Better random pick function for the dependency tree

[Ly0n]

At the moment we are gathering all contributors in the dependency tree and pick a user defined of random amount out of that.
I don't think it is the ideal approach. It would be better to randomly pick a project and use the same weight function as in the major project. @kikass13 @cornerman @krux02 What do you think? How can we better deal with dependency contributors?

[kikass13]

@Ly0n @cornerman @krux02
i think the way it's done right now is fine.
but:

I also think that we should be able to run libreselery recursively ... meaning that if project_dependencies: true is enabled, we should search for dependencies and execute other libreselery instances with out wallet.

In case there is no libreselery in a dependency project, we just randomly split stuff because we have no reason to do it otherwise,

In case libreselery is also a tool used by the dependency, we should be able to run it with our wallet and payout information as well, so that the dependency project receives it's share of the main project, but it will distribute the share as it sees fit.

In other words my requested feature is:

• libreselery has to run in a recursive/cascaded manner (master/slave)
  • libreselery can be a master instance, distributing money as it is configured
  • libreselery masters can trigger other libreselery instances as slaves
  • libreselery slaves can be triggered by receiving wallet and payout information from a libreselery master
  • libreselery slaves will use the provided wallet and payout information, but will distribute the provided money as if it was their own (as their own configuration allows it)