Bitcloud is going to be based on Tahoe-LAFS

[Avi Freedman]

As an observer (user, not coder, of LAFS)...

That said, what you're proposing sounds interesting in many dimensions.  

Is any of it coded?  I don't see any on GitHub, just marketing/high
level docs.

My observation is that even the core coders find new features, much
less important features for scalability, to be time-consuming and
occasionally break things.

(The below is said with respect, if not love...)

There are some important things that have not yet been done to the
core of LAFS:

- making it easy and bulletproof to install
- ensuring diversity of chunk spreading (maybe done or almost done I think)
- verification and rebuild of data by proxy for the client without having 
  to give up the ability to recover to plaintext
- accounting
- multiple introducers
- a few other things I'm forgetting 

Making a demo of addressing maybe one of the above, and building a 
running demo of some of the CA, accounting, mesh networking, and maybe
another element of the 10+ you suggest you'll add would go a long way
to making what you're proposing seem real/doable and would get more
interest than the current approach of innovation by assertion vs
coding :)

The funding model you mention also seems confusing.

Good luck with the project.

Thanks,

Avi

> Let me paste here what is going to be Bitcloud. Bassically it is just a
> super-set of Tahoe-LAFS, a modification of it to become the base for
> DAOs needing to use storage:

> Javier.