What direction are people thinking with interaction with Ethereum. One approach is that it(edit: an Ethereum application) is just something you go to when you want to do something with Ethereum, and Ethereum just provides something like 'Ethereum apps' that help people make interfaces to their contracts.
Another approach, which i like more, is to have Ethereum be a more persistent backend to a browser, and provide two other things; sending data between nodes directly, and putting data in the system.(like magnet links) Gavin Wood introduced this idea as whisper and swarm at a point.(
'The Ethereum experience',
web3.0 posts) The addition of a browser with javascript bindings is a step towards this, but so is parts of the
Douglas project.(Which is currently using Ethereum and magnet links for web-pages)
Their approaches are slightly different though, and generally, there are different approaches to it. The ethereum wiki is closed, so
I wrote in the BitVote wiki about it. So i am interested in what you think. Feel free to edit the wiki. Wikis are for convergent(as opposed to divergent) thinking, though. (You dont need to agree with me, but you shouldnt throw streams of conciousness in there.)
Comments
Reading your publication i think that the best option for deployment will be a server listening on a port (or ports) like in Freenet. So programs can make queries to the program and the Browsers included mobile ones can use the program as a proxy to serve the marketplace or whatever you want to show in the Browser.
http://en.wikipedia.org/wiki/Freenet
They have an API to make client applications (storage, messaging) that connect to that local (and/or remote) server wich is a Freenet node.
One interesting note is that you can add in the installation package of Ethereum a Browser preconfigured to use the proxy, so users do not have to mess with configurations, Tor includes a modified Firefox browser.
The other options are difficult to adopt by some users and you will probably end losing valuable user segments.
I think if the database is is large, downloading that is the biggest limitation, not the computation.. Maybe could use merkle tree mechanisms to allow random servers to prove they're sending the right data upon request.
I'm looking for virtual team mates to help work on customized Dapp "back-end" as discussed herein. I've done the "easy" part of creating a front-end UI that is being reviewed by an executive-level business partner, but - given my Engineering (not software) background, I need to partner with someone with a pretty deep grasp of Ethereum back-end requirements & architecture.
Please do contact me at JJBJManagement(at)Gmail(Dot)Com to discuss details.