, if the interested population decides that the work needs changes, people will have to face bureaucracy and headache to have any real benefit (if they have).
The bazaars, however, are businesses, they have to adapt quickly to reality, otherwise they will simply be useless and crushed by the competition. Bazaars follow all the consequences of a free market.
In the software world, cathedrals are closed, solid, stable software in the hands of a small group, while bazaars are open, decentralized software.
Although enthusiasts in the Linux world unanimously say that the bazaar model, this is not always true. I prefer not to go into details, but for some business models, like centralized banking applications, it's not interesting for everyone to know the structure of the system (no, I'm not a purist).
My definition
For this specific post, I'm going to distort a little bit the original meaning of the cathedral and the bazaar. While the context of being easier to modify is somewhat relative, let's switch to something more tangible: decentralized systems.
Gmail, for example, even though it is based on an open technology (email), is still a centralized system (and is still proprietary), while a decentralized email service, such as those on the Zilliqa network, remains . Let's consider these centralized systems, where the user has no voting power or voice, as the modern equivalent of the cathedral model in the article, and more decentralized systems as bazaars. Can be?
Cathedrals in the era of decentralization
Although are popular today, the most common model still remains that of the great cathedrals. Banking apps are centralized, sales apps are centralized, even Steam and OPENSEA are centralized. Even applications that work with cryptos, which deal directly with decentralization, tend to be centralized. This presents a number of advantages, such as maintaining control over your project and having more autonomy to do what you need. Cathedrals are more scalable too: you can make a cathedral that can hold thousands of people, while bazaars are smaller.
Google, for example, is a phenomenal example of this. By privileging large corporations in their internal and external policies, it ends up generating great resources in the short term, but opens great wounds in users' trust. However, it is a fast and scalable solution, much simpler than each one hosting a server with the videos they want and sharing them with the people they like (of course there are simpler and more practical ways, but we won't get into that topic now).
Bazaars on web 3.0
And the bazaars?
While in other projects, such as Linux, things move reasonably faster than in their closed counterparts, in the decentralized world things are not so simple.
If we take, for example, our own country currency (be it the dollar, the real or whatever you use in your country), it is centralized, and all decisions about the future of the currency depend on this central entity. Even if you have benefits or expenses with such currency and its transactions, you do not have the right to have a say in it.
Now let's take Bitcoin, Litecoin or any decentralized currency. Voting on the coin takes place using your wallet, you decide by updating your wallet to the new protocol or change, and if a large majority of the network votes to accept the change, it will be accepted. This happened, for example, in the split between Bitcoin and Bitcoin Cash.
The truth is that we don't handle decentralization very well. Changes in decentralized models tend to generate benefits for a greater portion of those involved than changes in centralized models, but the difficulty of executing things in a truly decentralized way is immense. You have Bitcoin, right? Does this mean you have a full node wallet? Probably not. And what about your blog? Did you hire Blogger or Wordpress or did you leave a server running at your house? Do you have your own mail server or do you use Gmail? We all tend to think centrally because it's simpler and faster, even if it doesn't generate as much profit in the long run.
Which one is the best?
Contrary to the context of the previous article, both models have advantages and disadvantages, and both models also have different implementation complexities, so it is so difficult to say which is better or worse.
Okay, we have Odysee as an example of a decentralized competitor to YouTube, but would you really prefer to create a right off the bat for your innovative project even if there is no demand for it? And would you really want to spend resources and more resources to create something centralized when creating it decentrally will benefit a lot more people?
There is no correct answer, so much so that today there are many more centralized solutions than decentralized ones. Of course, we are still in their infancy in decentralized technologies, but I do not believe that centralized technologies will simply die, nor that decentralized technologies will ever be a hegemony.
Thinking about decentralized technologies, we have three pillars from which we can only choose two. These pillars are decentralization, scalability and security. These three pillars are the so-called Scalability Trilemma, and although they are focused on the cryptocurrency context, we can see this in other completely different systems, or even compare systems of different natures. Pix, the Brazilian payment system, is scalable and secure, but is fully centralized in the Central Bank entity, while Bitcoin is secure and decentralized, but not easily scalable. There are several other examples of decentralized and centralized applications, but they all run into the same problem: of the three, you can only choose two.
That's why I say, there is not necessarily a better or worse, there is room for everyone, and there will always be.
What is the future?
Although the future is always uncertain, what I believe will happen is that projects and businesses will increasingly stop being cathedrals or bazaars, and increasingly they will begin to encompass a little bit of both worlds. In the past, the difference between a digital bank and a non-digital bank already existed, but today this difference is much more about the origin of each bank than actually functionality and operation. Simply comparing the two no longer makes much sense: where does digital begin and where does it end? And that's what I believe will happen with the popularization of decentralized systems.
Today it is still very clear where the centralized ends and where the same ends. But this tends to change over time. Even though many believe that the boundaries will remain clear and well-defined, I do not believe that, I believe that mixed options and partnerships between centralized and non-centralized services will emerge, just as there are solutions that mix free software and proprietary software today.
Of course I could be wrong and one of the models has a clear predominance in the future, or an absolute success, but my bet for the future of decentralization is mixed systems, where part of the system architecture can come from centralized sources and others can come from decentralized sources.
An example of this is , where the heart of the system is the LBRY protocol, which is decentralized, but the entire part of the frontend and the cache servers themselves is centralized. This encompasses the best of both worlds, allowing for scalability and decentralization. The same is true of cryptocurrency networks and so-called "lightning wallets", such as the Tip.CC wallet, which, despite taking advantage of the protocols of decentralized cryptocurrency networks, are still implemented centrally.
But and you? What do you think? What will be the future of the internet and applications? Tell me there in the comments.
