Hi all
I would really like to get a feeling for the software architecture of the apps that will be running on the Safe Network. Would be nice to have some drawings depicting it’s structure.
Is all the business logic going to run in the client ( aka web browser)?
How would an MVC design be implemented?
Would be nice to have an example of how our code in the client will interface with the DHT on the network to perform CRUD operations.
Is there any API documentation I can read?
Cheers
Preferrably, yes. It is possible to create a communication channel through SAFE messaging with a server, but serverless apps and sites should usually be preferred for security and privacy purposes. Try to keep the clients in control of their own data and processing of data.
Classic websockets should probably never be used. I have little experience in webprogramming (embedded systems are my thing), but if possible we should probably block websockets (through a function overload with an empty implementation or something like that) for safe: addresses in the browser plugin(s), or at least popup explicit permission requests when they are used.
Much of the logic will be client side definitely, but less than I first thought.
One of the problems with this kind of question - which many of us share - is that there will be many ways to build an app, although in time I expect that to settle down into a smaller number of options around several common architecture/design patterns.
Here are my first thoughts for http://safepress.io:
However, since then I’ve stopped work on the backend side of things (which were almost all in the client as you can see above), to await more details of Structured Data which I understand will allow some level of querying or enumeration (but don’t quote me on that because I really don’t understand what is planned).
There are also questions about how dynamic web apps could be implemented. I submitted an RFC on this but the jury is out because I think MaidSafe may have other (probably better) ideas, and they are too busy with other stuff for the time being.
One of the devs, who I shall not name for fear of incriminating them, did say he expects them to begin clarifying some of these issues in a couple of sprints, which suggests before the end of the year (being cautious).
bringing this back, very thought provoking read from David Irvine regarding applications.
Wow Daniel, how much is laid out in there that I’d completely forgotten. Well worth re-reading everyone. David explaining the economics and functioning of the network all in one blog.
