My take on David Johnston's DApps paper

Kent

Regarding the original paper, as seen here: https://github.com/DavidJohnstonCEO/DecentralizedApplications/blob/master/README.md

Overall this is a good read; it contributes to the ongoing discussion of DApps, DAO's, and DAC's, which is evolving before our very eyes.

The following is an excellent description of Ether - and indeed, most all tokens in the blockchain world.

<< From a technical perspective, those issuing tokens as part of a crowd-sale are selling access to software for the users of that software. The private keys associated with the tokens that the users purchase are literally the passwords that the users need to access a DApp’s software. >>

His trifurcation of the DApp types also makes a lot of sense:

<< Type I decentralized applications have their own block chain. Bitcoin is the most famous example of a type I decentralized application but Litecoin and other “alt-coins” are of the same type.

Type II decentralized applications use the block chain of a type I decentralized application. Type II decentralized applications are protocols and have tokens that are necessary for their function. The Master Protocol is an example of a type II decentralized application.

Type III decentralized applications use the protocol of a type II decentralized application. Type III decentralized applications are protocols and have tokens that are necessary for their function. A hypothetical Cloud Protocol that uses the Master Protocol to issue ‘cloudcoins’ that can be used to acquire cloud computing services would be an example of a type III decentralized application. >>

However, based on that classification, Ethereum is mischaracterized as a Type III DApp.
Ethereum, by his definition, falls into Type I; it has its own blockchain. A DApp running on that blockchain would therefore by a Type II.

Please note my intent isn't to delve into semantics in a field where clear, universally-accepted definitions are yet to be defined, but rather to accurately fit this emerging platform into a clear schema where everyone has a clear idea of WTF is going on.

Finally, David asserts that...

<<DApps have the potential to become self-sustaining because they empower their stakeholders to invest in the development of the DApp. Because of that, it is conceivable that DApps for payments, data storage, bandwidth and cloud computing may one day surpass the valuation of multinational corporations like Visa, Dropbox, Comcast, and Amazon that are are currently active in the space.>>

I sense a big leap here...something like: 1.) Build a DApp 2.) ??? 3.) Become bigger than Amazon

We as a community need to more clearly articulate how these DApps, DAO's, and DAC's will improve the life of the average human being on this planet. This is a necessary precondition to becoming bigger than the Dropboxes of the world. On a related note, it's incongruous to refer to a giant centralized organization, a la Amazon, as the end state of a DApp. I have a feeling we're headed for something that's both bigger AND vastly different that the centralized beasts we've grown accustomed to dealing with.

StephanTual

I'm writing an article about just that, will post the link here when I'm done (warning: I'm a slow writer )

VitalikButerin

I don't agree with the trifurcation; I think he's missing the very important category of dapps that do not use the blockchain at all (eg. bittorrent), and how some dapps use the blockchain for some things and not for other things. Also, dapps do not need to have tokens; you can have a dapp written as a charity project, and you can also monetize a dapp by selling the namespace. So he got lots of things right, but also tries insufficiently hard to generalize

bcurdy

Thanks for sharing! We are still defining concepts indeed and it is really interesting to read how other people think about them. For now,it is quite clear that more abstraction work is still needed. It feels as if Bitcoin was used to generate a definition of Dapps and then used again as an example of it. And there are indeed some massive leaps in reasoning.

My only true regret about the paper is that the author didn't follow bcurdy's law: "Every time you feel like naming a statement after yourself, don't"

Anyway, looking forward to reading what Stephan has to write about this. And by the way, I've seen that there's a philosopher in the Ethereum team. He is writing about those topics somewhere ?