Decentralised Processwire and the blockchain?

[Vlad]

Hi everyone,

I'm quite interested in building a directory platform and Processwire seems like a good choice with plenty of flexibility. I'm more in the concept phase and not going to do myself the coding, but want to understand the limitations of technology for now.

With the last developments in the blockchain technology and trend of going towards decentralised apps, I'm more and more attracted by the idea of reducing server costs and dependency through this decentralisation and was wondering if there's any way of adapting Processwire towards something like this. I know blockchains in general have their own language but maybe theres a way of doing it via the IPFS protocol? I'm not a programmer so don't throw with rocks if I ask dumb stuff, just wondering if technically is feasible. Or it has to be developed something from scratch based on emerging blockchains like RChain, Eos, Synereo... where most of the content is stored on the users side. 

Thanks!

[LostKobrakai]

What's your usecase, which a blockchain would solve? I doubt it's anywhere near worth the time/complexity if you don't have a specific problem to be solved by this technology.

[Vlad]

Some I can imagine would be is: lower transfer fees between the users of the platform, security since they send back and forth sensitive blueprints of products, and lower storage costs. Maybe you're right and I'm just overcomplicating things, but for example when you have international payments the fees are absurd for transferring basically what's bits of info. An internal currency for transferring value would be much faster and cash only when they take it off from the platform. 

If you implement a Stripe payment you get in the best case scenario with huge volumes 1-2% fee, right? As for beginning 3-4%. 

The downside now it's the adoption of coins that is pretty low but i can imagine if you offer 2 versions in time users will switch to the cheaper one. Some blockchains run on 0 transaction fee, so that's a huge advantage since your users have transfers of few hundred dollars generally. Also the platform would self sustain and expand together with the number of it's users. For now what's the method, going with Amazon web services? 

[LostKobrakai]

I still don't fully understand what your platform should offer in the end, but to be honest I'm even more curious what you find so appealing in processwire compared to what you're writing about using a blockchain. ProcessWire is a quite nice, but comparatively simple CMS, whereas a blockchain implementation is a highly complex thing for decentralized and trustless computation/storage. The latter is something where experts develop a non trivial amount of time to just get it running (maybe not for the blockchain per se, but also the implementation and business logic around it). 

6 hours ago, Vlad said:

Some blockchains run on 0 transaction fee

You should not miss one critical factor in the blockchain equation, which is the computational cycles needed to calculate new blockchain hashes. That's the reason, why this stuff does work, because it's so (mathematically proven) hard to abuse/break. 

So I'd really suggest you to ask yourself, what benefits you'd get from using a blockchain, because building that will be an investment and it'll certainly limit your base of potential customers if you don't fully hide the blockchain technology from them (e.g. most likely run your own servers again).

[Vlad]

I agree with you that Processwire is quite simple and easy to take off with, probably that's one of the main reason I consider using it, because even with my limited knowledge in programming I can get a minimum running product to test it's appeal to the audience. Why I thought about blockchain is mainly due to the payment system, somehow in my logic would build some sort of trust in the platform. But probably i'm thinking too much ahead and should test the idea first, and if it takes off i can implement the payments later. 

[LostKobrakai]

Using blockchain technology is a way to not need trust in a distributed system. In a blockchain cluster you still don't trust instances of other people, but the system can weed out people trying to abuse it (more reliably the longer data was stored). Without the distribution in the equation it's just an overly complex immutable data (/event) store. So testing your core business idea first is probably a good idea. Also if you can avoid distribution you'll save yourself a lot of complexity in building the product.