https://www.reddit.com/r/TheDao/comments/4ljzic/can_the_split_dao_reward_token_mechanism_be/d3ociic"Let me try to frame the problem now in an abstract fashion. I may be totally wrong in this analysis, but hope this produces a useful conversation for the DAO community:
"The measurement of 'investment done by a DAO' as the 'outflow of funds from the main DAO contract' might be bound to be fraught by troubles. If this is indeed what the DAO 1.0 framework does, then it will reflect itself in a wide range of attacks / community fights when DAOs split
Some examples:
1)
Not every investment is revenue generating; some are maintenance oriented: Examine the Slock IT security proposal and one realises that if the DAO accepts it, then it is not producing a 'return generating asset'. It is producing a 'maintenance cost' for the DAO. Hedge funds and VC firms have a distinction between 'maintenance costs' (cost to analyse investments, maintain funds, communicate with LPs etc. and 'actual investments that can produce returns' (for example an investment in Uber). Hedge funds and VC firms logically strive to minimise 'maintenance costs' and maximise 'actual investments that can produce returns'.
The current method of DAO accounting lumps 'maintenance costs' and 'actual investments that can produce returns' into just one bucket i.e. 'investments'. This is fine as long as the DAO has never been split.
If the DAO splits, the child DAO obviously will not want to treat the maintenance costs of the parent DAO as 'investments' made by the parent DAO. They will want their own reward tokens to be diluted only when the parent DAO makes 'actual investments that can produce returns'. After all, why should the child subsidise the the maintenance costs of the parent when it has split?
I see no mechanism making this distinction between 'maintenance cost' and 'actual investments that can produce returns' in Framework 1.0.
2)
The firing and rehiring of contractors: The DAO framework explicitly encourages the firing and re-hiring of contractors incase quality of work produced is sub-par, ether volatility changes project economics etc.
So, assume the DAO commits 1 million ether to a contract today. This is the first and only investment it has done. Reward tokens corresponding to these million ether are generated (106 * 1018). Now, the DAO splits and the childDAO must receive 40% of rewards from the contract and the parent 60%. The contractor spends 0.1 million ether and then is fired returning 0.9 mil ether back to the parentDAO. The parent then renegotiates the same contract for 0.1 million ether but with a different address controlled by the same contractor (maybe the price of ether has jumped 10x). How does the reward token managed account upgrade to reflect the new reality? Who gives it this new information?"
